Keywords: Autonomous systems
Abstract: The complexity of modern autonomous systems has grown exponentially in the past decade. Today’s control engineers need to deliver high performance autonomy which is safe despite environment uncertainty, is able to effectively interact with humans, and improves system performance by using data processed on local and remote computing platforms.

Employing predictions of system dynamics, human behavior and environment components can facilitate such task. In addition, historical and real-time data can be used to bound forecasts uncertainty, learn model parameters and allow the system to adapt to new tasks.

Our research over the past decade has focused on control design for autonomous systems which systematically incorporate predictions and learning. In this talk I will first provide an overview of the theory and tools that we have developed for the designing of learning predictive controllers. Then, I will focus on recent results that use data to efficiently formulate stochastic control problems which autonomously improve performance in iterative tasks. Throughout the talk I will focus on autonomous cars to motivate our research and show the benefits of the proposed techniques.

More info on: www.mpc.berkeley.edu

Keywords: Machine learning, Robotics, UAV's
Abstract: In this paper, we propose an online learning approach that enables the inverse dynamics model learned for a source robot to be transferred to a target robot (e.g., from one quadrotor to another quadrotor with different mass or aerodynamic properties). The goal is to leverage knowledge from the source robot such that the target robot achieves high-accuracy trajectory tracking on arbitrary trajectories from the first attempt with minimal data recollection and training. Most existing approaches for multi-robot knowledge transfer are based on post-analysis of datasets collected from both robots. In this work, we study the feasibility of impromptu transfer of models across robots by learning an error prediction module online. In particular, we analytically derive the form of the mapping to be learned by the online module for exact tracking, propose an approach for characterizing similarity between robots, and use these results to analyze the stability of the overall system. The proposed approach is illustrated in simulation and verified experimentally on two different quadrotors performing impromptu trajectory tracking tasks, where the quadrotors are required to accurately track arbitrary hand-drawn trajectories from the first attempt.

Keywords: Neural networks, Machine learning
Abstract: The energy output of photovoltaic (PV) power plants depends on the environment and thus fluctuates over time. As a result, PV power can cause instability in the power grid, in particular when increasingly used. Limiting the rate of change of the power output is a common way to mitigate these fluctuations, often with the help of large batteries. A reactive controller that uses these batteries to compensate ramps works in practice, but causes stress on the battery due to a high energy throughput. In this paper, we present a deep learning approach that uses images of the sky to compensate power fluctuations predictively and reduces battery stress. In particular, we show that the optimal control policy can be computed using information that is only available in hindsight. Based on this, we use imitation learning to train a neural network that approximates this hindsight-optimal policy, but uses only currently available sky images and sensor data. We evaluate our method on a large dataset of measurements and images from a real power plant and show that the trained policy reduces stress on the battery.

Keywords: Robust control, Predictive control for nonlinear systems, Uncertain systems
Abstract: Handling uncertainties is one of the most important challenges in nonlinear model predictive control (NMPC). While several robust NMPC methods have been recently presented, their implementation on embedded systems is usually difficult due to the necessary conservative assumptions or because of the required computational complexity.

In this work, we use a complex robust NMPC approach to generate data pairs that are used to learn an approximate robust controller which is robust to model uncertainties. We propose to use deep neural networks to learn the approximate controller based on recent results that prove the powerful representation capabilities of such networks over traditional shallow ones. The approximate controller, which just requires the simple forward evaluation of neural network, can be easily combined with an Extended Kalman Filter to obtain an efficient embedded implementation of an output-feedback robust NMPC scheme. We propose a statistical verification strategy to compute backoffs that lead to the satisfaction of important constraints despite the presence of estimation, measurement and approximation errors. The potential of the approach is illustrated with numerical results for the embedded robust control of a towing kite.

Keywords: Iterative learning control, Output regulation, Predictive control for linear systems
Abstract: This work presents a data driven control strategy able to track a set point without steady state error. The control sequence is computed as an affine combination of past control signals, which belong to a set of past closed loop trajectories stored in a process historian database. This affine combination is computed so that the variance of the tracking error is minimized. It is shown that offset free control (zero mean tracking error) is achieved under the assumption that the underlying dynamics are linear and the closed loop trajectories of the database are in turn offset free. That is, the proposed strategy inherits the offset free tracking capability of the stored past closed loop trajectories. No prior or subsequent knowledge about the process dynamics is required. The procedure to build the database is to store only the best trajectories that meet a design criteria, chosen from a series of iteratively tuned controllers. In this way the proposed controller will learn how to obtain a well tuned control in spite of the different operating conditions.

Keywords: Computational methods, Statistical learning, Identification for control
Abstract: Passivity is an important concept in control design as it pertains to stability properties of the closed loop. We propose a framework to determine to which extent a dynamic system is or is not passive from data. In particular, we develop a probabilistic approach based on Gaussian processes to underestimate the input feedforward passivity index from experiments with measurement noise. We also show how prior knowledge on the input-output behavior can be incorporated in this framework. Besides the offline approach, we present an iterative scheme that in expectation tightens the lower bound on the feedforward passivity index with every additional data sample and gives an upper bound on the conservatism of the resulting passivity measure.

Keywords: Identification for control, Adaptive control, Iterative learning control
Abstract: For processes that repeat the same task an iterative control method can be applied that learns from the previous runs and corrects the tracking error for the next run. Meanwhile, a set membership identification technique is be combined with the regulation part to identify the plant response using the informative input-output data of previous observations. This combination method has been experimentally tested on a radio frequency machine and the results are presented in the paper.

Keywords: Distributed control, Optimal control, H2/H-infinity methods
Abstract: We consider the problem of designing a stabilizing and optimal static controller with a pre-specified sparsity pattern. Since this problem is NP-hard in general, it is necessary to resort to approximation approaches. In this paper, we characterize a class of convex restrictions of this problem that are based on designing a separable quadratic Lyapunov function for the closed-loop system. This approach generalizes previous results based on optimizing over diagonal Lyapunov functions, thus allowing for improved feasibility and performance. Moreover, we suggest a simple procedure to compute favourable structures for the Lyapunov function yielding high-performance distributed controllers. Numerical examples validate our results.

Keywords: Distributed cooperative control over networks, Delay systems, Linear systems
Abstract: This paper examines accelerating the well-known Laplacian average consensus algorithm by breaking its conventional delay-free input into two weighted parts and replacing one of these parts by an outdated feedback. We determine for what weighted sum there exists a range of time delay that leads to increase in the rate of convergence of the algorithm. For such weights, using the Lambert W function, we obtain the rate increasing range of the time delay and also the maximum reachable rate and its corresponding maximizer delay. We also specify what combinations of the current and an outdated feedback increase the rate of convergence without increasing the control effort of the agents. Lastly, we determine the optimum combination of the current and the outdated feedback weights to achieve maximum increase in the rate of convergence without increasing the control effort. We demonstrate our results through a numerical example.

Keywords: Distributed cooperative control over networks, Robust control, Predictive control for nonlinear systems
Abstract: This paper addresses the task of distributed model predictive control (DMPC) for a set of subsystems with affine dynamics including disturbances. For a set-up in which the subsystems are coupled through their cost functions and state constraints, we focus on the robustness of DMPC with respect to the delays of a communications network used for transmitting information between subsystems. As an alternative to existing schemes using a constant upper bound of communication delay, we here propose to employ predictions of time-varying communication delay, and we show that robustness can still be preserved.

Keywords: Distributed cooperative control over networks, Control over communication, Autonomous systems
Abstract: This paper proposes an event-based method to ensure a collision-free movement of mobile objects with a certain spatial separation. The vehicles are locally controlled and connected by a communication network. For collision avoidance each object is provided with an event-based trajectory planning unit, which acts on a predictor representing the region in which the neighbouring object is located. If the spatial separation between the two objects deceeds a threshold, communication to the neighbouring object is invoked. The local trajectory is replanned based on Bézier curves to guarantee the collision avoidance. In contrast to existing literature this approach is performed online and with a low communication effort. The advantages of the method are demonstrated by a simulation study with two quadrotors.

Keywords: Distributed cooperative control over networks
Abstract: The coverage control problem for robotic networks focuses on distributively coordinating the positioning of multiple dynamic agents to provide sensor coverage across a bounded region in two dimensional space. The associated optimal coverage problem seeks to position these agents so as to minimize an associated coverage cost. This coverage cost is typically defined with respect to a density function that is used to bias the network towards desired configurations. Previous approaches to this optimal coverage problem have addressed both static and dynamic environments through the choice of density function; however, stability guarantees for time-varying densities are restricted by significant technical assumptions that simplify the underlying proofs at the expense of limited applicability. In this paper, a generalized algorithm is presented that guarantees practical stability under relaxed technical assumptions. The algorithm, and its convergence, is illustrated via simulation examples.

Keywords: Optimization algorithms, Large-scale systems, Distributed control
Abstract: In this paper, we deal with large-scale Mixed Integer Linear Programs (MILPs) with coupling constraints that must be solved by processors over networks. We propose a finite-time distributed algorithm that computes a feasible solution with suboptimality bounds over asynchronous and unreliable networks. As shown in a previous work of ours, a feasible solution of the considered MILP can be computed by resorting to a primal decomposition of a suitable problem convexification. In this paper we reformulate the primal decomposition resource allocation problem as a linear program with an exponential number of unknown constraints. Then we design a distributed protocol that allows agents to compute an optimal allocation by generating and exchanging only few of the unknown constraints. Each allocation is iteratively used to compute a candidate feasible solution of the original MILP. We establish finite-time convergence of the proposed algorithm under very general assumptions on the communication network. A numerical example corroborates the theoretical results.

Keywords: Agents and autonomous systems, Agents networks, Distributed control
Abstract: The paradigmatic formation control problem of steering a multi-agent system towards a balanced circular formation has been the subject of extensive studies in the control engineering community. Indeed, this is due to the fact that it shares several features with relevant applications such as distributed environmental monitoring or fence-patrolling. However, these applications may also present some relevant differences from the ideal setting such as the curve on which the formation must be achieved not being a circle, or the measurements being neither ideal nor as a continuous information flow. In this work, we attempt to fill this gap between theory and applications by considering the problem of steering a multi-agent system towards a balanced formation on a generic closed curve and under very restrictive assumptions on the information flow amongst the agents. We tackle this problem through an estimation and control strategy that borrows tools from interval analysis to guarantee the robustness that is required in the considered scenario.

Keywords: Agents and autonomous systems, Cooperative autonomous systems, Constrained control
Abstract: We study the problem of controlling multi-agent systems under a set of signal temporal logic tasks. Signal temporal logic is a formalism that is used to express time and space constraints for dynamical systems. Recent methods to solve the control synthesis problem for single-agent systems under signal temporal logic tasks are, however, subject to a high computational complexity. Methods for multi-agent systems scale at least linearly with the number of agents and induce even higher computational burdens. We propose a computationally-efficient control strategy to solve the multi-agent control synthesis problem that results in a robust satisfaction of a set of signal temporal logic tasks. In particular, a decentralized feedback control law is proposed that is based on time-varying control barrier functions. The obtained control law is discontinuous and formal guarantees are provided by nonsmooth analysis. Simulations show the efficacy of the presented method.

Keywords: Agents and autonomous systems, Cooperative autonomous systems, Coverage control
Abstract: In this paper, a task of monitoring complex unknown environments by a large group of autonomous robotic agents is tackled, where it aims to find a final deployment of agents with good distribution and a large area of coverage. A novel Virtual Force Interaction Scheme (VFIS) among agents and an environment is proposed to accomplish the task. In addition to the virtual repulsive force exerted on agents for local exploitation, a new type of virtual force, vortex force, is firstly introduced for better exploration of the whole environment. In VFIS, each agent consecutively updates its position according to the varying virtual forces it bears and the final distribution is obtained after some iteration. The designed scheme has been validated in a series of experiments with different configurations successfully. Furthermore, the new structure particle swarm optimization is utilized for finding more efficient, obstacle-free trajectories from the initial positions to the final distribution after VFIS is applied.

Keywords: Agents and autonomous systems, Cooperative control, Autonomous systems
Abstract: Control of multi-agent systems with the focus on flocking has been studied heavily in recent years. The principles of interactions between independent agents have applications in the fields of swarm robotics, biology, economics and in control of human crowd motion. Multi-agent dynamic systems can be described by large-scale systems of N coupled nonlinear differential equations. The goal of this study is to move a flock of Cucker-Dong dynamics through a virtual leader without collision with different shaped obstacles. This paper proposes an extension of the Cucker-Dong particle system to perform group maneuvers in a safe manner.

Keywords: Agents and autonomous systems, Distributed control, Agents networks
Abstract: This work deals with formations of mobile agents with six independently controllable degrees of freedom able to retrieve relative bearing measurements w.r.t. some neighbors in the group. Exploiting the bearing rigidity framework, two control objectives are here addressed: (i) the stabilization of these fully actuated multi-agent systems towards desired configurations, and (ii) their coordinated motion along directions guaranteeing the system shape maintenance. The proposed approach relies on a new formulation of the bearing rigidity theory based on the adoption of the unit quaternion formalism to describe the agents attitude. Through this representation choice, the formation dynamics is linear w.r.t. the input control velocities and the rigidity theory suggests the design of a distributed control scheme for both control goals whose efficacy is confirmed by numerical simulations.

Keywords: Agents networks, Concensus control and estimation, Network analysis and control
Abstract: In this paper we analyze the Friedkin-Johnsen model of opinions when the coefficients weighting the agent susceptibilities to interpersonal influence approach 1. We will show that in this case, under suitable assumptions, the model converges to a quasi-consensus condition among the agents. In general the achieved consensus value will be different from the one obtained by the corresponding DeGroot model.

Keywords: Aerospace, H2/H-infinity methods, Flexible structures
Abstract: The Benchmark Active Control Technology model based on NASA research and wind tunnel test is used to demonstrate the possibilities of modern fixed-order controllers for active flutter suppression. Direct design methods based on nonsmooth, nonconvex optimization are utilized that give rise to H-infinity optimal control laws with prescribed structure and complexity. We compare achieved performance with classical unconstrained H-infinity design.

Keywords: Aerospace, Neural networks, Modeling
Abstract: The use of Flush Air Data System (FADS) sensors has been limited to estimation of air flow parameters like angle of attack, and side-slip angle. This paper presents a detailed analysis and methodology for using a network of spatially distributed FADS sensors to estimate the instantaneous aerodynamic loading over an aircraft for use in controls and health monitoring applications.

Keywords: Aerospace
Abstract: In ground vehicles, tire consumption is in general mainly due to the mileage covered, and in fact the life span of tires, at least in common situations, is rather long. In the aeronautical context, and for aircraft in particular, instead, tire consumption plays a crucial role in determining the maintenance costs. This is due to the fact that, in aircraft braking, nearly all maneuvers activate the anti-skid controller, which remains in use for long time intervals. In ground vehicles, instead ABS systems are usually active for short time intervals which cover a part of the braking maneuvers only. Thus, tire consumption in the automotive context is usually studied under constant speed assumptions. In this work, we formulate a tire consumption models that encompasses explicitly the wheel acceleration/deceleration dynamics, and we show that tire wear can be directly related to the anti-skid controller parameters. Based on this, a sensitivity analysis of tire-consumption versus braking performance is carried out, showing that the braking control problem can be reformulated as a tire consumption regulation one.

Keywords: Fault detection and identification, Aerospace
Abstract: The design of reliable and fault tolerant space flexible structures is a key point to guarantee successful space missions. This paper presents a study of detectability of fault affecting sensors and actuators of a continuous flexible structure subject to unknown disturbances. Several sensors and actuators configurations, also including the so called co-located setup, are investigated. The structural properties of these class of systems is studied by means of the geometric approach-based system theory which provides, in a coordinate-free framework, necessary and sufficient conditions for the resolvability of the fault detection problem.

Keywords: Aerospace, Military applications, Sliding mode control
Abstract: A finite-time-convergent impact time control guidance law based on pure proportional navigation is proposed for interception of stationary target. The proposed guidance law achieves impact time control by tracking desired look angle. The desired look angle is defined to satisfy the time-to-go condition at each instant by pure proportional navigation. The guidance law can deal with feasible impact times for all initial conditions without any linearisation and approximation. The finite-time convergence of tracking error ensures that the missile can intercept the target while maintaining the characteristics of pure proportional navigation. Numerical simulation is performed to verify the properties and performance of the proposed guidance law.

Keywords: Aerospace, Fault estimation, Fault tolerant systems
Abstract: Single gimbal control moment gyros (SGCMGs) have been widely used on agile satellites to get a rapid retargeting capability. To enhance the reliability and safety of the SGCMG system, this paper addresses the fault estimation and fault- tolerant steering problem. The SGCMG is modeled as a two-loop system including a wheel speed control loop and a gimbal rate control loop, and a cascade multiplicative fault model of SGCMG is developed. Then, in view of the complexity of the gimbal fault, a local adaptive fault estimator is proposed to reconstruct the total time-varying fault effects for each SGCMG. Using the estimated fault effects, a fault-tolerant steering logic is further developed to not only allocate the commanded attitude control torque properly but also compensate the fault effects. To verify the proposed fault estimator and fault-tolerant steering logic, numerical simulations are carried out on an SGCMG-actuated spacecraft.

Keywords: Nonlinear system theory, Lyapunov methods, Robotics
Abstract: In this paper, we develop sufficient conditions for uniform asymptotic stability of interconnected dynamical systems that are not in cascade form. We show that the stability analysis of a two-subsystem interconnection can be reduced to ensuring the stability of the first non-isolated subsystem with respect to its own state vector (partial stability) and the stability of the isolated second subsystem. In addition, based on the above results, we provide a control design framework for nonlinear systems where the control objective reduces to stabilization of only a part of the system state while guaranteeing the stability for the entire state of the system.

Keywords: Nonlinear system theory, Optimization, Robust control
Abstract: In this paper, we consider a problem of designing discrete-time systems which are optimal in frequency-weighted least squares sense subject to a maximal output amplitude constraint. It can be shown for such problems, in general, that the optimality conditions do not provide an explicit way of generating the optimal output as a real-time implementable transformation of the input, due to the instability of the resulting dynamical equations and sequential nature in which criterion function is revealed over time. In this paper, we extend the well-known method of balanced truncation for linear systems to the case of nonlinear systems with diagonal contractive operators. We then propose a causally stable finite-latency nonlinear system which returns high-quality approximations to the optimal solution. The proposed system is obtained by a careful truncation of an infinite dimensional representation of the optimal system, as suggested by the derived extension of the balanced truncation method.

Keywords: Stability of nonlinear systems, Nonlinear system theory
Abstract: The ability of a nonlinear system to recover from a large disturbance to a desired stable equilibrium point depends on system parameter values, which are often uncertain and time varying. A particular disturbance acting for a finite time can be modeled as an implicit map that takes a parameter value to its corresponding post-disturbance initial condition in state space. The system recovers when the post-disturbance initial condition lies inside the region of attraction of the stable equilibrium point. Critical parameter values are defined to be parameter values whose corresponding post-disturbance initial condition lies on the boundary of the region of attraction. Computing such values is important in numerous applications because they represent the boundary between desirable and undesirable system behavior. Many realistic system models involve controller clipping limits and other forms of switching. Furthermore, these hybrid dynamics are closely linked to the ability of a system to recover from disturbances. The paper develops theory which underpins a novel algorithm for numerically computing critical parameter values for nonlinear systems with clipping limits and switching. For an almost generic class of vector fields with event-selected discontinuities, it is shown that the boundary of the region of attraction is equal to a union of the stable manifolds of the equilibria and periodic orbits it contains, and that this decomposition persists and the boundary varies continuously under small changes in parameter.

