Keywords: Robotics
Abstract: One of the central societal challenges is to prolong independent living for elderly and promote well. Personalized robotic rehabilitation and assistance is considered one of the enabling technologies with control design playing a significant role. Focusing on sensorimotor rehabilitation and assistance, personalized control should be able to adapt to the high inter-personal variability in human motor behavior but also to intra-personal changes over time. Control adaptation is further challenged by the sparsity of person-specific data because calibration routines need to be brief for user acceptance. Above all, guaranteed safety is one of the key requirements.

In this talk we will present recent results on learning-based control with performance and safety guarantees for highly uncertain systems with particular focus on challenges arising from personalized rehabilitation and assistance. In order to achieve high sample efficiency as well as transparency of the system, available knowledge of neuromuscular dynamic models will be exploited and and augmented by Bayesian non-parametric model components. Epistemic uncertainty due to limited training data will explicitly taken into account in the control design in order to achieve uncertainty-aware behavior of the closed loop system. Online learning as well as realtime capabilities are further important aspects. The results will be demonstrated in user intention-driven shared control designs for upper limb rehabilitation and assistance with exoskeletons and functional electrical stimulation. Furthermore, the limits of learning control and personalization will be discussed.

Keywords: Automotive, Nonlinear system theory, Agents and autonomous systems
Abstract: This contribution introduces a centralized input constrained optimal control framework based on multiple control barrier functions (CBFs) to coordinate connected and automated agents at intersections. For collision avoidance, we propose a novel CBF which is safe by construction. The given control scheme provides provable guarantees that collision avoidance CBFs and CBFs to constrain the agents' velocity are jointly feasible (referred to as control-sharing property) subject to input constraints. A simulation study finally provides evidence that the proposed control scheme is safe.

Keywords: Automotive, Autonomous systems, Predictive control for nonlinear systems
Abstract: An approach for automated driving in highway scenarios in the context of a two levels hierarchical architecture is proposed. In particular, we define suitable artificial potential functions (APF) combinations that can effectively handle the most relevant maneuvers of highway driving, such as speed and distance tracking, lane keeping, overtaking and returning. Parameters of the APF functions are dynamically tuned according to the acquired scenario. The defined APF are included in the cost function of a Model Predictive Control (MPC) control problem to generate the path trajectory. A behavioral logic described by a finite state machine (FSM), based on sensor acquired data and suitable dynamic conditions is defined to select the most appropriate maneuver to realize. Extensive simulation tests are introduced to show the effectiveness of the proposed approach.

Keywords: Automotive, Decentralized control, Predictive control for linear systems
Abstract: This contribution presents a decentralized control strategy for the automation of roundabouts, in which autonomous vehicles are required to circulate without collisions and to safely interact with pedestrians. The inherent hierarchical controller incorporates a consensus-based auction algorithm, which allows vehicles to agree on a crossing order at each entry point of the roundabout. An on-board scenario-based model predictive controller avoids the occurrence of collisions with other vehicles and pedestrians, while optimizing performance metrics over time. We run stochastic simulations that confirm theoretical findings.

Keywords: Traffic control, Predictive control for nonlinear systems, Transportation systems
Abstract: In this paper a platoon-actuated mainstream traffic control is proposed to decongest bottlenecks due to recurrent and nonrecurrent events. In this scheme the control actions to be communicated to the platoons, i.e., speed and configuration, are defined by means of a predictive control law based on traffic and platoon state detected in an area identified immediately upstream of the bottleneck. The main peculiarity of this scheme is that the size of the controlled area is dynamically adjusted based on the predicted congestion at the bottleneck. This approach keeps the control law computation burden low, while not sacrificing the control performance. Simulation results reported in the paper show the effectiveness of the proposed scheme, eliminating from 60% to 80% of the delay incurred from congestion compared with the uncontrolled case, depending on the level of traffic.

Keywords: Cooperative autonomous systems, Stochastic control, Hybrid systems
Abstract: Cooperative driving, enabled by Vehicle-to-Everything (V2X) communication, is expected to significantly contribute to the transportation system's safety and efficiency. Cooperative Adaptive Cruise Control (CACC), a major cooperative driving application, has been the subject of many studies in recent years. The primary motivation behind using CACC is to reduce traffic congestion and improve traffic flow, traffic throughput, and highway capacity. Since the information flow between cooperative vehicles can significantly affect the dynamics of a platoon, the design and performance of control components are tightly dependent on the communication component performance. In addition, the choice of Information Flow Topology (IFT) can affect certain platoon's properties such as stability and scalability. Although cooperative vehicles’ perception can be expanded to multiple predecessors’ information by using V2X communication, the communication technologies still suffer from random loss. Therefore, cooperative vehicles are required to predict each other's behavior to compensate for the effects of non-ideal communication. The notion of Model-Based Communication (MBC) was proposed to enhance cooperative vehicle's perception under non-ideal communication by introducing a new flexible content structure for broadcasting joint vehicle's dynamic/driver's behavior models. By utilizing a non-parametric (Bayesian) modeling scheme, i.e., Gaussian Process Regression (GPR), and the MBC concept, this paper develops a discrete hybrid stochastic model predictive control approach and examines the impact of communication losses and different information flow topologies on the performance and safety of the platoon. The results demonstrate an improvement in response time and safety using more vehicles' information, validating the potential of cooperation to attenuate disturbances and improve traffic flow and safety.