Keywords: Stability of nonlinear systems, Output regulation, Lyapunov methods
Abstract: This paper presents a new passivity property for distributed piezoelectric devices with integrable port-variables. We present two new control methodologies by exploiting the integrability property of the port-variables. The derived controllers have a Proportional-Integral (PI) like structure. Finally, we present the simulation results and an in-depth analysis on the tuning gains explaining their transient and the steady-state behaviors.

Keywords: Nonlinear system theory, Cooperative control, Agents and autonomous systems
Abstract: A cooperative consensus algorithm is proposed for a group of double integrators. This algorithm is based on minimal relative information and does not use communication networks. Each agent counts the number of predecessor and successor neighbors. The difference of these numbers is used in the controller instead of any accurate range or position measurements. Non-smooth Lyapunov method is used to analyze the stability under the proposed consensus algorithm. Under certain conditions, the proposed protocol can ensure bounded control input, finite-time convergence, or fixed-time convergence. Simulation results illustrate the effectiveness of the proposed algorithm.

Keywords: Stability of nonlinear systems, Lyapunov methods, Nonlinear system theory
Abstract: Lyapunov characterizations of output finite-time stability are presented for the system dot x=f(x), y=h(x) which is locally Lipschitz continuous out of the set Y={x in R^n: h(x)=0} and continuous on R^n. The definitions are given in the form of K and KL functions. Necessary and sufficient conditions for output finite-time stability are given using Lyapunov functions. The theoretical results are supported by numerical examples.

Keywords: Biomedical systems, Medical signal processing, Robotics
Abstract: This contribution considers a neuroprosthesis for people who have (partially) paralyzed hand functions as a typi- cal result of a stroke or spinal cord injury. Functional Electrical Stimulation (FES) is applied to three forearm-muscles to induce artificial muscle contractions causing wrist extension/flexion and grasping, respectively. The aim is to control and stabilize the wrist joint angle. Herein, a co-activation to increase the stiffness of the wrist joint is applied to yield a better tolerance to external influences. Doing so, a better motor precision is expected compared to other approaches. To maintain the level of co-activation also under the progression of muscle fatigue, the muscle activation is estimated and controlled using an underlying feedback of the electrical muscle responses as caused by motor unit recruitment. The control system was preliminary tested on two healthy subjects, and a wrist positioning error of 4.5◦ on average was obtained demonstrating the feasibility of the co-activation based approach.

Keywords: Biomedical systems, Biological systems, Medical signal processing
Abstract: State estimation of biological systems is a difficult task due to their complexity and stochasticity. In particular, bilinear and Michaelis-Menten terms are the base for many biological models such as in infectious diseases, cancer, diabetes, and many others.

In this paper, mentioned non-linear terms are formulated into a polynomial form with state-dependent matrices driven by additive white Gaussian noises over linear observations. To show the effectiveness of the approach, two different models widely used for modeling viral infectious diseases are considered and compared with the extended Kalman filter (EKF) algorithm. Numerical results show the applicability of the polynomial approach.

Keywords: Biomedical systems, Optimization
Abstract: This paper deals with the application of an Internal Model Control strategy to the regulation of the depth of hypnosis in general anesthesia, when propofol is used as hypnotic drug and the bispectral index scale is used as controlled variable. In particular, the inverse of the Hill function is used to linearize the system and then a standard internal model controller is applied. The internal model control filter time constants are determined by applying a particle swarm optimization algorithm on a set of patient models that describes a wide range of population. The Monte Carlo method is then used to demonstrate the robustness of the controller with respect to intra-patient and inter-patient variability.

Keywords: Biomedical systems, V&V of control algorithms, Sliding mode control
Abstract: Paramedics frequently perform physically demanding tasks during patient transport. This leads to extremely high rates of work-induced injuries and early retirement. A novel self-balancing rescue aid (SEBARES) with an additional stair climbing mechanism, which is externally guided by a paramedic, is supposed to reduce the workloads and allowing an intuitive operation and safe transport for the patient. For the self-stabilizing functionality precise and stable control at any time is crucial. In this study a sliding mode controller, which is known for robust and stable control of nonlinear applications, is designed and implemented as a first control approach for such a device. To rate the performance application specific control requirements are defined, which address critical scores such as ergonomic user forces, maximal tilt angles and maximal overshoot. In the subsequent experimental evaluation the performance of the device and the controller are determined. The sliding mode controller successfully stabilized the system and met the predefined requirements. The system was able to accelerate adequately to match typical accelerations of walking humans, forces below 20 N had to be applied during constant walking and the system was able to decelerate from 1.4 m/s to a full halt within a sufficient distance of maximal 0.41 m. Upcoming studies will include a comparative analysis of different control approaches as well as the development of the stair climbing kinematics and obstacle detection.

Keywords: Applications in neuroscience, Biomedical systems
Abstract: Understanding the performance of the human brain to stabilize the body remains an open fundamental research question. In this article, we study the hypothesis of internal model of the Central Nervous System (CNS) by a novel proposed architecture based on a recurrent neural network. The overall objective of the article and the main contribution stems from demonstrating the capability of replicating the balancing mechanisms of the brain by training the proposed bio-inspired network architecture with human balancing data and in the sequel applying the resulting control structure for controlling a single link inverted pendulum. Towards this direction, the body kinetics and kinematics measurements of forty-five subjects during upright stance trails were collected and utilized for training the proposed neural network. The efficacy of the proposed scheme will be proven through multiple simulation results with a single link inverted pendulum, where it will be demonstrated that the brain-inspired control scheme achieves a proper balance.

Keywords: Nonlinear system theory, Biomedical systems, Modeling
Abstract: In this work, discrete sliding mode control method in combination with non-invasive pulsatile estimator model is used to develop non-linear tracking control algorithm in the presence of bounded uncertainties and disturbances. A sate space model to estimate mean pulsatile flow based on the pulsatility index of current is constructed. The control method is introduced the reference model as part to design the algorithm. The reaching control law is proposed, and the conditions of system stability are considered. The algorithm is evaluated using dynamical closed loop system to study the robustness for the system uncertainties. Simulation results reveal the efficiency of the developed controller.

Keywords: Automotive, Autonomous systems, Transportation systems
Abstract: In this paper we address the problem of co- ordinating automated vehicles at intersections, with a spe- cial focus on turning maneuvers. The inclusion of rear-end collision avoidance constraints into the problem is decided during turning maneuvers by a smooth function of the vehicle state, rather than integer variables. Moreover, curvature-based acceleration constraints are introduced, which limit the velocity of the vehicle during the turn, and a term in the objective function accounts for passenger comfort. We discuss how the coordination problem is formulated as a nonlinear program and show though simulations that for practical problem instances the proposed approximation is either exact or introduces very little conservativeness.

Keywords: Automotive, Autonomous systems
Abstract: This paper addresses learning of the tire-friction curve for road vehicles, using a batch of wheel-speed and inertial measurements. We formulate a Bayesian approach based on recent advances in particle filtering and Markov chain Monte-Carlo methods. The unknown function mapping the wheel slip to tire friction is modeled as a Gaussian process (GP) that is included in a dynamic vehicle model relating the GP to the vehicle state. The approach is nonparametric and learns the probability density function of the tire friction, from which explicit estimates can be extracted. One benefit of the method is that it is not subject to overfitting issues. We illustrate the efficacy of the method for a set of simulated step-steer maneuvers. The results show that the method can accurately identify the nonlinear tire-friction curves, even for a limited amount of data.

Keywords: Automotive, Identification, Nonlinear system theory
Abstract: Physical parameters of a vehicle drivetrain are required in many applications. In the context of fault diagnosis, for example, knowledge about the installed components or parts can provide insights in order to verify if they behave in accordance with standards or deviate from them in a way that adversely affects the operating performance. For this purpose a frequency domain identification approach is presented, which is based only on standard mounted sensors. In the presented method in particular nonlinear effects are taken into account resulting from backlash. In order to guarantee a unique parameter set a local identifiability analysis is performed. The main idea of the method is to exploit the dependency between the frequency response of the nonlinear system and the magnitude of the test-signal to improve optimization of the physical parameters. Finally, identification results using real measurement data are presented.

Keywords: Automotive, Optimal control, Uncertain systems
Abstract: We investigate the problem of coordinating human-driven vehicles in road intersections without any traffic lights or signs by issuing speed advices. The vehicles in the intersection are assumed to move along an a priori known path and to be connected via vehicle-to-vehicle communication. The challenge arises with the uncertain driver reaction to a speed advice, especially in terms of the driver reaction time delay, as it might lead to unstable system dynamics. For this control problem, a distributed stochastic model predictive control concept is designed which accounts for driver uncertainties. By optimizing over scenarios, which are sequences of independent and identically distributed samples of the uncertainty over the prediction horizon, we can give probabilistic guarantees on constraint satisfaction. Simulation results demonstrate that the scenario-based approach is able to avoid collisions in spite of uncertainty while the non-stochastic baseline controller is not.

Keywords: Automotive, Linear parameter-varying systems, Optimization algorithms
Abstract: The paper deals with energy optimal control allocation of an in-wheel electric vehicle with autonomous trajectory tracking. The proposed method is based on both high-level control allocation between steering intervention and torque vectoring minimizing cornering resistance of the vehicle, and a low-level multi-criteria torque distribution method considering power consumption of the electric in-wheel motors. The aim of the design is to enhance battery state-of-charge (SOC), extending the range of the electric vehicle. The reconfiguration control design is founded on Linear Parameter Varying (LPV) framework, while the wheel torque distribution is calculated using constrained optimization techniques. The operation of the energy optimal reconfiguration control is demonstrated in CarSim simulation environment with a detailed battery and electric motor model.

Keywords: Automotive, Predictive control for nonlinear systems, Autonomous systems
Abstract: For motion planning and control of autonomous vehicles to be proactive and safe, pedestrians' and other road users' motions must be considered. In this paper, we present a vehicle motion planning and control framework, based on Model Predictive Control, accounting for moving obstacles. Measured pedestrian states are fed into a prediction layer which translates each pedestrians' predicted motion into constraints for the MPC problem.

Simulations and experimental validation were performed with simulated crossing pedestrians to show the performance of the framework. Experimental results show that the controller is stable even under significant input delays, while still maintaining very low computational times. In addition, real pedestrian data was used to further validate the developed framework in simulations.

Keywords: Complex systems, Linear systems, Algebraic/geometric methods
Abstract: This paper introduces assume/guarantee contracts on continuous-time control systems, hereby extending contract theories for discrete systems to certain new model classes and specifications. Contracts are regarded as formal characterizations of control specifications, providing an alternative to specifications in terms of dissipativity properties or set-invariance. The framework has the potential to capture a richer class of specifications more suitable for complex engineering systems. The proposed contracts are supported by results that enable the verification of contract implementation and the comparison of contracts. These results are illustrated by an example of a vehicle following system.

Keywords: Hybrid systems, Algebraic/geometric methods, Computational methods
Abstract: This paper deals with the problem of asymptotically estimating a linear function of the state of a linear impulsive system, in the presence of unknown inputs, by means of an observer whose state space has the minimal possible dimension. The linear impulsive systems considered are subject to the following constraint: the length of the time interval between any two consecutive jumps must be greater than or equal to a given finite positive constant. First, a necessary and sufficient condition for the existence of an observer whose state space has a generic dimension (i.e., not necessarily minimal) is proven. Then, the issue of the minimization of the observer state dimension is investigated and solved.

Keywords: Hybrid systems, Automata, Biomedical systems
Abstract: Symbolic models are becoming more and more popular in the research community working on hybrid systems because they provide a systematic approach to enforce logic specifications on purely continuous or hybrid systems while fulfilling the constraints at the hardware/software implementation level. This paper contributes to this research line and proposes symbolic models approximating possibly unstable time-delay systems with quantized measurements of the outputs. An application of the proposed results to the glucose control problems for the Artificial Pancreas is discussed in the paper.

Keywords: Hybrid systems, Electrical power systems, Stability of hybrid systems
Abstract: In this article, a hybrid control law is proposed for the three-phase three-level Neutral Point Clamped (NPC) converter working as a rectifier in order to regulate the output DC voltage. The control problem deals with the unbalance capacitor voltages as well as the phase currents. The proposed algorithm is based on the Hybrid Dynamical System theory, which takes into account the hybrid nature of the NPC converter, i.e., the continuous and discrete dynamics, proving uniform global asymptotic stability of the operating point. Finally, the effectiveness of the proposed control algorithm is validated in simulation.

Keywords: Hybrid systems, Output regulation, Stochastic systems
Abstract: The problem of output regulation for linear stochastic systems is addressed. The controlled system belongs to a general class of linear systems, namely the state, the control input and the exogenous input appear in both the drift and diffusion terms of the differential equations. Building upon the solution of the ideal, non-causal, stochastic regulator problem, we define an approximate full-information problem. By means of measurements of the state vector, we provide a way to compute a sequence of scalars approximating a posteriori the variations of the Brownian motion. Then, we propose a hybrid control architecture which solves the approximate problem. The continuous-time part of the controller is deterministic, whereas the discrete-time part has the function of "correcting" the control action by means of the approximate discrete-time Brownian motion. The solution of the ideal stochastic regulator problem is recovered as the sampling time tends to zero. We illustrate the results by means of a numerical example and conclude the paper with some final remarks: the proposed control architecture is the first causal solution of the full-information output regulation problem and is an essential intermediate step for the solution of the error-feedback problem.

Keywords: Hybrid systems, Switched systems, Identification for hybrid systems
Abstract: This paper deals with the problem of continuous state and mode estimation for piecewise linear systems, which can be considered as a class of state-dependent linear switched systems. Some necessary and sufficient observability conditions for the mode observation, i.e. mode observability, are provided. On the other hand, some sufficient observability conditions for the continuous state are also introduced. Based on such conditions, a mode observer and a bank of continuous state observers are proposed to estimate the continuous and the mode by means of sliding-mode control techniques. Some simulation results illustrate the effectiveness of the proposed approach.

Keywords: Model/Controller reduction, Predictive control for linear systems, Optimal control
Abstract: Despite the success of model predictive control (MPC), its application to high-dimensional systems, such as flexible structures and coupled fluid/rigid-body systems, remains a largely open challenge due to excessive computational complexity. A promising solution approach is to leverage reduced order models for designing the model predictive controller. In this paper we present a reduced order MPC scheme that enables setpoint tracking while robustly guaranteeing constraint satisfaction for linear, discrete, time-invariant systems. Setpoint tracking is enabled by designing the MPC cost function to account for the steady-state error between the full and reduced order models. Robust constraint satisfaction is accomplished by solving (offline) a set of linear programs to provide bounds on the errors due to bounded disturbances, state estimation, and model approximation. The approach is validated on a synthetic system as well as a high-dimensional linear model of a flexible rod, obtained using finite element methods.

Keywords: Predictive control for linear systems, Behavioural systems, Identification for control
Abstract: We consider the problem of optimal trajectory tracking for unknown systems. A novel data-enabled predictive control (DeePC) algorithm is presented that computes optimal and safe control policies using real-time feedback driving the unknown system along a desired trajectory while satisfying system constraints. Using a finite number of data samples from the unknown system, our proposed algorithm uses a behavioural systems theory approach to learn a non-parametric system model used to predict future trajectories. The DeePC algorithm is shown to be equivalent to the classical and widely adopted Model Predictive Control (MPC) algorithm in the case of deterministic linear time-invariant systems. In the case of nonlinear stochastic systems, we propose regularizations to the DeePC algorithm. Simulations are provided to illustrate performance and compare the algorithm with other methods.

Keywords: Predictive control for linear systems, Distributed control, Large-scale systems
Abstract: In this paper a new algorithm for computing a terminal cost function in Distributed Model Predictive Control is presented. In order to ensure stability it is required that the terminal cost function acts as a Lyapunov function with respect to a nominal linear feedback control law. An additional separable structure is imposed on the terminal cost to make the global online MPC problem amenable to distributed optimization. A possible method to achieve this design consists in recasting the synthesis problem as a Linear Matrix Inequality (LMI) problem. In such formulation the required structure can be easily added as an additional linear constraint. In networked systems there is an increasing number of applications requiring online reconfiguration of the controllers typically in response to changes in the network structure. Here solving an LMI problem online can be prohibitive due to the required computation time. This paper proposes for these scenarios a novel design approach based on matrix iteration which provides a fast solution and it can be effectively employed online.

Keywords: Optimal control, Predictive control for linear systems, Maritime
Abstract: A novel indirect variational optimal control theory is proposed for integral-differential equations of motion. This algorithm is applied to an engineering application: the control of a an underawater autonomous vehicle’s 2-Dof lifting surface. The variational approach proposed is an extension of the classical Potryagin optimal control theory which normally refers to differential equations. The control has been extended by developing a novel integral MPC technique. Numerical results show good performace of the optimal control proposed compared with the standard LQR method.

Keywords: Predictive control for linear systems, Robust control, Constrained control
Abstract: We present a novel method for the computation of small RPI sets tailored for tube-based robust MPC. In the framework of robust MPC for linear systems with additive disturbances, RPI sets are used to guarantee robust constraint satisfaction (and robust stability) through constraint tightening. To minimize the tightening, small RPI sets are beneficial. Hence, classical approaches aim for epsilon-approximations of the minimal RPI set. Unfortunately, the computation of those approximations usually requires Minkowski sums that are numerically expensive. In contrast, our method avoids demanding Minkowski sums. Nevertheless, the resulting RPI sets lead to tightened constraints that are identical to those for epsilon-approximations of the minimal RPI set. The (simple) key observation underlying our approach is that the Pontryagin difference, that is instrumental for the constraint tightening, can lead to identical results for different subtrahend sets.

Keywords: Predictive control for linear systems, Robust control, Output feedback
Abstract: We propose a new robust output feedback Model Predictive Control (MPC) scheme that can handle both additive and parametric (multiplicative) uncertainties using the tube-enhanced multi-stage (TEMS) MPC framework. In TEMS MPC, the significant parametric uncertainties are handled using the multi-stage MPC formulation and the additive uncertainties are handled using the tube-based MPC framework resulting in an improved trade-off between optimality and complexity. The proposed output feedback approach can handle estimation errors in addition to multiplicative and additive uncertainties. A key advantage of the proposed formulation is that it is independent of the state estimation scheme employed and hence it can be easily combined with any estimation scheme. The recursive feasibility and stability of the proposed approach in the linear case are proven. The advantages of the proposed approach are demonstrated for two examples.

Keywords: Linear parameter-varying systems, Nonlinear system identification, Identification
Abstract: The article presents a method for the identification of Linear Parameter-Varying (LPV) models in a Linear Fractional Representation (LFR), which corresponds to a Linear Time-Invariant (LTI) model connected to a scheduling variable dependency via a feedback path. A two-stage identification approach is proposed. In the first stage, Kernelized Canonical Correlation Analysis (KCCA) is formulated to estimate the state sequence of the underlying LPV model. In the second stage, a non-linear least squares cost function is minimized by employing a coordinate descent algorithm to estimate latent variables characterizing the LFR and the unknown model matrices of the LTI block by using the state estimates obtained at the first stage. Here, it is assumed that the structure of the scheduling variable dependent block in the feedback path is fixed. For a special case of affine dependence of the model on the feedback block, it is shown that the optimization problem in the second stage reduces to ordinary least-squares followed by a singular value decomposition.

Keywords: Identification, Nonlinear system identification, Cellular dynamics
Abstract: This paper analyzes the accuracy with which the drug-resistant sub-population of cancer cells in a tumor can be estimated from measurements of total tumor size. The paper is motivated by two key facts. First, drug resistance is one of the main reasons for the failure of cancer chemotherapy treatment: a fact that makes it critical to monitor and estimate such resistance. Second, recent research has shown that above a threshold level of drug resistance, the optimal treatment protocol is one that regulates total cancer size rather than attempting to eliminate the cancer. This makes the accurate estimation of resistance critical for treatment protocol selection. The literature already examines the causes and dynamics of resistance in cancerous tumors. However, the problem of determining the accuracy with which the prevalence of resistance can be estimated remains relatively unexplored. To address this gap in the literature, we apply Fisher information analysis to the problem of estimating the fraction of a total cancer cell population that is drug-resistant, assuming a constant drug administration rate. Our analysis reveals that drug-resistant cell population estimation accuracy worsens with increasing drug administration rate up to the point where the drug-sensitive and drug-resistant cell population growth rates are equal. Beyond that point, additional drug administration improves resistance estimation accuracy.

Keywords: Identification, Signal processing
Abstract: In Prediction Error Identification, the consistency of the identified parameter vector is only guaranteed if the data are informative enough i.e. if the excitation signal is sufficiently rich. For SISO systems, one can verify whether a given excitation is sufficiently rich for a system based on the number of frequencies at which its power spectrum is nonzero. The extension of this criterion to multivariate systems is not straightforward. In the literature, one has proposed criteria based on the number of frequencies at which the power spectrum matrix of the excitation signal is strictly positive definite. However, this criterion is too restrictive as it does not cover the case of multisine excitations, while it is well known that such excitation signals can lead to consistent estimates. This paper proposes less restrictive conditions for the consistency of the identified parameter vector when FIR MISO systems have to be identified with multisine signals in the open-loop configuration.

Keywords: Identification, Identification for control, Signal processing
Abstract: Experiments in industrial plants to collect data for system identification are constrained due to production and safety requirements. In such situations, logged historical data can be instead used for system identification instead. However, these recorded data are predominantly stationary in continuously operated plants since processes are seldom excited during normal operation. Performing system identification with such data will yield numerical problems. Alternately, the ''most'' informative sequences can be extracted and used for system identification. Current data selection methods have several drawbacks. They are constrained to Single-Input Single-Output (SISO) modeling problems. The methods are not robust against correlated noise which is a disadvantage when using real data sets. Moreover, setting design parameters requires some information about the process that is not always available. In this contribution, an alternative data selection method for system identification is presented and evaluated in a case study. In contrast to current approaches, the proposed method does not require data normalization to detect transient changes. It can be used in Multi-Input Single-Output (MISO) systems operating in open or closed loop. An instrumental variables (IV) method is used in the algorithm which provides robustness against non-white noise. Results from a simulation case study of a multivariable system show that models with similar accuracy are obtained when using the intervals retrieved by the data selection method as when using the entire data set.

Keywords: Identification, Identification for control, Cooperative autonomous systems
Abstract: In this paper, we consider the problem of designing the least costly experiment that leads to a sufficiently accurate estimate of one module in a network of locally controlled systems. A module in such a network can be identified by exciting the corresponding local closed loop system. Such an excitation signal will not only perturb the input/output of the to-be-identified module, but also other modules due to the interconnection. Consequently, the cost of the identification can be expressed as the sum of the influence of the excitation signal on the inputs and outputs of all locally controlled systems. We develop a methodology to design the spectrum of the excitation signal in such a way that this cost is minimized while guaranteeing a certain accuracy for the identified model. We also propose an alternative identification configuration which can further reduce the propagation of the excitation signal to other modules and we make steps to robustify this optimal experiment design problem with respect to the cost of the identification.

Keywords: Model validation, Nonlinear system identification, Identification
Abstract: Model identification of a multivariate non-linear process is significantly related to its quality assessment. It can be measured using various methods starting from classical and canonical mean square error or the variance up to alternative fractal and persistence measures. Nested model of the real atmospheric distillation column plays the role of the benchmark example. The process is multivariate and strongly nonlinear. Its operation impacts significantly the overall refinery performance. One may find several papers describing column modeling with full spectrum of approaches staring from the first principle models up to regression black-boxes. Presented analysis compares nested static and NARIMA models using classical and alternative prediction error measures. The comparison of the considered indexes gives practical indication for the process modeling of real installations.

Keywords: Optimal control, Stochastic control, Stochastic systems
Abstract: Multistage risk-averse optimal control problems with nested conditional risk mappings are gaining popularity in various application domains. Risk-averse formulations interpolate between the classical expectation-based stochastic and minimax optimal control. This way, risk-averse problems aim at hedging against extreme low-probability events without being overly conservative. At the same time, risk-based constraints may be employed either as surrogates for chance (probabilistic) constraints or as a robustification of expectation-based constraints. Such multistage problems, however, have been identified as particularly hard to solve. We propose a decomposition method for such nested problems that allows us to solve them via efficient numerical optimization methods. Alongside, we propose a new form of risk constraints which accounts for the propagation of uncertainty in time.

Keywords: Stochastic filtering, Stochastic systems, Sensor and signal fusion
Abstract: We propose a geometry-driven deterministic sampling method for Bingham distributions in arbitrary dimensions. With flexibly adjustable sampling sizes, the novel scheme can generate equally weighted samples that satisfy requirements of the unscented transform and approximate higher-order shape information of the Bingham distribution. By leveraging retraction techniques from Riemannian geometry, the sigma points are constrained to preserve the second-order moment. Meanwhile, samples in each principal direction are located in a way that minimizes a distance measure between the on tangent-plane Dirac mixtures and the underlying on-manifold density. For that, the modified Cramér-von Mises distance based on the localized cumulative distribution (LCD) is employed. We further integrate the proposed approach into a quaternion-based orientation estimation framework. Compared to the existing unscented sampling approach drawing only fixed and limited numbers of sigma points, simulation results show that the proposed scheme enables better accuracy and robustness for nonlinear Bingham filtering.

Keywords: Stochastic systems, Game theoretical methods, Markov processes
Abstract: We consider in this paper the problem of a zero-sum two players linear quadratic (LQ) difference game for a stochastic discrete-time dynamical system subject to both random switching of its coefficients and multiplicative noise. Two kinds of admissible strategies are discussed. Following a Riccati equation approach, we show that in the solution of such a control problem, a crucial role is played by the stabilizing solution of an adequately defined class of coupled Riccati difference equations. For each type of admissible strategy the specific sign properties of the stabilizing solution of the involved Riccati equations are emphasized. The solution to this problem is given in a closed form.

Keywords: Stochastic systems, Markov processes, Linear time-varying systems
Abstract: In this paper we tackle some control problems related to the class of continuous-time, Markov jump linear systems. First of all, the annular stochastic finite-time stability problem is considered, and two different sufficient conditions are derived. Both conditions require the solution of a feasibility problem based on differential linear matrix inequalities. The analysis conditions are the starting point to solve the state-feedback design problem. Some numerical examples, considering also an electrical circuit, show the effectiveness of the proposed approach.

Keywords: Filtering, Nonlinear system identification, Optimization
Abstract: This paper addresses the set-valued filtering problem for discrete time-varying dynamical systems, whose process and measurement equations are polynomial functions of the system state. According to a set-membership framework, the process and measurement noises, as well as the initial state, are assumed to belong to bounded uncertainty regions, which are supposed to be generic semialgebraic sets described by polynomial inequalities. A sequential algorithm, based on sum-of-squares (SOS) representation of positive polynomials is proposed to compute a semialgebraic set described by an a-priori fixed number of polynomial constraints which is guaranteed to contain the true state of the system with certainty.

Keywords: Optimization algorithms, Optimal control, Filtering
Abstract: We propose an efficient algorithm, that given a strictly proper, second-order system, finds a sparse input signal so that the system's output optimally approximates a given trajectory in least-squares sense. As an illustration, we apply the algorithm to an estimation problem from medicine.

Keywords: Power electronics, Adaptive control, Observers for nonlinear systems
Abstract: This paper proposes a nonlinear, output feedback, adaptive controller to increase the stability margin of a direct-current (DC), small-scale, electrical network containing an unknown constant power load. Due to their negative incremental impedance, constant power loads are known to reduce the effective damping of a network, leading to voltage oscillations and even to network collapse. To overcome this drawback, we consider the incorporation of a controlled DC-DC power converter in parallel with the constant power load. The design of the control law for the converter is particularly challenging due to the existence of unmeasured states and unknown parameters. We propose a standard input-output linearization stage, to which a suitably tailored adaptive observer is added. The good performance of the controller is evaluated with numerical simulations.

Keywords: Power electronics, Adaptive control, Identification for control
Abstract: As the share of grid-connected converters increases, potential stability issues in the grid interface of the converters are emphasized. Impedance-based stability criterion can be used to assess the stability of the interface accurately, and the grid conditions should be considered in control design of the converter. However, the grid impedance is often an unknown parameter, and thus, adaptive control is a favorable method for addressing the uncertainties in the control design. This work presents an algorithm, in which the full-order grid impedance is measured during the start-up of the device, and control bandwidth of phase-locked loop and grid-voltage feedforward gains are adjusted to optimize sensitivity function in the grid interface, based on generalized Nyquist criterion. The method can ensure robust control in weak grid conditions, while still achieving good control performance in strong grids. The method is verified with power hardware-in-the-loop experiments with a kW-scale grid-connected three-phase inverter.

Keywords: Modeling, Power electronics
Abstract: The use of Hardware-in-the-loop (HIL) has been increasing recently owing to its flexibility in allowing an efficient and economical approach to promote real-time emulation. In the field of power electronics, HIL can be a useful solution since it allows testing control techniques in real controller devices without running the risk of damaging expensive components. However, real-time emulation requires powerful hardware to be performed, which has the drawback of increasing the HIL implementation's cost. Therefore, this paper proposes a low-cost HIL implementation for real-time emulation and control of a boost converter that uses the STM32 high-performance microcontroller unit (MCU). Stand-alone simulation on PSIM software and validation using an experimental boost converter are executed.

Keywords: Electrical power systems, Power electronics
Abstract: This paper tackles the current sharing problem for interconnected power converters. Specifically, it considers a single load, fed by parallel buck converters via a common DC bus. In such a case, it has been recently shown that dynamics related to current distribution can be controlled without impacting voltage regulation. Such a decoupling is performed without resorting to frequency separation argument, which inevitably lowers achievable performance. This decoupling can be destroyed by input constraints, though. In this context, this paper shows that a mere unilateral coupling can still be achieved: Using a novel control structure, foremost control goal of voltage regulation can be made independent of current distribution control. Controller design example exploiting the control scheme is provided together with numerical simulations.

Keywords: Power electronics, Electrical power systems
Abstract: Harmonic disturbance is an issue that can be solved using active filters connected at the point of common coupling (PCC). Power inverter based distributed generation systems (DGs) can operate as active filters as well as supplying power to the grid. Several works perform the total harmonic compensation at the PCC, which is in general not economically attractive considering that there are regulatory standards that establish the allowed limits of distortion. This paper proposes the use of a resonant proportional controller (PR) strategy for three-phase current-controlled voltage source inverters, present in microgrids to control the injected power into the grid and compensate partial harmonic components of current at the PCC. The compensation strategy is performed by adding proportional plus integral (PI) controllers to adjust the amplitude of the currents according to adopted harmonic distortion references. Scenarios with total and partial harmonic distortions compensation at the PCC were analyzed and the results showed that the use of the partial compensation strategy allowed a 10% increase in the value of the fundamental current in relation to the total compensation strategy.

Keywords: Power plants, Energy systems, Adaptive systems
Abstract: Since most of fuel cell models are generally non- linearly parameterized functions, existing modeling techniques for fuel cells mainly rely on the optimization approaches and impose heavy recursive computational costs. In this paper, an adaptive online parameter estimation approach for PEM fuel cells is developed in order to directly estimate unknown parameters and reduce computational burden. The general framework of this approach is that the electrochemical model is first reformulated by the Taylor series expansion. Then, one recently proposed adaptive parameter estimation method is fur- ther tailored to estimate unknown parameters. In this method, the adaptive law is directly driven by parameter estimation errors without using any predictors or observers. Moreover, the parameter estimation errors can be guaranteed to achieve exponential convergence. Besides, the online validation of the regressor matrix invertibility are avoided such that computation costs can be effectively reduced. Finally, comparative simulation results demonstrate that the proposed approach can achieve better performance than least square algorithms for estimating the unknown parameters of fuel cells.

Keywords: Differential algebraic systems, Uncertain systems, Robust control
Abstract: For a differential-algebraic system described by a polynomial or an improper transfer function matrix subject to perturbations acting on the normalized coprime factors we solve the robust stability problem. A formula for the maximum achievable stability margin and an explicit construction of the robust stabilizing controller are given in terms of realizations and solutions of the appropriate Riccati equations.

Keywords: Robust control, H2/H-infinity methods, Linear systems
Abstract: Bicoprime factorizations of the plant have recently resurfaced in control theory, after being a largely dormant subject for three decades. Recent results have shown that bicoprime factorizations can be beneficial in solving some control problems. In this paper, an additional advantage is presented where the use of bicoprime factorizations of the plant allows for robust control synthesis via reduced dimension algebraic Riccati equations by discarding some (or all) of the stable dynamics of the plant. This is a major advantage in the stabilisation of high order systems for which robust control synthesis can become computationally intractable.

Keywords: Robust control, H2/H-infinity methods, Mechatronics
Abstract: Various applications use high-pressure pumps for waterjet machining. A cutting head generates the waterjet, which is capable to cut hard and brittle materials. These cutting heads induce pressure fluctuations that originate from enabling and disabling the waterjet with respect to an a priori unknown switching law. Any pressure fluctuation will degrade the cutting quality. Thus, the primary control objective is to reject these disturbances by means of the high-pressure pump. Furthermore, waterjet applications show significant uncertainties that necessitate a robust control design. This paper presents the robust control design for a specific electrically driven high-pressure pump. Consequently, H-infinity controller synthesis is employed, taking into account expected parameter variations, exogenous disturbance and control requirements. This work introduces joint shaping functions, resulting in a reduced H-infinity optimization problem. Subsequently, the obtained controller is verified on a validated simulation model.

Keywords: Robust control, Model/Controller reduction, Linear time-varying systems
Abstract: Principal component active control is an important technique for noise and vibration reduction control problems. Most existing robust analysis regarding active control systems is based on the assumption of static open-loop behaviour, whose results can be very limited in practice. More promising results, exploiting integral quadratic constraints, have been reported recently and are more accurate and reliable for practical applications. However, depending on the nature of the system, such analysis may require one to establish the feasibility of an extremely large dimensional linear matrix inequality, effectively limiting the use of the result. This paper proposes an alternative approach, using model reduction methods, in which a set of IQC multipliers are obtained in a first step and the arising frequency domain inequality being verified in a second step. This two-step partition makes the approach much less computationally demanding for many complex practical systems. A detailed rotorcraft application is provided to illustrate the benefits of the proposed approach.

Keywords: Robust control, Sampled data control, Stability of nonlinear systems
Abstract: Robustification of controllers for the consensus problem of multi-agent systems (MASs) with actuator disturbances is analysed in this work. If a general consensus protocol is available for the disturbances-free system, a new sampled-data control law guarantees the agreement, in the sample-and-hold sense, with arbitrarily large and bounded actuator disturbance. Also the observation error is investigated. The agent drift dynamics are required to be globally Lipschitz and bounded. A numerical simulation example is finally proposed to validate the results.

Keywords: Robust control, Uncertain systems, Statistical learning
Abstract: The main objective of equity traders is to find a trading law leading to a safe profit, whatever the dynamics of the price be. Recenlty, a novel approach based on feedback control has been proposed, which allows the trader to treat the price as an exogenous disturbance to reject, rather than a stochastic process to be modelled. Although very promising, some questions arise, among which how to design the feedback controller defining the trading policy. In this work, we propose a reformulation of feedback trading as a trend following robust control problem, in which mild knowledge on the price range is exploited. The proposed approach is shown to outperform the state of the art methods on real-world stocks.

Keywords: Applications in neuroscience
Abstract: This tutorial provides an introduction to the topic of neuromodulation as an important control paradigm for natural and artificial neuronal networks. We will discuss neuromodulation as a control problem. We will review how neuromodulation modulates excitability, and how neuromodulation interacts with homeostasis. We stress how modulating nodal excitability provides a robust and versatile control principle to dynamically reconfigure the connectivity of rhythmic circuits and to shape the spatio-temporal synchrony of large and heterogeneous populations.

Keywords: Applications in neuroscience
Abstract: The architecture and functional principles of neural circuits are increasingly well understood both from an experimental and theoretical point of view. For example, we are now able to construct artificial neural circuits that operate analogously to biological principles and at the same time outperform many classical algorithms in decision and control. However, a core biological principle - neuromodulation - remains entirely absent from artificial neural circuit applications and lacks a mature theory in general. Neuromodulation refers to the capacity for biological neural circuits to undergo immediate parameter changes that affect the electrical properties of neurons as well as the connectivity in a circuit. This phenomenon underlies almost all animal behaviours from walking and chewing to sleep, defensive responses, attention and reward based learning. In this tutorial I will present an overview of the biological principles of neuromodulation, giving a sense of the history of its emergence as a concept and providing plenty of specific examples of how neuromodulation underpins the neural control of behaviour. I will articulate some of the main open questions and attempt to connect these to emerging system-theoretic models.

Keywords: Applications in neuroscience
Abstract: This talk will revisit classical questions of neurophysiology from a feedback control perspective. Specific questions include (i) the distinction between Type I and Type II excitability (ii) the role of the so-called 'cellular rebound' in thalamo-cortical oscillations and sleep; and (iii) the role of dopamine in regulating beta oscillations in motor control. In the three questions, we will show the specific role of slow positive feedback at a cellular level and discuss why it is counter-intuitive and has often led to conflicting interpretations. The three illustrations suggest novel ways to understand the role of cellular excitability in circuits and networks.

Keywords: Applications in neuroscience
Abstract: This tutorial provides an introduction to the topic of neuromodulation as an important control paradigm for natural and artificial neuronal networks. We review how neuromodulation modulates excitability, and how neuromodulation interacts with homeostasis. We stress how modulating nodal excitability provides a robust and versatile control principle to dynamically reconfigure the connectivity of rhythmic circuits and to shape the spatio-temporal synchrony of large populations.

Keywords: Statistical learning, Energy systems, Optimization
Abstract: Given the total power demand, P_d, current practice of equal load sharing in the process industry is to distribute the load among power supply units and machines (e.g., diesel/gas/wind turbines) in proportion to the maximum power, i.e., P_i = frac{P_{max}^i}{sum_j P^j_{max}} P_d, where P^i_{max} denotes the maximum power of the i^{textrm{th}} unit. However, the efficiency of power supply units, vary in time and are highly individual, even in the case of units from same brand and model. Thus, by considering and utilizing these individual differences, it is possible to share the load in a more fuel/cost/energy optimal manner. To capture this potential, the work presented in this paper proposes an optimization- and learning-based approach for sharing the load among a set of heterogeneous power supply units. The main contributions of this paper include formulation of the overall energy management system algorithm, including power grid connection and Energy Storage System (ESS), as a Mixed Integer Non-Linear Program (MINLP), as well as algorithms for robust moving horizon estimation of efficiency curves and machine learning-based algorithms for classification of turbine state and prediction of future power demand needed for the predictive planning scheme. The soundness and performance of the proposed algorithms are verified using actual data from a power plant including gas turbines and power grid connection in combination with ESS.

Keywords: Machine learning, Decentralized control, Distributed control
Abstract: In this paper we develop a fully decentralized algorithm for policy evaluation with off-policy learning and linear function approximation. The proposed algorithm is of the variance reduced kind and achieves linear convergence with O(1) memory requirements. We consider the case where a collection of agents have distinct and fixed size datasets gathered following different behavior policies (none of which is required to explore the full state space) and they all collaborate to evaluate a common target policy. The network approach allows all agents to converge to the optimal solution even in situations where neither agent can converge on its own without cooperation. We provide simulations to illustrate the effectiveness of the method in a Linear Quadratic Regulator (LQR) problem.

Keywords: Machine learning, Process control, Manufacturing processes
Abstract: Machine learning techniques have been widely applied to production processes with the aim of improving product quality, supporting decision-making, or implementing process diagnostics. These techniques proved particularly useful in manufacturing industry where huge variety of heterogeneous data, related to different production processes, can be gathered and recorded but where traditional models fail due to the complexity of the production process. In this study, we describe a novel Machine Learning methodology to associate some product attributes (either defects or desirable qualities) to process parameters. Namely we combine Support Vector Ma- chine (SVM) and the Support Vector Representation Machine (SVRM) to perform instance ranking. The combination of SVM and SVRM guarantees a high flexibility in modeling the decision surfaces (thanks to the kernels) while limiting overfitting (thanks to the principle of margin maximization). Thus, this method is well suited for modeling unknown, possibly complex relationships, that may not be captured by simple handcrafted models. We apply our method to production data of an investment casting industry placed in South Italy. We obtain an instance ranking that may be used to infer proper values of process parameter set-points.

Keywords: Machine learning, UAV's, Iterative learning control
Abstract: In this paper, the maze navigation problem for the unmanned ground vehicle (UGV) is considered. A new maze navigation scheme with Reinforcement Learning (RL) is proposed to find the optimal path from the entrance to the exit for the UGV. First, the quadrotor with a camera at its bottom is used to capture the image data of a random maze in a 3D view, and an image processing approach is implemented to reconstruct the maze in a virtual platform. Then, a novel exploration algorithm, Q-Learning(lambda) with improved epsilon-greedy (iepsilon-greedy),is proposed to find the shortest path in the reconstructed maze. The advantages of this paper are that the navigation scheme can provide an optimal path to the UGV without experiencing the maze cell by cell, and the proposed exploration algorithm can greatly reduce the randomness comparing to the traditional RL method. The effectiveness of both the proposed navigation scheme and the proposed exploration algorithm has been verified by simulations.

Keywords: Machine learning, Optimal control, Constrained control
Abstract: Enforcing state and input constraints during reinforcement learning (RL) in continuous state spaces is an open but crucial problem which remains a roadblock to using RL in safety-critical applications. This paper leverages invariant sets to update control policies within an approximate dynamic programming (ADP) framework that guarantees constraint satisfaction for all time and converges to the optimal policy (in a linear quadratic regulator sense) asymptotically. An algorithm for implementing the proposed constrained ADP approach in a data-driven manner is provided. The potential of this formalism is demonstrated via a numerical example.

Keywords: Machine learning, Randomized algorithms, Optimal control
Abstract: Q-functions are widely used in discrete-time learning and control to model future costs arising from a given control policy, when the initial state and input are given. Although some of their properties are understood, Q-functions generating optimal policies for continuous problems are usually hard to compute. Even when a system model is available, optimal control is generally difficult to achieve except in rare cases where an analytical solution happens to exist, or an explicit exact solution can be computed. It is typically necessary to discretize the state and action spaces, or parameterize the Q-function with a basis that can be hard to select a priori. This paper describes a model-based algorithm based on generalized Benders theory that yields ever-tighter outer-approximations of the optimal Q-function. Under a strong duality assumption, we prove that the algorithm yields an arbitrarily small Bellman optimality error at any finite number of arbitrary points in the state-input space, in finite iterations. Under additional assumptions, the same guarantee holds when the inputs are determined online by the algorithm's updating Q-function. We demonstrate these properties numerically on scalar and 8-dimensional systems.

Keywords: Decentralized control, Robust control, Complex systems
Abstract: In this paper, the safety synthesis problem for a discrete-time system comprised by multiple interconnected systems is considered. Using compositional reasoning, a quantitative framework is applied to each of the subsystems. With this framework it has been possible to derive robust controlled invariant subsets for each of the subsystems with respect to the control invariant subsets of the other subsystems. These invariant subsets can be computed from a parameterized family of sets and they share a common safety controller. Contract-based design is utilized to built assume-guarantee contracts for all the subsystems, namely to assume that the other subsystems belong to their invariant sets and guarantee that the subsystem will belong to its invariant set. This circularity of the implications can be resolved by a fixed point algorithm which computes the parameters to guarantee that all the subsystems fulfil their contracts simultaneously. Then, the invariant set and the safety controller are given for the original system. To illustrate the effectiveness of the proposed approach, an application for the temperature regulation of adjacent rooms of a building is given as an example.

Keywords: Distributed control, Energy systems, Concensus control and estimation
Abstract: We consider small-scale power systems consisting of several inverter-interfaced units connected in parallel to a common bus, the point of common coupling (PCC), and sharing a joint load. This is a frequently encountered configuration in microgrid applications. In such a setting, two important control objectives are reactive power sharing and voltage regulation at the PCC. In this paper, we first show that the nonlinear equilibrium equations corresponding to the aforementioned objectives admit a unique positive solution. Then, we propose a consensus-based distributed voltage controller which renders this desired unique solution locally asymptotically stable. Finally, control performance is illustrated via a simulation example.

Keywords: Distributed control, Game theoretical methods, Constrained control
Abstract: In this paper, we show the equivalence between a constrained, multi-agent control problem, modeled within the port-Hamiltonian framework, and an exact potential game. Specifically, critical distance-based constraints determine a network of double-integrator agents, which can be represented as a graph. Virtual couplings, i.e., pairs of spring-damper, assigned to each edge of the graph, allow to synthesize a distributed, gradient-based control law that steers the network to an invariant set of stable configurations. We characterize the points belonging to such set as Nash equilibria of the associated potential game, relating the parameters of the virtual couplings with the equilibrium seeking problem, since they are crucial to shape the transient behavior (i.e., the convergence) and, ideally, the set of achievable equilibria.

Keywords: Distributed control, Optimal control of communication networks, Optimization algorithms
Abstract: In this paper we deal with distributed optimal control for nonlinear dynamical systems over graph, that is large-scale systems in which the dynamics of each subsystem depends on neighboring states only. Starting from a previous work in which we designed a partially distributed solution based on a cloud, here we propose a fully-distributed algorithm. The key novelty of the approach in this paper is the design of a sparse controller to stabilize trajectories of the nonlinear system at each iteration of the distributed algorithm. The proposed controller is based on the design of a stabilizing controller for polytopic Linear Parameter Varying (LPV) systems satisfying nonconvex sparsity constraints. Thanks to a suitable choice of vertex matrices and to an iterative procedure using convex approximations of the nonconvex matrix problem, we are able to design a controller in which each agent can locally compute the feedback gains at each iteration by simply combining coefficients of some vertex matrices that can be pre-computed offline. We show the effectiveness of the strategy on simulations performed on a multi-agent formation control problem.

Keywords: Distributed control, Robust control, Randomized algorithms
Abstract: This paper proposes probabilistic distributed control design method to a linear plantwide system with generic parametric uncertainty in the behavioural framework. Parametric quadratic differential forms (QDF) are used to characterise the dynamic properties of the processes and controllers in the form of dissipativity properties as well as imposing plantwide requirements. The scenario approach is then used to convert the parametric linear matrix inequalities (LMI) to deterministic ones, hence making them solvable using finite algorithms. The controllers are synthesised in parallel directly from their respective supply rates through J-factorisation.

Keywords: Distributed cooperative control over networks, Concensus control and estimation, Optimal control
Abstract: In this paper, we consider a leader-follower problem for a group of homogeneous linear time invariant (LTI) follower agents that are interacting over a directed acyclic graph. In our problem of interest, only a subset of the follower agents has access to the state of the leader in specific sampling times. The dynamics of the leader that generates its states is unknown to the followers. For interaction topologies in which the leader is a global sink in the graph, we propose a distributed algorithm that allows the agents to arrive at the sampled state of the leader before the next sample arrives. We prove that the control input to take the followers from one sampled state to the next one is minimum energy for all the followers. We also show that after the first sampling epoch, the states of all the follower agents are synchronized with each other. We demonstrate the application of our proposed algorithm for two leader-follower problems for mobile agents. Our first example shows the application of our algorithm in control of unicycle robots in a formation motion. In the second example, we demonstrate the use of our algorithm for reference state tracking via a group of second order integrator followers with bounded control. In this example, we show that the properties of our proposed leader-follower algorithm allow us to design the arrival times at the reference states in such a way that the input bounds of the agents never get violated.

Keywords: Concensus control and estimation, Adaptive systems, Distributed parameter systems
Abstract: This paper presents a framework for online adaptive estimation of unknown or uncertain systems of nonlinear ordinary differential equation (ODEs) that characterize a multiagent sensor network. This paper extends recent results in cite{Parag2018,kurdila1} and here the nonlinear ODEs are embedded in the real, vector-valued reproducing kernel Hilbert space (RKHS) mathbb{H}:=H^N with H a real, scalar RKHS. Each agent casts its local representation of the unknown function f as a member of the RKHS H. The result defines a distributed parameter system that governs the state estimates and estimates of the unknown function. The convergence of state estimates is proven along similar lines to that encountered in conventional adaptive estimation for systems of unknown nonlinear ODEs. The analysis of the parameter estimates, which is studied by an evolution in Euclidean space in conventional methods, now concerns the convergence of error functions in the RKHS. We show that the convergence of the function estimates to the unknown function in the RKHS is guaranteed provided a newly introduced persistency of excitation (PE) condition holds. This PE condition is defined on functions defined over a subset Omega that contains the trajectory of the true dynamic system. It can be viewed as an extension of the notion of partial persistence of excitation to the RKHS embedding framework.

Keywords: Concensus control and estimation, Agents and autonomous systems, Optimal control
Abstract: We introduce a reachable set based method of leader selection, for achieving minimum time consensus in networks of first order discrete time systems. The systems, also referred as agents, interact over a weighted communication graph, using distributed control protocol proposed by Vicsek. In addition, a bounded external input is applied to one of the agents, designated a-priory, as the leader. If a time- optimal external input is applied to an agent, the states of the agents are driven from given initial condition to a consensus state in minimum time. The algorithm proposed in this paper identifies the leader such that, for the given initial condition, the minimum time to consensus through the leader, is the least among all possible choices. We show that the state-space of the relative state dynamics can be partitioned using the boundaries of the reachable sets.

Keywords: Concensus control and estimation, Cooperative autonomous systems, Distributed cooperative control over networks
Abstract: This paper considers the tracking control problem of a network of identical linear state-space models with directed topology. It is assumed that communication topology contains a rooted directed spanning tree, where root has the knowledge of the states and control input of reference trajectory. Though the control input of reference is not known to all, the structure of the dynamics of reference control input is known to all agents. A distributive consensus-type tracking control protocol is proposed based on the relative information of states in the neighbourhood of each agent with a consensus-based estimator that estimates the reference input. Condition on coupling strength has been derived such that errors between states of the agents and the reference trajectory, and the estimated control inputs and control input of reference trajectory decay at a determined rate. Computational complexity reduces abundantly for an undirected communication topology. The effectiveness of the proposed controller is demonstrated through numerical simulations.

Keywords: Concensus control and estimation, Distributed cooperative control over networks, Cooperative control
Abstract: Continuous-time average consensus protocols have been deeply analyzed in recent years. This kind of protocols enable the network to reach the agreement on the average of the initial conditions in a fully distributed way. Moreover, the convergence of such protocols is proved to be asymptotic. However, several applications need the network to reach the agreement in finite time. On the other hand, often a convergence to a ball centered at the average value is sufficient in most application contexts. In this paper, a method for distributively detecting when the continuous-time average consensus protocol has reached the agreement within a given error margin is presented. This distributed detection takes finite time and occurs at each node simultaneously. Numerical results show the effectiveness of the proposed approach.

Keywords: Network analysis and control, Optimization algorithms, Distributed cooperative control over networks
Abstract: In this paper, we deal with the containment control problem in presence of antagonistic interactions. In particular, we focus on the cases in which it is not possible to contain the entire network due to a constrained number of control signals. In this scenario, we study the problem of selecting the nodes where control signals have to be injected to maximize the number of contained nodes. Leveraging graph condensations, we find a suboptimal and computationally efficient solution to this problem, which can be implemented by solving an integer linear problem. The effectiveness of the selection strategy is illustrated through representative simulations.

Keywords: Network analysis and control, Control over networks, Biological systems
Abstract: Understanding the fundamental mechanisms enabling fast and reliable communication in the brain is one of the outstanding key challenges in neuroscience. In this work, we address this problem from a systems and information theoretic perspective. Specifically, we first develop a simple yet tractable framework to model information transmission in networks driven by linear dynamics. We then resort to the notion of Shannon capacity to quantify the information transfer performance of these networks. Building on this framework, we show that it is possible to increase Shannon capacity via two fundamentally different mechanisms: either by decreasing the degree of stability of the network adjacency matrix, or by increasing its degree of non-normality. We illustrate and validate our findings by means of simple yet insightful examples.

Keywords: Mechatronics, Robotics, MEMS
Abstract: Positive acceleration, velocity and position feedback (PAVPF) control scheme has been successfully applied to piezoelectrically actuated nanopositioning systems to suppress resonance impelled vibrations. This control design shows a couple of improvements like high bandwidth control. The design framework is based on the principle in which damping is imparted by shifting the closed loop poles arbitrarily further into the left-half plane. In the proposed control design, a polynomial based pole placement approach was used to simultaneously damp and track the closed-loop poles to achieve a larger bandwidth control and reduced tracking error than the traditional PAVPF implementation as shown by experimental results.

Keywords: Algebraic/geometric methods, Mechatronics, Output feedback
Abstract: Partial differential equations (PDEs) persist to be a stumbling block in Interconnection and Damping Assignment Passivity-Based Control (IDA-PBC). Lately, for a class of mechanical systems an energy shaping controller has been investigated that avoids PDEs by exploiting implicit port-Hamiltonian representations. Following the same research line, in this paper we generalize the total energy shaping IDA-PBC for underactuated mechanical systems in the implicit framework and propose an algebraic solution for a class of systems. Besides, under some additional conditions we are also able to produce a respective output feedback. A reduction of the closed loop system in explicit coordinates is presented. We test our results on the inclined cart-pole system and portal crane.

Keywords: Mechatronics, Neural networks
Abstract: The hydraulics of heavy-duty transmissions typically lack pressure sensors, which implies that the hydraulic actuation system has to be controlled in a feedforward manner. In practical applications, this is often done by numerically inverting a static mapping, leading to undesired pressure errors and inaccurate clutch positioning especially in dynamic situations. This paper therefore investigates two feedforward control strategies for a hydraulic clutch actuation path. The first is a flatness-based feedforward control strategy that relies on a high accuracy model, and additionally accounts for input constraints. The second one is a data-driven neural network-based feedforward control strategy that relies on measurement data. The feedforward controls are first evaluated in simulations and subsequently evaluated experimentally for a real-world transmission that demonstrates the realizability on ECU level as well as the effectiveness of the flatness-based and neural network-based feedforward controls compared to the standard static mapping-based approach.

Keywords: Mechatronics, Observers for linear systems, Machine learning
Abstract: In this paper, a Kalman filter (KF) based method for the accurate estimation of the dynamic positioning error of the tool-center-point (TCP) of a high-precision measuring machine is presented. A generalizing approach consisting of a linear physical model of the dynamic TCP deviations and a data-based model, which is realized as an additive Gaussian process (GP) trained on the physical model error, is applied. On one hand, the TCP position can be measured using a novel camera-based sensor which yields the absolute positioning error at a relatively slow sampling rate. On the other hand, the GP predicts the model mismatch at the fast base sample rate and can be treated as an additional pseudo measurement. A multi-rate (MR) observer in the KF framework yields an improved estimate of the TCP position compared to a KF using only the camera measurements. Simulation results show the potential of the proposed MR-KF approach using a combined physical and data-based model.

Keywords: Mechatronics, Predictive control for linear systems, Energy systems
Abstract: This paper presents a Model Predictive Control of the flange torque of a Wind Turbine System Test Bench drive train as the core of a Hardware-in-the-Loop simulation. A System Test Bench enables the efficient and reproducible investigation of the complex behavior of an entire Wind Turbine drive train only without the rotor and tower, as these complicate the test significantly. Instead, a motor applies the torque, yet the altered drive train setup changes its dynamic considerably. This results in unrealistic loads and thus makes the operation with the original Wind Turbine control impossible. The proposed control restores the original behavior by controlling the flange torque between the Wind Turbine and Test Bench drive train according to a reference from a Hardware-in-the-Loop simulation of the rotor. It considers the motor constraints and is robust both against deviation of its internal model and time delay. In simulation, the Model Predictive Control emulates the realistic Wind Turbine drive train dynamic perfectly in part load, though with less performance, yet still robustly, in full load due to the motor constraints. It is already implemented on a dSPACE system integrated in the System Test Bench with a research Wind Turbine. The Hardware-in-the-Loop concept itself is validated experimentally with a simpler control with less performance.

Keywords: Stability of nonlinear systems, Automotive
Abstract: Electric vehicle (EV) operation is based on specific electromechanical and energy storage systems driven by suitable power electronic devices. The whole system constitutes a nonlinear Euler-Lagrange system with controlled inputs the duty-ratio components of i) the voltage source converter used as power interface between the electric motor and the output of the dc storage device, and, ii) the dc/dc converter duty-ratio used to automatically regulate the charging condition of the storage devices. Adopting cascaded mode control techniques, that have been appropriately modified for the present case, and developing the entire incremental model of the system, a systematic method for tuning the gains of the fast inner-loop controllers is deployed. To this end, suitable Lyapunov techniques are applied which simultaneously prove passivity at a first stage, and, global asymptotic stability at the desired nonzero equilibrium in a next one. The proposed controller response is examined by extensive simulations and the results fully confirm an excellent system performance, further verifying the rigorous analysis of the theoretical part.

Keywords: Stability of nonlinear systems, Autonomous systems, Lyapunov methods
Abstract: In this paper, we study the finite-time stability of a class of nonlinear systems dot{x}= f(x) = H(x)b(x); where H is homogeneous and b is bounded. We define the homogeneous extension of the non-homogeneous function f and use this extension to prove that, under some conditions on b; if the system dot{x}= f(x) is globally asymptotically stable, then it is finite-time stable. An example of global asymptotic stablesystem with some additional conditions is presented in the last section to illustrate the obtained results.

Keywords: Stability of nonlinear systems, Lyapunov methods, Robotics
Abstract: The problem of a state feedback design for control of a quadrotor system under state and time constraints is studied. Convex embedding approach and Implicit Lyapunov function are employed to design a finite-time controller. The feedback gain is solved by a system of LMIs(Linear Matrix Inequalities). Theoretical results are supported with numerical simulation.

Keywords: Stability of nonlinear systems, Lyapunov methods
Abstract: The problem of tracking an arbitrary curve in the state space is considered for underactuated driftless control-affine systems. This problem is formulated as the stabilization of a time-varying family of sets associated with a neighborhood of the reference curve. An explicit control design scheme is proposed for the class of controllable systems whose degree of nonholonomy is equal to 1. It is shown that the trajectories of the closed-loop system converge exponentially to any given neighborhood of the reference curve provided that the solutions are defined in the sense of sampling. This convergence property is also illustrated numerically by several examples of nonholonomic systems of degrees 1 and 2.

Keywords: Stability of nonlinear systems, MEMS, Nonlinear system theory
Abstract: This paper demonstrates the effectiveness of a modified Linear Integral Resonant Controller based on its original LTI cousin, known just as the `IRC', for suppressing Jump-Phenomena found in MEMS and other Duffing-Type systems wherein the primary nonlinearity is that of the cubic nonlinearity. A Method of Multiple Scales frequency response is derived, explored and compared with a Runge-Kutta based numerical integration method in order to understand any shortcomings in approximate analytical methods for the analysis of closed-loop nonlinear systems with the inclusion of a stability analysis. It is found that there exist some mild behavioural inconsistencies when comparing closed-loop Method of Multiple Scales to traditional numerical integration. Finally, it is shown that with sensibly chosen controller gains, a MEMS with Jump-Phenomena can be made to behave similarly to a linear second order resonant system opening up the possibilities of Laplace Domain and Linear State-Space techniques once more.

Keywords: Stability of nonlinear systems, Robust control, Lyapunov methods
Abstract: This paper studies the robustness properties of homogeneous finite-time stable systems with respect to multiplicative perturbation for sufficiently small inputs. Robust stability conditions are presented for the systems admitting homogeneous approximation at the origin and at infinity. The utility of the obtained results is illustrated via robustness analysis of homogeneous observer with time-varying gains.

Keywords: Biological systems, Biomolecular systems, Genetic regulatory systems
Abstract: Genetic oscillators have a fundamental role in the regulation not only of intracellular, but also of intercellular functions: for instance, in the segmentation clock, the synchronised oscillation of neighbouring cells generates spatial travelling waves that induce segmentation of precursors of the vertebral column. To investigate this type of phenomena, we consider the behaviour of two genetic negative feedback oscillators, each operating in a different cell, coupled by an intercellular positive-feedback interconnection with delays. The two coupled systems are nominally identical, but can be different due to noise and cell-to-cell variability. When they can be different in general, we study the effect of positive-feedback and of delay in inducing an oscillatory behaviour. When they are identical, we study how the intercellular feedback delay affects the phase difference between the two oscillators.

Keywords: Biological systems, Genetic regulatory systems, Output regulation
Abstract: The genetic toggle switch is a MIMO control system that can be controlled by varying the concentrations of two inducer molecules, aTc and IPTG, to achieve a desired level of expression of the two genes it comprises. It has been shown in Lugagne et al., Nature Communication (2017) that this can be accomplished through an open-loop external control strategy where the two inputs are selected as mutually exclusive periodic pulse waves of appropriate amplitude and duty-cycle. In this paper, we use a recently derived average model of the genetic toggle switch subject to these inputs to synthesize new feedback control approaches that adjust the inputs’ duty-cycle in real-time via two different possible strategies, a model-based hybrid PI-PWM approach and a so-called Zero-Average dynamics (ZAD) controller.

Keywords: Biological systems, Genetic regulatory systems, Cellular dynamics
Abstract: In this paper, we propose and analyse a regulatory network characterized by the ability of detecting chemical concentration ranges of ligand molecules. In particular, we consider quorum sensing mechanism found in engineered Escherichia Coli bacterium and involved in the regulation of the Green Fluorescence Protein. The considered network overall implements an incoherent feed-forward loop, a network motif in which an activator regulates both a gene and a repressor of the gene. Such a motif can lead to very localized spatial emergent behaviours regulated by the chemical concentration. As customary in biological circuits modelling, molecular interactions are described by means of Hill functions. The main contribution of the present work is twofold: first, gradient concentration detection requires specific conditions on molecular species’ properties (namely, the incoherent feed-forward loop structure alone is not sufficient); and secondly promoter cooperativity is crucial to control the width of the concentration range detected by the circuit. Theoretical analysis is complemented with numerical simulations.

Keywords: Biological systems, Biomolecular systems, Genetic regulatory systems
Abstract: Biological systems compute spatial and temporal gradients with a variety of mechanisms, some of which have been shown to include integral feedback. In traditional engineering fields, it is well known that integral components within a negative feedback loop can be used to perform a derivative action. In this paper, we define the concept of a practical differentiator that is inspired by this design principle. We then consider three simple biological circuit examples in which we prove that feedback combined with ultrasensitive, quasi- integral components yields a practical differential network under some assumptions. These examples include phosphorylation/dephosphorylation cycles, and two networks relying on molecular sequestration.

Keywords: Genetic regulatory systems
Abstract: The expression of individual genes can be maintained through positive feedback loop mechanisms. If genes are expressed in bursts, then feedback either affects the frequency with which bursts occur or their size. Here we use a tractable hybrid modelling framework to evaluate how noncooperative positive feedback in burst frequency or burst size impacts the protein-level distribution. We confirm the results of previous studies that noncooperative positive feedback in burst frequency can support bimodal distributions. Intriguingly, bimodal distributions are unavailable in the case of feedback in burst size in the hybrid framework. However, kinetic Monte Carlo simulations of a full discrete model show that bimodality can reappear due to low-copy number effects. The two types of feedbacks lead to dramatically different values of protein mean and noise. We show that small values of leakage imply a small protein mean for feedback in burst frequency but not necessarily for feedback in burst size. We also show that protein noise decreases in response to gene activation if feedback is in burst frequency but there is a secondary noise amplification if feedback acts on burst size. Our results suggest that feedback in burst size and feedback in burst frequency may play fundamentally different roles in maintaining and controlling stochastic gene expression.

Keywords: Stochastic systems, Biomolecular systems, Biological systems
Abstract: The realization of embedded molecular control systems is a challenging aim in Synthetic Biology, where a major goal is to design synthetic biological circuits performing specific tasks. In this field, the novel emergent approach is to assemble the circuit in a modular fashion, possibly restraining reciprocal interactions from interconnected modules (zero-retroactivity). Within this framework, recent results have been proposed, dealing with the realization of an embedded subtractor module, with the idea of exploiting it in a more general chemical reaction network that resembles a classical control scheme. So far, this research has been carried out according to the deterministic approach. More sophisticated analysis requires the use of stochastic models, which play a paramount role in investigating noise propagation in chemical reaction networks, especially when the species copy number is low and the intrinsic stochasticity of the phenomena under investigation cannot be neglected. This note deals with a first analysis of the subtractor module, according to the stochastic approach. To this end, Chemical Master Equations are exploited to model one of the possible molecular circuits implementing the subtractor, and moment equations are written in order to evaluate how noise propagates with respect to different values of the inputs and different model parameter settings.

Keywords: Automotive, Transportation systems, Intelligent systems
Abstract: In the last few years, smartphone usage has been unanimously considered to be one of the most dangerous driving habits, leading to numerous road accidents. In the context of insurance telematics, the online estimation of the driving style is of utmost importance to offer personalized policies and to evaluate riskiness levels. Smartphones can be used to gather dynamics measures needed to build riskiness indicators to evaluate drivers’ behaviour. Being able to recognize phone usage is of paramount importance. This work deals with the problem of detecting the use of the phone during a trip. To do this in a fully automatic fashion, we propose a new cepstrum-based classification method based on time series analysis. The resulting performance is tested on experimental data and compared with those obtained with another method, based on hand-crafted features, showing that the new approach yields fully comparable performance while significantly increasing the automation level of the classification process.

Keywords: Automotive, H2/H-infinity methods, Robust control
Abstract: This paper focuses on lateral control of autonomous vehicles. Two different approaches, among the most important ones dealing with a multi-objective synthesis problem, are presented and compared for the design of an efficient and comfortable assistance. The main aim here is to determine the best approach to meet the specifications of industrial applications. On one hand, a Linear Parameter-Varying (LPV) controller is synthesized, taking explicitly into account the speed and acceleration variations, through a dedicated LMI-based approach. On the other hand, a Gain-Scheduling (GS) controller is synthesized using existing non-smooth optimization algorithms. The comparison is carried out in the H₂ framework and the sought controller has a fixed structure, namely a Static Output Feedback (SOF) form to ease the real-time implementation. A detailed comparison of the two approaches is presented and based on simulations using real measurements of a camera.

Keywords: Automotive, H2/H-infinity methods
Abstract: Shared control is a solution which already showed beneficial effects during experiments for steering control in the automotive sector. It improves the tracking performance of pilots, by assisting them but without taking them out of the loop, making them able to deal with unexpected situations. This article presents a novel shared control strategy for car steering, using an original two parts method. On the one hand, the feedforward part consists of a trajectory generator based on the simulation of a virtual autonomous vehicle. On the other hand, a mixed H2/H∞ control law constitutes the feedback part that is applied to the difference between the real and the virtual vehicle’s states. A first benefit of such an architecture is that the two parts can be designed sequentially. Second, the virtual vehicle can be used to enlarge the possibilities of shared control's implementation to various ADAS: LKA, LCA, lane change, obstacle avoidance, etc. Finally, the sharing level between the human driver and the assistance is parameterizing the control synthesis process, and could be adapted during driving, e.g. for gradual switching between autonomous and manual modes. The proposed shared control strategy showed interesting results during simulation, both in terms of performance and quality of driving sharing.

Keywords: Automotive, Optimal control, Autonomous systems
Abstract: This paper presents a drifting stabilizing controller for a rear-wheel-driven car, leveraging on front tyre steering angle and longitudinal force developed by rear tyres, the same control inputs available to a human driver. The proposed controller is based on a linear-quadratic regulator designed on a linearised single-track model of the vehicle, so that both longitudinal and lateral velocities along with yaw rate are stabilized. The controller has been experimentally validated on a scaled car. An extensive experimental campaign has been performed to demonstrate the robustness of this approach along with its shortcomings, that will be addressed in future works.

Keywords: Automotive, Optimization, Mechatronics
Abstract: This paper deals with the optimal energy management control of series hybrid electric vehicles (HEVs) by which their fuel consumption is minimized. Based on the modeling of the essential dynamics of a series HEV and its powertrain, closed-form solutions of the optimal energy source power split are formulated. These solutions yield a set of simple heuristic control rules, by which a novel rule-based control strategy, the truncated battery power following strategy (TBFS), is developed for optimally allocating the hybrid energy sources to satisfy the demanded propulsion power. Numerical examples show that the TBFS is able to reproduce the globally optimal solutions found by dynamic programming (DP), but with significantly reduced computation effort, as the TBFS only requires one-dimensional parameter tuning, rather than solving an optimization problem. The results also indicate its benchmark potential for high-fidelity HEV models, where DP is no longer applicable.

Keywords: Automotive, Transportation systems, Autonomous systems
Abstract: The paper proposes the design of an MPC-based method for overtaking maneuvers of autonomous vehicles. The strategy incorporates a graph-based optimization algorithm, in which the probability of collisions with the surrounding vehicles is incorporated. The purpose of the graph search is to determine the road and velocity profile of the autonomous vehicle, which is guaranteed by the proposed MPC control. The novel algorithm is able to consider the multi-vehicle scenarios during the overtaking maneuver. The effectiveness of the method is presented in CarMaker simulation scenarios.

Keywords: Hybrid systems, Modeling, Automata
Abstract: Zeno hybrid systems exhibit an infinite number of switches between different configurations in finite time. This is the result of the abstraction of modeling and the assumption that the switching is instantaneous. Simulation of such systems will generally stop or give false results after a finite time, called the Zeno time. In reality, physical systems do not exhibit such behavior, because the switching is not in reality instantaneous, but also often because they are subject to a small amount of noise. This noise makes possible the continuous evolution of the real systems even if extremely fast switching arises. Thus, the problem is how to predict the behavior of the system after the Zeno time. In this paper a new technique is proposed to extend simulations beyond the Zeno time. It consists of a probabilistic regularization by a convolution method that includes noise in the system. For example, low-amplitude Gaussian noise can be used to transform discontinuous vector fields to continuous vector fields. This convolution approach makes the value observed for the continuous evolution over time slightly different from what would be predicted without the noise, but in a way that should correspond to what happens in reality. An example of a Zeno hybrid system is employed to illustrate the result. The evolution beyond the Zeno time varies depending on the noise that exists in physical systems and on how the noise is included in the regularized model. However, a useful idealization is obtained in the limit of low noise amplitude.

Keywords: Optimal control, Hybrid systems, Robotics
Abstract: A numerical framework for finding and stabilizing periodic trajectories of underactuated mechanical systems with impacts is presented. By parameterizing a trajectory by a set of synchronization functions, whose parameters we search for, the dynamical constraints arising due to underactuation can be reduced to a single equation on integral form. This allows for the discretization of the planning problem into a parametric nonlinear programming problem by Gauss-Legendre quadratures. A convenient set of candidates for transverse coordinates are then introduced. The origin of these coordinates correspond to the target motion, along which their dynamics can be analytically linearized. This allows for the design of an orbitally stabilizing feedback controller, which is also applicable for degrees of underactuation higher than one.

Keywords: Stability of hybrid systems, Constrained control, Lyapunov methods
Abstract: For a vehicle moving in an n-dimensional Euclidean space, we present a construction of a hybrid feedback that guarantees both global asymptotic stabilization of a reference position and avoidance of an obstacle corresponding to a bounded spherical region. The proposed hybrid control algorithm switches between two modes of operation: stabilization (motion-to-goal) and avoidance (boundary-following). The geometric construction of the flow and jump sets of the hybrid controller, exploiting a hysteresis region, guarantees robust switching (chattering-free) between stabilization and avoidance. Simulation results illustrate the performance of the proposed hybrid control approach for a 3-dimensional scenario.

Keywords: Switched systems, Hybrid systems, Stability of hybrid systems
Abstract: As an extension of standard switched systems, this paper presents a new formulation of generalized switched systems. Such a generalization allows switching among different types of systems. Hence a hybrid system can be treated as a special case of such generalized switched systems. With the design freedom gained from this formulation, it is possible to design an appropriate switching signal between a family of continuous-time systems and a family of discrete-time systems to achieve a better performance. This paper introduces fundamental definitions and stability results. The concept of output persistent excitation condition is proposed, which links closely to uniform global asymptotic stability. To illustrate the usefulness of this new formulation, stability results of the hybrid systems are revisited to show how this formulation works.

Keywords: Switched systems, Lyapunov methods
Abstract: This paper extends some existing methods of safety verification of nonlinear systems to the case of nonlinear switched systems with time-dependent switching. Barrier functions and barrier densities are applied to safety verification of nonlinear switched systems. In particular, sufficient conditions for safety of switched systems are given based on the existence of a common Barrier function or a common Barrier density. Theoretical results are exemplified applying a sum of squares method with the Putinar positivstellensatz to find barrier functions and barrier densities certifying safety.

Keywords: Predictive control for nonlinear systems, Robust adaptive control, Process control
Abstract: We address the problem of robust nonlinear model predictive control (NMPC) under plant-model mismatch in the absence of full-state measurements. We propose an approach that is based on the use of a model-error model (MEM) to handle the estimation errors and the structural plant-model mismatch in a Multi-stage NMPC framework. The MEM which consists of a linear model followed by a nonlinear operato with bounded gain captures the estimation error along with the unmodeled dynamics of the plant. Multi-stage NMPC explicitly considers the presence of feedback in the problem formulation, hence it is less conservative than other robust NMPC schemes. The advantages of the proposed scheme are demonstrated on a benchmark reactor problem.

Keywords: Predictive control for nonlinear systems, Distributed control, Large-scale systems
Abstract: We consider the interconnection of dissipative systems through coupling costs in a distributed economic MPC context. Our goal is to provide a structured dissipativity property for the overall system emerging from the subsystems' local dissiaptivity properties and their interconnection. However, following this bottom-up approach, only in very few cases we can expect to establish dissipativity of the overall system based on a minimal set of assumptions. Hence, in this work we introduce the concept of approximate dissipativity, which still allows us to make statements on the system performance, albeit somewhat unsharp. We verify approximate dissipativity for the overall system under cost interconnection of the subsystems, and we demonstrate how this concept can constructively be exploited when adding a new subsystem to the network.

Keywords: Predictive control for nonlinear systems, Constrained control, Output feedback
Abstract: In this paper, we present a simple methodology to design nonlinear output-feedback model predictive control (MPC) schemes. The design procedure is applicable to a large class of nonlinear systems and guarantees constraint satisfaction despite noise and disturbances. We utilize an existing observer with guaranteed exponential stability properties in combination with an initial bound on the estimation error in order to predict valid bounds on the possible future estimation error. The predicted estimation error is then used online to appropriately tighten the state and input constraints, using recently developed nonlinear robust MPC methods based on incremental stabilizability properties. The resulting nonlinear output-feedback MPC scheme is simple to implement, only marginally increases the computational demand (compared to a nominal MPC scheme), and ensures robust constraint satisfaction and input-to-state stability w.r.t. disturbances/noise. We demonstrate the simplicity and applicability of the proposed approach with a numerical example.

Keywords: Predictive control for nonlinear systems, Neural networks
Abstract: The goal of this paper is to investigate the theoretical properties, the training algorithm, and the predictive control applications of Echo State Networks (ESNs), a particular kind of Recurrent Neural Networks. First, a condition guaranteeing incremetal global asymptotic stability is devised. Then, a modified training algorithm allowing for dimensionality reduction of ESNs is presented. Eventually, a model predictive controller is designed to solve the tracking problem, relying on ESNs as the model of the system. Numerical results concerning the predictive control of a nonlinear process for pH neutralization confirm the effectiveness of the proposed algorithms for the identification, dimensionality reduction, and the control design for ESNs.

Keywords: Predictive control for nonlinear systems, Optimization algorithms, Optimal control
Abstract: In this paper, we present an early termination condition for the primal-dual interior-point method for application in nonlinear model predictive control (NMPC) problems. The condition verifies the prescribed suboptimality level of a feasible iteration of the algorithm and enables one to employ the feasible suboptimal solution without jeopardising the stability of the system. The distinguishing property of the proposed condition is its primal-dual formulation, which allows the proposed early termination of the algorithm to be independent from any values computed on the previous time instances. Numeral experiments on a nonlinear planar multirotor system and a comparison of the proposed early termination condition with the existing methods are provided.

Keywords: Predictive control for nonlinear systems, Autonomous systems, Automotive
Abstract: In this paper, we provide a novel robust collision avoidance approach that is based on a general tube-based MPC framework. We consider collision avoidance for general nonlinear uncertain systems with moving obstacles. The resulting optimization problem can be handled by standard nonlinear programming solvers. Moreover, we provide formal guarantees, such as recursive feasibility, constraint satisfaction, as well as robust collision avoidance. We demonstrate the efficacy of the proposed method through a simulation of an autonomous car during realistic manoeuvres.

Keywords: Identification, Identification for control, Linear systems
Abstract: Most industrial processes, even though complex in nature, can be represented using second order plus dead-time models, usually determined from step response data, which capture the essential process dynamics. The parameters of these models are computed based on specific algorithms. However, the great majority of these algorithms require system identification basic knowledge and are thus difficult to be used by the process engineer. The focus of this paper is to introduce a new method for computing the parameters of such models. The major advantage over existing methods is that it does not require any system identification expertise, being fully automatic. Additional advantages include robustness to noise, disturbances and system order. All of these are emphasized through several numerical examples, as well as an experimental validation.

Keywords: Identification, Markov processes, Delay systems
Abstract: In this paper we propose a solution to the problems of detecting a stochastic output delay sequence characterized by a Markov chain and estimating the state of a linear system driven by Gaussian noise through an augmented delay-state dynamics. This is the model for uncertain observations resulting from losses in the propagation channel due to fading phenomena or packet dropouts that is common in wireless sensor networks, networked control systems, or remote sensing applications. The solution we propose consists of two parallel stages: a nonlinear detector, which identifies at each time instant the delay and a filtering stage. Numerical simulations show the performance of the proposed method.

Keywords: Identification, Modeling, Neural networks
Abstract: Atmospheric air pollution is one of the main environmental problems that our society is facing. Moreover, according to the World Health Organization it is a major worldwide environmental risk to health. Due to these facts, Decision Support Systems (DSSs) have been developed to help Environmental Authorities in designing short and long terms air quality plans to cost-efficiently control the impacts of atmospheric pollution. A key component of the DSSs is the air quality forecasting system, needed to compute the pollutant concentrations in advance with respect to the occurrence of critical events. These models can adopt either a deterministic or a statistical approach. In both cases, the resulting models are characterized by intrinsic strengths and weaknesses. This work proposes an approach to develop and implement air quality forecasting models by integrating these two approaches to reap their benefits while avoiding or minimizing the disadvantages, focusing on the often neglected field of short term air pollution. This integration is done by implementing a reanalysis algorithm allowing to rely on the complexity and accuracy of deterministic models and on the performances of statistical models, limiting, at the same time, the frequent concentration underestimation of deterministic models and the statistical models spatial limi- tations. Such approach has been tested by identifying models to reproduce the daily mean concentrations of particulate matter on Lombardy region, a highly polluted area in Northern Italy.

Keywords: Identification, Reduced order modeling, Optimization
Abstract: Laguerre models are a simple solution for orthonormal basis function modelling, but such simplicity is accompanied by limitations to model processes whose poles are sparse and when the system has zeros. In this article conditions for position and ratio optimisation of two real poles are analytically derived and used to develop methods for the estimation of real two-parameter Kautz models. Both the models obtained by means of the proposed methods approximated a process with sparse poles and zero with greater accuracy than the second-order Laguerre model.

Keywords: Identification, Sensor and mesh networks, Signal processing
Abstract: The identification of vector autoregressive (VAR) processes from partial samples is a relevant problem motivated by several applications in finance, econometrics, and networked systems (including social networks). The literature proposes several estimation algorithms, leveraging on the fact that these models can be interpreted as random Markov processes with covariance matrices satisfying Yule-Walker equations. In this paper, we address the problem of identification of distributed vector autoregressive (DVAR) processes from partial samples. The DVAR theory builds on the assumption that several processes are evolving in time, and the transition matrices of each process share some common characteristics. First, we discuss different models for describing the coupling among single processes. Subsequently, we propose an estimator for the transition matrices of the DVAR processes adopting an Empirical Bayes approach. More precisely, the local parameters are treated as random variables with a partially-unknown a priori density function, chosen as the conjugate family of distributions defined over symmetric, nonnegative-definite matrix-valued random variables and parameterized by suitable unknown hyperparameters. We develop an optimization algorithm to obtain the maximum likelihood estimates of the hyperparameters. The main feature of the proposed approach is that it does not require exact knowledge of the model describing the coupling between the different VAR processes, and it proves particularly well suited in scenarios in which the number of samples are allowed to be highly inhomogeneous or incomplete. The proposed techniques are validated on a numerical problem arising in social networks estimation.

Keywords: Identification, Signal processing, Stochastic systems
Abstract: In the literature, the problem of frequency estimation is usually cast in a nonlinear parametric fashion. We show in this paper that it can also be formulated in a Bayesian nonparametric framework by assigning a uniform a priori probability distribution to the unknown frequency. We find that the covariance matrix of signal model is the discrete-time analogue of the integral operator whose eigenfunctions are the famous prolate spheroidal wave functions, introduced by Slepian and coworkers in the 1960's. Two methods are proposed to estimate the hyperparameters of the prior distribution. One uses techniques which are essentially linear from subspace identification. The other is based on Prediction Error Minimization. An explicit formula for the predictor is derived exploiting the exceptional decay property of the eigenvalues of the covariance matrix. In addition, a covariance estimation scheme is suggested assuming panel data in response to the nonergodicity of the process. The approach seems quite promising.

Keywords: Stochastic filtering, Nonlinear system identification, Optimization algorithms
Abstract: This paper presents a method for parameter estimation using a particle filter, where the parameters are elements on the manifold SO(3). The presented method is based on Stochastic Gradient Descent, which makes it possible to perform the static parameter estimation on-line. The presented method is coordinate-free and is therefore not prone to singularities.

The proposed method is demonstrated for estimating the motion of a crane load, where markers on the crane load facilitate visual tracking. The pose of each marker relative to the crane load is initially unknown, but are found within four seconds using the proposed method.

Keywords: Distributed estimation over sensor nets, Sensor and mesh networks, Sensor and signal fusion
Abstract: This paper addresses a novel multiple target tracking (MTT) problem in a decentralized sensors network (DSN) framework. The algorithm jointly estimates the number of targets and the states of the targets from a noisy measurement in the presence of data association uncertainty and missed detection. The standard GM-PHD filters estimate the multi-targets in a cluttered environment with an assumption that the target birth intensity is known or homogeneous. It results in inefficient tracking for new, occluded or missed targets. The issue is addressed by the proposed adaptive Gaussian birth components based estimation. A method based on covariance intersection fusion is proposed to address inter-sensor target data association.

Keywords: Stochastic systems, Stochastic control, Biological systems
Abstract: First-passage time (FPT) of an Ornstein- Uhlenbeck (OU) process is of interest in various contexts. This paper investigates tuning of FPT moments of an OU process. The region of interest is defined by two boundaries, out of which at least one is absorbing. We find that the FPT distribution of an OU process is scale invariant with respect to the drift parameter, i.e., the drift parameter just controls the mean FPT and does not affect the shape of the distribution. This facilitates independent control of the mean and the coefficient of variation (CV) of the FPT. We also explore the effect of control parameters on the FPT distribution, and find parameters that minimize the distance between the FPT distribution and a desired distribution.

Keywords: Algebraic/geometric methods
Abstract: In this paper, we propose a variation-based extended Kalman filter (V-EKF) on the two-sphere manifold. We consider the spherical pendulum dynamical system whose nonlinear geometric dynamics evolve on the two-sphere. These dynamics are linearized about the current state using a variation-based linearization resulting in a time-varying linear system with state constraints that describe the dynamics of the variation states. The Kalman filter is applied on the resulting variation states with the pendulum position as measurements for measurement updates. The V-EKF also has a const update where the estimated state and covariance are updated to ensure they satisfy the constraints. Desirable properties of V-EKF, such as preserving the geometric structure of the estimated state are thus achieved. The proposed method is illustrated through numerical simulations and also validated through experiments.

Keywords: Control over communication, Stochastic control, Signal processing
Abstract: In this paper, we investigate the role of a physical watermarking signal in quickest detection of a deception attack in a scalar linear control system where the sensor measurements can be replaced by an arbitrary stationary signal generated by an attacker. By adding a random watermarking signal to the control action, the controller designs a sequential test based on a Cumulative Sum (CUSUM) method that accumulates the log-likelihood ratio of the joint distribution of the residue and the watermarking signal (under attack) and the joint distribution of the innovations and the watermarking signal under no attack. As the average detection delay in such tests is asymptotically (as the false alarm rate goes to zero) upper bounded by a quantity inversely proportional to the Kullback-Leibler divergence(KLD) measure between the two joint distributions mentioned above, we analyze the effect of the watermarking signal variance on the above KLD. We also analyze the increase in the LQG control cost due to the watermarking signal, and show that there is a tradeoff between quick detection of attacks and the penalty in the control cost. It is shown that by considering a sequential detection test based on the joint distributions of residue/innovations and the watermarking signal, as opposed to the distributions of the residue/innovations only, we can achieve a higher KLD, thus resulting in a reduced average detection delay. We also present some new structural results involving the associated KLD and its behaviour with respect to the attacker's signal power and the watermarking signal power. These somewhat non-intuitive structural results can be used by either the attacker to choose their power to minimize the KLD, and/or by the system designer to choose its watermarking signal variance appropriately to increase the KLD. Numerical results are provided to support our claims.

Keywords: Electrical power systems, Optimization, Distributed control
Abstract: Economic dispatch of interconnected microgrids that is based on distributed model predictive control (DMPC) requires the cooperation of all agents (microgrids). This paper discusses the case in which some of the agents might not comply with the decisions computed by performing a DMPC algorithm. In this regard, these agents could obtain a better performance at the cost of degrading the performance of the network as a whole. A resilient distributed method that can deal with such issues is proposed and studied in this paper. The method consists of two parts. The first part is to ensure that the decisions obtained from the algorithm are robustly feasible against most of the attacks with high confidence. In this part, we employ a two-step randomization-based approach to obtain a feasible solution with a predefined level of confidence. The second part consists in the identification and mitigation of the adversarial agents, which utilizes hypothesis testing with Bayesian inference and requires each agent to solve a mixed-integer problem to decide the connections with its neighbors. In addition, an analysis of the decisions computed using the stochastic approach and the outcome of the identification and mitigation method is provided. The performance of the proposed approach is also shown through numerical simulations.

Keywords: Electrical power systems, Control over networks, Distributed control
Abstract: Secondary control architectures for islanded direct-current microgrids are getting interest since they are necessary to manage the voltage references in order to properly distribute the time-varying load demand. To this aim, we propose three different optimization-based secondary control approaches considering the internal units constraints, losses minimization and the continuous satisfaction of the load demand. The described approaches are based on a centralized, distributed and cluster-based optimization strategy, respectively.

Keywords: Electrical power systems, Decentralized control, Hybrid systems
Abstract: In this paper, we address the problem of voltage stability and power sharing in DC microgrids with time-varying power demand. By exploiting the monotonicity property enjoyed by the system, and under the assumption of full observability of the bus voltages, we design a centralized, abstraction-based symbolic controller that, once refined into a controller for the original system, ensures the required specifications. Whereas load voltages cannot be measured, we propose an appropriate decomposition of the system, such that the control problem can be reformulated in terms of assume-guarantee contracts to be satisfied by the observable and unobservable components. A constructive procedure to determine suitable contracts is further investigated and the obtained results are validated with two numerical examples.

Keywords: Power plants, Identification, Electrical power systems
Abstract: An identification algorithm is proposed to estimate varying parameters and some unknown input in a MISO continuous-time system, for the purpose of monitoring frequency control performance in hydroelectric generating units. In the proposed algorithm, an integral transform with a Fourier series-based kernel is developed from irregularly sampled data and its efficiency is shown by comparison with real hydropower plant parameters and with a Savitzky-Golay filter.

Keywords: Electrical power systems, Discrete event systems, Emerging control applications
Abstract: In this paper, a discrete event approach is proposed for the optimal charging of electrical vehicles in microgrids. In particular, the considered system is characterized by renewable energy sources (RES), non-renewable energy sources, electrical storage, a connection to the external grid and a charging station for electric vehicles (EVs). The decision variables are relevant to the schedule of production plants, storage systems and EVs’ charging. The objective function to be minimized is related to the cost of purchasing energy from the external grid, the use of non-renewable energy sources and tardiness of customer’s service. The proposed approach is applied to a real case study and it is shown that it allows to considerably reduce the dimension of the problem (and thus the computational time required) as compared to a discrete-time approach.

Keywords: Electrical power systems, Energy systems, Optimization
Abstract: A two-stage unit commitment (UC) method for power systems under load uncertainty is proposed, where one of the worst-case or expected operating costs is minimized with the other bounded above by a predefined value. It is shown that the UC problem can be formulated as a mixed-integer linear programming problem and solved via Benders decomposition. While the weighted sum model (WSM) can be used to consider both types of operating costs simultaneously, it is not straightforward to find a pair of weights that lead to an acceptable UC solution reflecting the system operator's preference. The proposed method is more intuitive than the WSM in that its parameter, i.e., the allowable maximum worst-case or expected operating costs, has the same unit as that of the objective value, i.e., a monetary unit. A numerical experiment demonstrates the effectiveness of the proposed method.

Keywords: Robust control, Optimal control, Robotics
Abstract: This work proposes a nonlinear W-infinity controller for a convertible tilt-rotor Unmanned Aerial Vehicle (UAV) in order to solve the trajectory tracking problem during the helicopter-flight mode. The control design for such aircraft is challenging since it is a multi-body, highly-coupled, underactuated mechanical system. Therefore, a controller based on the novel nonlinear W-infinity control approach is designed aiming to guide the translational position and yaw angle toward a desired trajectory, while the remaining degrees of freedom are stabilized. Numerical experiments are carried out to corroborate the efficiency of the control strategy demonstrating robustness, good transient performance and fast response against external disturbances.

Keywords: Robust control, Optimization, Energy systems
Abstract: This paper develops a robust optimization framework for the day-ahead energy scheduling of a grid-connected residential user. The system incorporates a renewable energy source (RES), a battery energy storage system (BESS) as well as elastic controllable and critical non-controllable electrical appliances. The proposed approach copes with the fluctuation and intermittence of the RES generation and non-controllable load demand by a tractable robust optimization scheme requiring minimum information on the sources of uncertainty. The main objective is minimizing the total energy payment for the user considering operational/technical constraints and a contractual constraint penalizing the excessive use of energy. The presented framework allows the decision maker to define different robustness levels for uncertain variables, and to flexibly establish an equilibrium between user’s payment and price of robustness. To validate the effectiveness of the proposed framework under uncertainty, we simulate the dynamics of a residential user as a case study. A comparison between the proposed robust approach and the same method with deterministic RES and loads profiles is carried out and discussed.

Keywords: Robust control, Stochastic systems, Stochastic filtering
Abstract: This paper addresses L1 analysis and synthesis problems for a class of Markov jump linear systems, under the assumption that the jump process degenerates to a sequence of independent, identically distributed switches. This class of systems has received a great deal of attention in recent years, for instance, in the study of networked control systems. We consider here the scenario that comes up when all system coefficients are positive - to the best of our knowledge, this scenario has never been considered for Bernoulli jump systems. The main results include the tight characterization of two different metrics for L1 performance, along with efficient methods for the analysis and synthesis of a positive L1 filter that ensures robustness, in a quite general scenario where the system coefficients are subject to linear fractional uncertainty. The results are expressed in terms of linear programs of considerably smaller dimension than the existing methods that treat Markov jump linear systems with more general Markov chain structures.

Keywords: Robust control, Uncertain systems, Transportation systems
Abstract: Dynamic robustness of a control system is a crucial requirement for self-driving cars. To this aim, longitudinal control is one of the key tools to effectively track a desired reference profile, that in a smart road scenario can be imposed by a road infrastructure communication. Indeed, uncertainties and disturbances due to environmental, road and traffic conditions strongly impact on the autonomous vehicle motion. For this reason, it is necessary considering an accurate electromechanical model, with realistic uncertain conditions, actuator limitations, and measurement noises. In this paper, an easy and robust smooth control method is proposed, allowing an autonomous electric vehicle to track a sufficiently smooth reference signal with an error in absolute value smaller than a prescribed value, despite bounded parametric uncertainties, disturbances and velocity measurement noises. The theoretical analysis is illustrated through an interesting application, confirming the effectiveness of the proposed method.

Keywords: Robust control
Abstract: A hierarchical composite nonlinear feedback (HCNF) control technique is proposed in which the auxiliary integrators are introduced to remove the steady-state bias due to the disturbances. The main idea of HCNF is to separate the control design of the auxiliary integrators and the composite nonlinear feedback (CNF) control design. The auxiliary integrators are applied only to remove the steady-state bias by automatically adjusting the commands to be tracked by the CNF control law in which the CNF control design does not involve with the auxiliary integrators. The HCNF control technique is applied to design the attitude control law of a maneuvering flight control law for a quadrotor. The resulting closed-loop system is verified in simulation successfully to complete the maneuvering flight.

Keywords: Aerospace, Robust control, Flexible structures
Abstract: The paper proposes an alternative methodology to build Linear Fractional Transformation (LFT) models of uncertain aeroelastic systems described by Fluid-Structure Interaction (FSI) solvers with the aim of studying flutter with the mu analysis technique from robust control. Two main issues can be identified for the fulfillment of this task. On the one hand, there is the difficult reconciliation between sources of physical uncertainty (well distinguishable in the original high-order system) and the abstracted uncertainties (defined in the reduced-order size representation used for the robust analyses). On the other hand, the large size of the resulting LFT model can prevent the application of robust analysis techniques. The solution proposed here consists of a symbolic LFT algorithm applied at FSI solver level, which guarantees the connection between the physical uncertainties and the parameters captured by the LFT. It also alleviates the final size of the LFT by exploiting the modal-oriented approach taken in introducing the uncertainties. Application of the framework using an unconventional aircraft layout as case study is finally discussed.

Keywords: Safety critical systems
Abstract: Reports of cyber-attacks, such as Stuxnet, have shown their devastating consequences on digitally controlled systems supporting modern societies, and shed light on their modus operandi: first learn sensitive information about the system, then tamper the visible information so the attack is undetected, and meanwhile have significant impact on the physical system. Securing control systems against such complex attacks requires a systematic and thorough approach. Risk management is a fundamental approach to build security, which critically depends on the ability to characterize, analyze, and rank attack scenarios in terms of their risk (i.e., impact and likelihood). The security of the control system is then built by deploying mitigation schemes targeted at high-risk (high-impact, high-likelihood) scenarios. In the first part of this talk, we shall provide an overview of the recent work on secure control systems centered around the risk management framework and its main stages: scenario characterization, risk analysis, and risk mitigation. In particular, we shall consider malicious attacks on key security properties such as confidentiality, integrity, availability, and map different attack policies into important characteristics of the adversaries, namely their access to resources enabling the disclosure and disruption of data, and model information. The second part of the talk will then provide insights into specific attack scenarios not covered by the specialized talks of the session, such as certain stealthy attacks, and discuss various approaches to analyze and mitigate them.

Keywords: Safety critical systems
Abstract: Emerging systems such as smart grids or intelligent transportation systems often require end-user applications to continuously send information to external data aggregators performing monitoring and control tasks. Despite their benefits, these applications can thus lead to an undesirable loss of privacy for the users, who might become legitimately concerned about providing their data. To address this issue, this talk gives an introduction to a formal definition of privacy called differential privacy, which provides strong privacy guarantees, even in the presence of adversaries with arbitrary side information. We then illustrate how tools from systems and control theory can be used to design real-time estimators releasing aggregate signals about a population of individuals, while balancing the accuracy of the information provided against the level of differential privacy offered.

Keywords: Safety critical systems
Abstract: Control systems work silently in the background to support much of the critical infrastructure we have grown used to. Water distribution networks, sewer networks, gas and oil networks, and the power grid are just a few examples of critical infrastructure that rely on control systems for its normal operation. These systems are becoming increasingly networked both for distributed control and sensing, as well as for remote monitoring and reconfiguration. Unfortunately, once these systems become connected to the internet they become vulnerable to attacks that, although launched in the cyber domain, have for objective the manipulation of the physical domain. In this talk I will discuss the problem of state-estimation and control for linear dynamical systems when some of the sensor measurements are subject to an adversarial attack. I will show that a separation result holds so that controlling physical systems under active adversaries can be reduced to a state-estimation problem under active adversaries. I will characterize the maximal number of attacked sensors under which state estimation is possible and propose computationally efficient estimation algorithms.

Keywords: Safety critical systems
Abstract: Recent advances in sensing, communication and computing allow cost effective deployment in the physical world of large-scale networks of sensors and actuators, e.g. IoT, enabling fine grain monitoring and control of a multitude of physical systems and infrastructures. Such CPS lie at the intersection of sensing, communication, computing and control. The close interplay among these fields and the resulting complexity render independent design of subsystems a risky approach, as separation of concerns does not constitute a realistic assumption in real world scenarios. It is therefore imperative to derive new models and methodologies to allow analysis and design of robust and secure CPS. In this talk I will present an overview of recent research on the topic security, ranging from issues of detection to resilient control and discuss future directions. I will discuss unique opportunities for detection in CPS, discuss fundamental limitation of passive detection methods and propose a number of active detection techniques, based on the concepts of physical watermarking and moving target defense. In the first small perturbations are introduced in the control systems and the their effect at the output will be used to authenticate sensor measurements. Moving target defense is an active detection technique capable of detecting attacks from powerful malicious agents, who can compromise secrecy and integrity of all inputs and measurements. Here some system parameters are changed at random so that the attacker neither can create fake measurements compatible with the predicted system behavior nor can identify a suitable model of the system. We will also propose two resilient control methodologies capable of countering attacks without compromising safety and liveness properties for the system, thus guaranteeing its survivability.

Keywords: Safety critical systems
Abstract: Owing to advances in computing and communication technologies, recent years have witnessed a growing interest towards CPS, namely systems where physical processes are monitored/controlled via embedded computers and networks. The concept of CPS is very appealing for automation but raises many theoretical and practical challenges, in particular how to design control systems that are resilient against cyber attacks. This tutorial provides an overview of some recent developments in the area of networked control in the presence of Denial-of-Service (DoS) attacks. The tutorial has two parts. Central to the first part is the modelling of DoS attacks. Inspired by the information science community, we discuss a model that constrains attacks only in terms of their frequency and duration. This has the merit to describe DoS patterns ranging from short and frequent attacks (e.g. jamming burst noise) to long but sporadic attacks (e.g. temporary channel saturation). The second part of the tutorial discusses some recent developments in this area for both centralized and distributed settings. The aim of the second part of the tutorial is also to discuss new research directions and open problems.

Keywords: Emerging control applications, Control over networks, Safety critical systems
Abstract: This tutorial provides a high-level introduction to novel control-theoretic approaches for the security and privacy of cyber-physical systems (CPS). It takes a risk-based approach to the problem and develops a model framework that allows us to introduce and relate many of the recent contributions to the area. In particular, we explore the concept of risk in the context of CPS under cyber-attacks, paying special attention to the characterization of attack scenarios and to the interpretation of impact and likelihood for CPS. The risk management framework is then used to give an overview of and map different contributions in the area to three core parts of the framework: attack scenario description, quantification of impact and likelihood, and mitigation strategies. The overview is by no means complete, but it illustrates the breadth of the problems considered and the control-theoretic solutions proposed so far.

Keywords: Fault detection and identification, Linear systems, Control over networks
Abstract: Cyber-physical systems (CPSs) have become targets for malicious adversaries, creating new challenges for attaining reliable CPS performance. Achieving CPS security requires tools which extend beyond what is offered in state of the art software and cyber security. In this tutorial, we consider a science of CPS security which combines tools from both cyber security and system theory to defend against adversarial behavior. We discuss realistic and intelligent attack models from an adversarial perspective and then present mechanisms for defenders to achieve resilience by recognizing and responding to such malicious behavior. The focus of this tutorial is particularly on the recognition and detection of attacks, which is the first and foremost step in achieving resilience and is a necessary prerequisite to any active response.

Keywords: Optimal control
Abstract: Optimal control is an attractive framework for addressing complex control tasks, as it offers the promise of determining the best possible action given the dynamic and other constraints of the system. With some notable exceptions, however, optimal control problems are notoriously difficult to solve even for moderate size systems. It is known that many optimal control problems encoded as dynamic programs can equivalently be characterised through the solution of linear programs. Replacing the (often infinite) linear program by a simpler (finite) counterpart leads to methods for approximating the solution of the original optimal control problem, in the spirit of Approximate Dynamic Programming. In this talk we outline such an approach to approximate optimal control and discuss how results in randomised optimisation can be leveraged to derive bounds on the errors incurred in the process. The approach also offers a vista towards data driven control, as in some cases the approximation can be carried out using sample paths of the system evolution, bypassing the need to explicitly know or identify the dynamic constraints.

Keywords: Machine learning, Nonlinear system identification, Adaptive control
Abstract: Their flexibility to learn general function classes renders nonparametric regression algorithms particularly attractive in system identification and data-based control settings, where little a priori knowledge about a dynamical system is to be presumed. Building on approaches known as NSM- or Lipschitz regression, we propose a new nonparametic machine learning approach. While it inherits theoretical learning guarantees from the methods it is built upon, it is designed to limit the computational effort both for learning and for generating predictions. This renders our method applicable to online system identification and control settings where the desired sample frequency precludes previous nonparametric approaches from being deployed. Apart from deriving a guarantee on the ability of our method to learn any continuous function, we illustrate some of its practical merits on a number of benchmark comparison problems.

Keywords: Optimization algorithms, Machine learning, Randomized algorithms
Abstract: In this paper we consider constrained convex optimization problems, where the constraints are described as the intersection of a finite number of convex sets. Each set of the intersection is assumed to be simple, that is the projection can be computed efficiently. For these problems we derive gradient methods with random projections, that is we perform a gradient step for minimizing the objective function and then access the feasible set only through the projection onto a random individual set. Thus, our algorithms can easily address streaming settings, where the whole feasible set is not known in advance, but it is rather learned in time through observations, as in machine learning applications. Also, these algorithms are of interest for optimization problems having a very large number of given constraints, as in constrained predictive control. We analyze the convergence behavior of the proposed algorithms for the case when the objective function is strongly convex and with Lipschitz continuous gradients. We prove linear convergence rate into a noise dominated region for the expected quadratic distance of the iterates from the optimal set for constant stepsize. However, our analysis allows to also provide global sublinear rates for diminishing~stepsize. Numerical evidence supports the effectiveness of our methods in real-world problems.

Keywords: Optimization, Machine learning, Statistical learning
Abstract: We study the approximate dynamic programming approach to revenue management in the context of attended home delivery. We draw on results from dynamic programming theory for Markov decision problems, convex optimisation and discrete convex analysis to show that the underlying dynamic programming operator has a unique fixed point. Moreover, we also show that -- under certain assumptions -- for all time steps in the dynamic program, the value function admits a continuous extension, which is a finite-valued, concave function of its state variables. This result opens the road for achieving scalable implementations of the proposed formulation, as it allows making informed choices of basis functions in an approximate dynamic programming context. We illustrate our findings using a simple numerical example and conclude with suggestions on how our results can be exploited in future work to obtain closer approximations of the value function.

Keywords: Modeling, Machine learning, Predictive control for nonlinear systems
Abstract: Autonomous operation of industrial plants requires a cheap and efficient way of creating dynamic process models, which can then be used to either be part of the autonomous systems or to serve as simulators for reinforcement learning. The trends of digitalization, cheap storage, and industry 4.0 enable the access to more and more historical data that can be used in data driven methods to perform system identification. Model predictive control (MPC) is a promising advanced control framework, which might be part of autonomous plants or contribute to some extent to autonomy. In this article, we combine data-driven modeling with MPC and investigate how to train, validate, and incorporate a special recurrent neural network (RNN) architecture into an MPC framework. The proposed structure is designed for being scalable and applicable to a wide range of multiple-input multiple-output (MIMO) systems encountered in industrial applications. The training, validation, and closed-loop control using RNNs are demonstrated in an industrial simulation case study. The results show that the proposed framework performs well dealing with challenging practical conditions such as MIMO control, nonlinearities, noise, and time delays.

Keywords: Medical signal processing, Machine learning, Fault diagnosis
Abstract: The aim of this work is to develop an efficient data-driven method for automatic medical decision making, especially for cardiac arrhythmia diagnosis. To achieve this goal, we have targeted the most common arrhythmia worldwide —atrial fibrillation (AF). Most of reported studies are dealing with inter-beat interval time series analysis coupled with univariate and/or multivariate data-driven methods. The state of the art of this subject revealed that although satisfactory detection findings have been achieved for long AF durations, there is still scope for improvement which needs to be addressed for brief episodes which is highly desired by healthcare professionals. Relevance vector machine (RVM) has been developed to address this issue. Several kernel functions and parameters have been tested to optimize RVM. Five geometrical and nonlinear features were extracted from 30s inter-beat time series. The RVM classifier was trained on 3000 randomly selected samples from four publicly-accessible sets of clinical data and tested on 1000 samples. The performance of the diagnosis model was evaluated by 10-fold cross-validation method. The results showed that the RVM model performed better than do existing algorithms, with 96.58% success rate. The automatic diagnosis on another dataset of 118985 samples of AF and Normal Sinus Rhythm (NSR) has yield 96.64% of classification accuracy. This automated data-driven decision making approach can be exploited for medical diagnosis of other arrhythmias.

Keywords: Fault detection and identification, Machine learning
Abstract: Component insertion is a common task in robotic assembly, and is widely used for manufacturing a variety of electronic devices. This task is generally characterized by low tolerances, thus requiring high precision during assembly. An early detection of a fault in the mating during the insertion process enables quality control of the end products, as well as safeguards the robotic equipment. We propose to use Gaussian Process Regression-based methods to learn the force profile during successful insertions, as well as quantify permissible deviations from this profile. The GPR model is then used to detect anomalies in case the observed force profile deviates significantly from the expected range. Apart from the standard GPR formulation, we consider two other variants - the Heteroscedastic GPR and the local GPR for better modeling accuracy and computational time efficiency, respectively. We report an accuracy of 100% in differentiating between normal and faulty insertions. The modeling and detection results indicate that our approach is accurate and robust to severe uncertainties due to process (e.g., force drift) and measurement noise.

Keywords: Agents and autonomous systems, Distributed control, Concensus control and estimation
Abstract: We consider a multi-population, represented by a network of groups of individuals. Every player of each group can choose between two options, and we study the problem of reaching consensus. The dynamics not only depend on the dynamics within the group, but they also depend on the topology of the network, so neighboring groups influence individuals as well. First, we develop a mathematical model of this networked bio-inspired evolutionary behavior and we study its steady-state. We look at the special case where the underlying network topology is a regular and unweighted graph and show that the steady-state is a consensus equilibrium. A sufficient condition for exponential stability is given. Finally, a related networked dynamical system with connections to the Bass model is studied and we conclude the paper with simulations.

Keywords: Game theoretical methods, Agents networks, Control over networks
Abstract: In this paper we consider a set of heterogeneous agents modelled as higher-order integrators, playing a game over a network. In such networked scenarios, agents have to make decisions compatible with seeking a Nash equilibrium, while using partial-networked information, and possibly rejecting disturbances. We propose dynamic agent decision-making based on gradient-play with an additional stabilizing component for the higher-order dynamics. In the partial-information setting, each agent makes its decision based on a dynamic estimate of the others' states, updated by local communication with its neighbours, which offsets the lack of global information. When external disturbances are present, the agent decision dynamics is augmented with an internal-model component, in the form of a reduced-order observer for the disturbance. We show convergence of agents' dynamics to the Nash equilibrium, irrespective of disturbances. Our proofs leverage input-to-state stability under strong monotonicity of the pseudo-gradient and Lipschitz continuity of extended pseudo-gradient. Applications to mobile robots in sensor networks are provided.

Keywords: Game theoretical methods, Markov processes, Agents networks
Abstract: This work studies a class of multi-player games in which the players' decisions can be influenced by a superplayer. We define a game with n players and parameterized utilities u(cdot, alpha) where the superplayer controls the value of alpha. The regular players follow Markovian repeated play dynamics that encompass a wide class of learning dynamics including strict best response. The objective of the superplayer is to control alpha dynamically to achieve a desired outcome in the game-play, which in this work we define as the realization of target joint strategies. We introduce the class of parametric games and reformulate the superplayer control problem as a Markov decision process (MDP). Reachability criteria are developed, allowing the superplayer to determine which game-play may occur with positive probability. With a reachable goal joint strategy, a emph{cost}-optimal policy can be computed using standard tools in dynamic programming. A sample MDP reward function is presented such that a reachable target joint strategy is guaranteed to be played almost surely. Finally, an application in a cyber-security context is provided to illustrate the use of the proposed methodology and its effectiveness.

Keywords: Game theoretical methods, Distributed control, Cooperative autonomous systems
Abstract: Networked systems are ubiquitous in today's world with examples spanning from ecology to the social and engineering sciences. Much of the research in networked systems is analytical, where the focus is on characterizing (and potentially influencing) the emergent collective behavior. A more recent trend of research focuses on the design of networked systems capable of achieving diverse and highly coordinated collective behavior in the absence of centralized control. Focusing on the well-studied class of maximum coverage problems, our first result demonstrates that any agent-based algorithm relying solely on local information induces a fundamental trade-off between the best and worst case performance guarantees, as measured by the price of anarchy and price of stability. Our second results demonstrates how to use an additional piece of system-level information to breach these limitations, thereby improving the system's performance.

Keywords: Game theoretical methods, Computational methods
Abstract: Many binary choice multi-agent problems, where one agent may have some private information, can be modeled as 2xn bimatrix games. We show here that all 2xn bimatrix games that are full rank are strategically equivalent to rank-1 games. Given the 2xn bimatrix game, we exploit the strategic equivalence among nonzero-sum games to derive another bimatrix game, which is a rank-1 game. We then devise a new polynomial time algorithm to solve 2xn bimatrix games. We conjecture that this algorithm has comparable performance to existing polynomial time algorithms for 2xn bimatrix games, such as support enumeration. We then comment on some applications and extensions.

Keywords: Traffic control, Optimization, Stochastic systems
Abstract: This paper introduces an optimization problem (P) and a solution strategy to design variable-speed-limit controls for a highway that is subject to traffic congestion and uncertain vehicle arrival and departure. By employing a finite data-set of samples of the uncertain variables, we aim to find a data-driven solution that has a guaranteed out-of-sample performance. In principle, such formulation leads to an intractable problem (P) as the distribution of the uncertainty variable is unknown. By adopting a distributionally robust optimization approach, this work presents a tractable reformulation of (P) and an efficient algorithm that provides a suboptimal solution that retains the out-of-sample performance guarantee. A simulation illustrates the effectiveness of this method.

Keywords: Concensus control and estimation, Cooperative autonomous systems, Agents networks
Abstract: In this paper we investigate the edge disconnection problem for a discrete-time multi-agent consensus network. We first show that if the communication network is symmetric, an edge disconnection that does not affect the connectedness of the whole network does not even affect the final consensus value. On the contrary, if the communication network is not symmetric, then, in general, an edge disconnection that preserves the strong connectedness of the communication graph nonetheless may lead to a consensus value that is different from the original one. Discernibility of the faulty network from the original one is investigated both in case the states of all the agents are available and in case only the states of a subset of the agents are available. Several equivalent conditions are derived and it is proved that the necessary and sufficient conditions for discernibility and for discernibility from the observation of a subset of agents are exactly the same and can be checked on the original state matrix and on its eigenvectors.

Keywords: Cooperative control, Aerospace, Control over networks
Abstract: This paper addresses the trajectory analysis, mission design, and control law for multiple microsatellites to cooperatively circumnavigate a host spacecraft. This cooperative circumnavigation problem is defined to drive a group of networked microsatellites to a predefined planar ellipse concerning a host spacecraft while maintaining a geometric formation configuration. We first design several potential functions to guide the microsatellites to the given planar elliptical orbit with a proper radius. Next, the affine Laplacian matrix is introduced to characterize the desired formation shape of microsatellites. Based on the potential functions and the Laplacian matrix, a cooperative circumnavigation control law is finally proposed. Then, simulation results of 8 microsatellites with Earth-orbiting mission scenarios are given.

Keywords: Cooperative control, Distributed cooperative control over networks, Linear systems
Abstract: This paper deals with the implementation of min-max time consensus tracking on a multi-quadrotor testbed. A leader quadrotor generates a reference trajectory onto which the remaining quadrotors converge in min-max time using a local feedback control strategy. This control strategy is known to be globally optimal. Further, the effect of finite communication/measurement rate on consensus tracking is analyzed and bounds on the resulting deviations of trajectories are characterized. The theoretical claims made are verified through experiments.

Keywords: Coverage control, Cooperative control, Cooperative autonomous systems
Abstract: Mobile Aerial Agents (MAAs) using localization sensors suffer from their inherent localization uncertainty. These MAAs are equipped with Pan-Tilt-Zoom (PTZ) cameras and thru their 3D-altitude and camera parameters control survey the area underneath them. For a larger convex area, the MAAs transmit to their neighbors their PTZ and altitude parameters. Using a Voronoi-free area tessellation framework and a gradient scheme, a collaborative control framework is provided to maximize the covered area. Simulation studies are offered to investigate the effectiveness of the suggested scheme.

Keywords: Distributed cooperative control over networks, Concensus control and estimation, Cooperative autonomous systems
Abstract: This paper deals with the consensus problem in networks divided into subnetworks (also called clusters), where each node of the network represents an agent with linear dynamics. Each subnetwork is represented by a fixed and directed graph. Moreover, the agents in each cluster cannot communicate with agents from other clusters, except one single agent of each subnetwork, which is called a leader. These leaders interact at some instant times via fixed and strongly connected directed graph. This paper proposes a consensus protocol that takes into account the continuous-time communications among agents in the clusters and discrete information exchanges between clusters. Based on this proposed protocol, the collective network dynamics of the multi-agent systems is described in the term of hybrid systems. The characterization of the global consensus value of this kind of networks is analyzed. Secondly, we show that the problem of consensus design can be indirectly solved by considering the stability of an equivalent system. Then, a sufficient condition for the asymptotical stability of this equivalent system is proposed. Finally, an illustrative example is given to show the effectiveness of the proposed theoretical results.

Keywords: Fault diagnosis, Cooperative autonomous systems, Hybrid systems
Abstract: We consider the problem of attack detection, isolation and counterattack in a network of agents. This particular network is divided into several smaller networks, which are disconnected from each other. The states continuously evolve following a linear consensus protocol and approach local agreements specific to each sub-network/cluster. So the agents are capable of achieving global consensus, at a given instant one agent from each sub-network called leader updates its state. Our objective here, to be more specific, is to develop a bank of unknown input observers (UIOs) such that each agent can monitor its own output link and detect if a neighbor is being attacked. Moreover, these observers should also be able to estimate the attack value and respond with a counter-attack to keep the global consensus from diverging too far from the expected global consensus value. We provide LMI sufficient conditions for the design of the bank of observers. Simulations are presented to illustrate our results.

Keywords: Mechatronics, Modeling, Robotics
Abstract: In this research work, model and identification for an artificial finger actuated by Shape Memory Alloy (SMA) wires is presented. The high energy density and flexibility of these alloys permits the design of compact actuation solutions with possible applications in robotics fields, ranging from industrial to biomedical ones. On the other hands, the hysteretic and highly nonlinear response of SMAs complicates system design, modeling, and control. As a first step towards the design of accurate position and interaction control algorithms, in this paper we develop a control-oriented model for the SMA actuated finger prototype. The model, based on a coupling between the finger structure with the Müller-Achenbach-Seelecke model for SMA material, allows to describe the dynamic and hysteretic response of the device in a physics-based fashion. After discussing the model, several experiments are performed tries in order to validate the model for several operating conditions.

Keywords: Mechatronics, Optimization, Stochastic control
Abstract: There is great interest in minimizing the impact forces and bounces of reluctance actuators during commutations, in order to reduce acoustic noise and mechanical wear. In that regard, a model-free run-to-run control algorithm is presented to decrease the contact velocity, by exploiting the repetitive operations of these devices. The problem is mathematically formulated and the algorithm is expressed in pseudocode. As the main contribution of this paper, a search method is proposed for the run-to-run strategy based on Bayesian optimization. Adjustments are carried out for its application in run-to-run control, e.g. the removal of stored points and the definition of a new acquisition function. For validation, simulations are performed using a dynamic model of a commercial solenoid valve, and defining the input parametrization. The results show the improvement of the proposed method with respect to a state-of-the-art search.

Keywords: Mechatronics, Robotics, Linear systems
Abstract: In this paper we consider the tracking problem with simultaneous unknown harmonic disturbance rejection for a linear parametric uncertain stationary SISO systems of arbitrary relative degree. The proposed structure is an extension of authors' previous results, which are based on output controller "consecutive compensator". Closed-loop system stability is proven and an iterative disturbance frequency estimation algorithm with 'sliding window' update scheme is introduced. Parallel kinematics Ball-and-Plate on a movable platform robotic system is considered as a test-bed, while the non-prehensile ball manipulation along a predefined cyclic trajectory under bounded disturbances is a task to be solved.

Keywords: Mechatronics, Robotics, Modeling
Abstract: This paper presents detailed analysis, modeling, and controller design of the Skid Steering Mobile Robot (SSMR) Argo J5 with a custom-built landing platform, which also considers wheel-terrain interaction for control purposes. Terramechanics theory and the Argo J5 kinetics are combined to analyze the wheel-terrain interaction for different types of terrain. A PD controller is, then, designed to demonstrate tracking capabilities of the Argo J5. The main contribution of the modeling approach is adding a vertical load on each wheel to obtain the wheel’s entry angle with respect to the terrain. Shear displacement is derived from knowing the point’s position on each wheel in 3D space. The Argo J5 model and controller are tested in a MATLAB/Simulink environment by using the real Argo J5 parameter values. Obtained results illustrate Argo J5 velocity, wheel rolling resistance, wheel turning moment resistance, and shear stress on different terrains.

Keywords: Mechatronics, Servo control, Robust control
Abstract: Cranes are required to lift and carry loads swiftly to desired positions without causing excessive swaying motion of the load. These are conflicting requirements, which make feedback control of crane systems challenging. Furthermore, variations in rope length and load mass complicate controller design, since they significantly influence swaying dynamics. This paper considers automatic control of Quanser 3DOF tower crane system using composite nonlinear feedback (CNF) methodology. To be more specific, a CNF controller is designed for the jib trolley position of the crane using partial state measurements. The performance of the CNF controller is compared with Quanser’s built-in linear quadratic regulator (LQR) controller both in simulation and experimental setups. The results show that the CNF controller provides better load handling capability in terms of fast positioning of the jib trolley and damping of load swaying.

Keywords: Robotics, Autonomous systems, Autonomous robots
Abstract: Unmanned aerial manipulators (UAMs) provide a remarkable level of dexterity on performing complex aerial manipulation tasks thanks to their high number of degrees of freedom (DOFs). By taking advantage of this redundancy, in this paper, a multi-task online redundancy resolution strategy is proposed to accomplish multiple UAM tasks while satisfying additional constraints such as joint angle limit and obstacle avoidance. The proposed approach is computationally inexpensive to run online, allowing UAM to dynamically react to unforeseen events by modifying its configuration during the execution of the primary task. In addition, the proposed approach incorporates the maximization of the manipulability measure into the objective function of the redundancy resolution framework. Moreover, in order to ensure the motion smoothness and satisfy the robot physical constraints the task relaxation approach for aerial manipulation is presented. Extensive simulation studies in ROS-Gazebo environment are performed to validate the performance of the proposed approach.

Keywords: Lyapunov methods, Nonlinear system theory, Stability of nonlinear systems
Abstract: Recently, several publications have been published, where continuous and piecewise-affine Lyapunov functions are constructed for nonlinear systems by numerically computing their values on a grid and then interpolating these values over the simplices of a simplicial complex. The value of such a Lyapunov function is computed at a grid point by numerically solving an initial-value problem and then integrating a positive definite function of the solution on a given time-interval. In this paper we systematically investigate how different initial-value solution methods compare in this application. Further, we propose a method to compute the integrals that is superior to former approaches.

Keywords: Observers for nonlinear systems, H2/H-infinity methods, Stability of nonlinear systems
Abstract: We consider the problem of estimating the state and unknown input for a large class of nonlinear systems subject to unknown exogenous inputs. The exogenous inputs themselves are modeled as being generated by a nonlinear system subject to unknown inputs. The nonlinearities considered in this work are characterized by multiplier matrices that include many commonly encountered nonlinearities. We obtain a linear matrix inequality, that, if feasible, provides the gains for an observer which results in certified mathcal{L}_2 performance of the error dynamics associated with the observer. We also present conditions which guarantee that the mathcal{L}_2 norm of the error can be made arbitrarily small.

Keywords: Output feedback, Uncertain systems, Observers for nonlinear systems
Abstract: While prior results on prescribed-time stabilization (i.e., convergence within a fixed prescribed time interval irrespective of the initial state) consider a chain of integrators with uncertainties matched with the control input (i.e., normal form), we consider here a general class of nonlinear strict-feedback-like systems with nonlinear state-dependent uncertain functions allowed throughout the system dynamics. Furthermore, we address the output-feedback problem and show that a dynamic observer and controller can be designed based on our dual dynamic high gain scaling based design methodology along with a novel temporal transformation and form of the scaling dynamics with temporal forcing terms to achieve both state estimation and regulation in the prescribed time interval.

Keywords: Differential algebraic systems, Nonlinear system theory
Abstract: The stabilization problem for differential-algebraic systems with Lipschitz nonlinearities is addressed. The proposed stabilization technique is based on the interpretation of differential-algebraic systems as the feedback interconnection of a linear system and an algebraic system. In this framework the algebraic variable and the nonlinearities can be treated as external disturbances acting on the linear system. A direct consequence of this approach is that the control problem reduces to a classical disturbance attenuation problem with internal stability. The application of the proposed theory to linear differential-algebraic systems recovers classical results. A simple example validates the technique.

Keywords: Computational methods, Nonlinear system theory, Signal processing
Abstract: Runaway Electrons (REs) are one of the main concerns in Tokamak plasma and can cause plasma instabilities. One key target of the future reactors is the REs diagnostic and control. For these reasons the analysis of their major properties is fundamental. An automatic platform for the RE pitch angle calculation which exploits imperfections on the visible images of the REs synchrotron radiation is presented.

Keywords: Nonlinear system theory, Lyapunov methods, Stability of nonlinear systems
Abstract: This paper proposes a new perspective of stability analysis for the dynamics influenced by Wiener processes. The main claim is introducing the dynamics of stochastic Lyapunov functions following the behaviors of the state variables. The aim is achieved by reconsidering the target systems as rough differential equations, which are the system representations in rough path analysis. In the procedure, we also show the operation of the ``finalization'' has an important role for the analysis. Then, we confirm the validity of our perspective by establishing a stability notion for linear-quadratic-Gaussian controlled systems compatible with asymptotic stability for linear-quadratic controlled systems.

Keywords: Algebraic/geometric methods, Biological systems
Abstract: A SIR epidemic model is one of the most well-known mathematical models that helps to understand the dissemination of an infectious illness. It is a three-compartment model composed by individuals that are susceptible, infective and recovered with respect to the disease. In this work, the infection rate is estimated for a particular SIR epidemic model by using as the output measurement the incidence rate, which is a nonlinear function of the state variables. The aim is then to eliminate variables in the given system for which there are no measurements, such as the proportion of each type of individuals (susceptible, infective and recovered). The method applied here is based on differential elimination concepts from differential algebra, more precisely the Rosenfeld-Gröbner algorithm is employed. Once the input-output (IO) equation is determined, the derivatives of the signal are estimated by a homogeneous finite-time differentiator and a gradient descent method can be applied to solve the IO equation for the infection rate.

Keywords: Biological systems, Genetic regulatory systems, Modeling
Abstract: One of the most crippling problems in quantitative and synthetic biology is that models aiming to describe the real mechanisms of biochemical processes inside cells typically contain too many unknown parameters to be reliably inferable from available experimental data. Recent years, however, have seen immense progress in the development of experimental platforms that allow not only to measure biological systems more precisely but also to administer external control inputs to the cells. Optimal experimental design has been identified as a tool that can be used to decide how to best choose these control inputs so as to excite the systems in ways that are particularly useful for learning the biochemical rate constants from the corresponding data. Unfortunately, the experiment that is best to learn the parameters of a system depends on the precise values of these parameters, which are naturally unknown at the time at which experiments need to be designed. In this paper, we use a recently constructed genetic toggle switch as a case study to investigate how close to the best possible experiment we can hope to get with the most widely used optimal design approaches in the field. We find that, for strongly nonlinear systems such as the toggle switch, reliably predicting the information that can be gained from a priori fixed experiments can be difficult if the system parameters are not known very precisely. This suggests that a better strategy to guarantee informative experiments might be to use feedback control and to adjust the experimental plan in real time.

Keywords: Biomedical systems, Machine learning, Optimization
Abstract: We augment linear Support Vector Machine (SVM) classifiers by adding three important features: (i) we introduce a regularization constraint to induce a sparse classifier; (ii) we devise a method that partitions the positive class into clusters and selects a sparse SVM classifier for each cluster; and (iii) we develop a method to optimize the values of controllable variables in order to reduce the number of data points which are predicted to have an undesirable outcome, which, in our setting, coincides with being in the positive class. The latter feature leads to personalized prescriptions/recommendations. We apply our methods to the problem of predicting and preventing hospital readmissions within 30-days from discharge for patients that underwent a general surgical procedure. To that end, we leverage a large dataset containing over 2.28 million patients who had surgeries in the period 2011--2014 in the U.S. The dataset has been collected as part of the American College of Surgeons National Surgical Quality Improvement Program (NSQIP).

Keywords: Chemical process control, Modeling, Identification for control
Abstract: To remove carbon dioxide from biogas so as to produce biomethane, a fixed-bed tubular reactor filled with a zeolite pelleted solid sorbent is considered. To generate biomethane continuously, three batch reactors are operated in coordinated cycles. The control system operates in a two-level structure: a high-level coordination algorithm determines the shift from a process stage to another for each single reactor and computes the setpoints for the low-level controllers of each reactor; the low-level controllers regulate the process variables, such as the inlet gas flow rate, the inlet or outlet gas pressure and the sorbent temperature, according to the setpoints. In this paper, we first investigate the modelling of the batch process by means of mass, energy and momentum conservation equations. The concurrent adsorption of methane is also taken into account. The model parameters are identified by means of a two-stage procedure using experimental measurements from two plants, a laboratory-scale one located in Piacenza (Italy) and a pilot-scale one located in Camposampiero (Italy). With the identified model we design a control system for the coordination of three batch reactors.

Keywords: Nonlinear system identification, Biological systems, Optimization
Abstract: Hypertension (high blood pressure) is the most prominent cardiovascular disease affecting the majority of the population. One of the primary mediators of controlling blood pressure is the kidney, which is innervated by sympathetic nerves, through control of blood volume. It is therefore important to develop a dynamic model which explains the interaction between the sympathetic nerve activity (SNA) and renal blood flow (RBF). The present study proposes a simple approach based on binary particle swarms (BPSO) to model the complex interaction between sympathetic nerve activity (SNA) and renal blood flow (RBF) using polynomial nonlinear autoregressive with exogenous input (NARX) model. The effectiveness of BPSO is demonstrated by fitting models to different sets of RBF data which are collected from several rabbits under two conditions (vasoconstriction and vasodilation). Frequency domain analysis of the fitted model is carried out to investigate if there exist any similarities in the renal dynamics under the infusion of vasoconstrictors or vasodilators.

Keywords: Sensor and signal fusion, Medical signal processing, Biomedical systems
Abstract: This paper presents the development of a localization system in GPS-denied environments using an Inertial Measurement Unit (IMU) and a Pulse-Doppler radar. A ground speed estimation from radar measurements is first proposed. This estimation is combined with noisy measurements from an IMU in a Luenberger observer, allowing accurate dead-reckoning. The methodology proposed provides short-term position of the sensors embedded in a white cane, the ultimate goal being obstacle detection through the computation of a model of the surroundings. The results show that this solution gives an error growth rate of the position estimation of 0.026m/s, which is a hundred times better than the one obtained with the naive double integration of the accelerometer data.

Keywords: Nonlinear system identification, Feedback linearization, Machine learning
Abstract: This paper presents a method for modelling and controlling charge air pressure of a heavy-duty Diesel engine. The modelling method combines expert information about the system with a training dataset to provide a reasonable extrapolation behaviour as well as high accuracy. The dynamics of the gas path are modelled as a Wiener process with separate dynamics for each input. After identifying these dynamics, a controller based on input-output linearisation can be designed. The controller is shown to be able to track a set point of the gas path even after a sensor failure. To this end, the system behaviour is learned online to increase the accuracy. An evaluation against a reference model shows the performance of the method.

Keywords: Cooperative autonomous systems, Optimal control, Automotive
Abstract: We propose a cooperative adaptive cruise control which exploits a front vehicle acceleration forecast, available via vehicle-to-vehicle communication, to reduce vehicle energy consumption. The proposed control design can be applied to groups or emph{platoons} of cooperative vehicles, which locally optimize performance while maintaining their relative distance bounded at all times. The proposed approach adopts a learning-based model predictive control scheme, and maximizes performance while guaranteeing robust safety against the uncertainty on the front vehicle acceleration forecast. Simulations show the validity of the proposed control design.

Keywords: Delay systems, H2/H-infinity methods, Automotive
Abstract: A methodology is proposed to design stabilising and robust fixed-order decentralised controllers for heterogeneous vehicular platoons with Cooperative Adaptive Cruise Control (CACC). We consider Linear Time Invariant (LTI) models with constant time-delays at state, input and output. The closed-loop systems of (identical) local controllers and heterogeneous parameter vehicles are modelled by a system of delay differential algebraic equations. The proposed frequency domain approach uses the non-conservative direct optimisation approach towards stabilisation and robustness optimisation of delay systems. In this paper, the design problem of stabilising (identical) controllers achieving mathcal{L}_2 string stability for one vehicle look-ahead platoon is reduced to a simultaneous controller design problem for a parameterised (sub)system, where the allowable values of the parameters correspond to heterogeneity (including time-delays) of the vehicles. By treating the heterogeneity in parameters as perturbations contained in specific intervals or regions, we determine the values for pseudo-spectral abscissa and robust induced-mathcal{L}_2 norm. Hence, we ensure that the achieved exponential stability and string stability properties along with the overall computational complexity (of designing the controller) are independent of the number of vehicles. The application of CACC is simulated in MATLAB software.

Keywords: Iterative learning control, Optimization, Automotive
Abstract: In this work we present a hierarchical optimization scheme to minimize the overall operational costs of a heavy-duty diesel engine combined with a selective catalytic reduction (SCR) catalyst. The optimization variables of the hierarchical optimization scheme are setpoints of the underlying engine controller. The presented hierarchical optimization scheme follows a modular approach with the capability of constraint integration and inclusion of additional subsystems. Moreover, to reduce the computational load, a learning method is given to adapt the solution of the hierarchical optimization

Keywords: Modeling, Reduced order modeling, Automotive
Abstract: In this paper, the Power-Oriented Graphs (POG) technique is used to model Planetary Gear transmission systems. The full elastic dynamic model of the system is obtained using a fast and direct method which can be easily applied to any type of planetary gear. The rigid and reduced dynamic model of the system when the stiffness coefficients go to infinity is then obtained using a POG congruent state space transformation allowing the user to select which angular speeds are to be maintained in the reduced model. Another interesting aspect of the presented method is that the obtained reduced model is still able to provide the time behaviors of the tangential forces present between each couple of gears of the considered planetary gear system. The presented fast and direct method is then applied to two practical case studies, and simulative results in Matlab/Simulink showing the effectiveness of the method are finally reported and commented.

Keywords: Predictive control for nonlinear systems, Adaptive control, Automotive
Abstract: With the recent emergence of electric powertrains, a faster and easy to model actuator, the electric motor, became available for the control of longitudinal dynamics. Therefore model-based control approaches promise an increase in control performance, especially for processes such as traction control that require highly dynamic control intervention. The task of traction controllers is to to prevent the driven wheels from slipping and thus ensure the vehicle’s steerability. In this paper, a model predictive control approach to traction control is developed. A semi implicit method to discretize the underlying model was proposed to handle numerical stability problems at low speeds in real time. Due to changing environmental conditions, the functionality of the traction controller is limited and may lead to performance degradation or even failure. Therefore, a maximum friction coefficient estimation utilizing an unscentend Kalman filter is integrated. The overall control scheme is experimentally evaluated with a Volkswagen Golf GTE Plug-In Hybrid on a test track with a wet steel road surface.