Keywords: Cooperative control, Adaptive control, Agents and autonomous systems
Abstract: We consider the problem of interaction-aware motion planning for automated vehicles in general traffic situations. We model the interaction between the controlled vehicle and surrounding road users using a generalized potential game, in which each road user is assumed to minimize a common cost function subject to shared (collision avoidance) constraints. We propose a quadratic penalty method to deal with the shared constraints and solve the resulting optimal control problem online using an Augmented Lagrangian method based on PANOC. Secondly, we present a simple methodology for learning preferences and constraints of other road users online, based on observed behavior. Through extensive simulations in a highway merging scenario, we demonstrate the practical efficacy of the overall approach as well as the benefits of the proposed online learning scheme.

Keywords: UAV's, Lyapunov methods, Hybrid systems
Abstract: We characterize an input-affine model of the UAV DarkO: a convertible drone designed and developed at the Ecole Nationale de L'Aviation Civile (ENAC) in Toulouse (France). Starting from a nonlinear model available in the literature, we present an approximate input-affine nonlinear model, whose dynamics simplifies the control design task. For this simplified model, we characterize the hovering equilibria in the absence of wind, and we derive the corresponding linearized dynamics. Then present a hysteresis-based switching mechanism combining a nonlinear feedback (providing a large basin of attraction) with a linearized feedback (providing improved performance but a smaller basin of attraction). Simulation results, using the original nonlinear model, confirm the effectiveness of the proposed feedback design.

Keywords: Optimization, Agents and autonomous systems, Distributed cooperative control over networks
Abstract: The task assignment problem is fundamental in combinatorial optimisation, aiming at allocating one or more tasks to agents in a system while minimizing the total cost or maximizing the overall assignment benefit. This problem is known to be computational hard since it is usually formulated as a mixed-integer programming problem. In this paper, we consider a novel time-triggered dimension reduction algorithm (TTDRA). We propose convexification approaches to convexify both the constraints and the cost function for the general non-convex assignment problem. The computational speed is accelerated via our time-triggered dimension reduction scheme, where the triggered condition is designed based on the optimality tolerance and the convexity of the cost function. Optimality and computational efficiency are verified via numerical simulations on benchmark examples.

Keywords: Optimization, Machine learning, Computational methods
Abstract: A design centering approach to robust optimization is presented that determines the largest tolerance box within the feasible region which is defined by performance functions. The associated semi-infinite optimization problem is solved with learning-based optimization algorithms. In every iteration of the tolerance maximization problem, the feasibility of the tolerance box is evaluated based on Gaussian process (GP) models of the performance functions. The GP models are adaptively learned during the optimization and a constrained acquisition optimization problem, using so-called probability threshold (PTR) measures, is solved to find promising boxes. This approach, called the PTR approach, is compared to a straight-forward baseline method, called bilevel approach, where separate GP models for the feasibility of a box and for the performance functions are maintained. Numerical evaluations of both approaches are presented and they are compared in terms of sample efficiency and computational effort.

Keywords: Optimization, Optimization algorithms, Computational methods
Abstract: The branch-and-bound (B&B) method is a commonly used technique in mixed-integer programming, whose performance is known to improve substantially if early termination methods can be applied to accelerate the pruning of branches. We propose an early termination technique that estimates a lower bound of the objective of current node problem and can stop the computation early instead of solving it to optimality. We show that our proposed technique can be generalized to ADMM-based mixed integer conic programming and speed up convergence in practice.

Keywords: Optimization, Optimization algorithms, LMI's/BMI's/SOS's
Abstract: We propose a bilinear decomposition for the Burer-Monteiro method for semidefinite programming and combine it with an Alternating Direction Method of Multipliers algorithm. Bilinear decomposition reduces the degree of the augmented Lagrangian from four to two, which makes each of the subproblems a quadratic programming and hence computationally efficient. Our approach is able to solve a class of large-scale SDPs with diagonal constraints. We prove that our ADMM algorithm converges globally to the set of first-order stationary points, and show empirically that the algorithm returns a globally optimal solution for diagonally constrained SDPs.

Keywords: Optimization, Optimization algorithms, Machine learning
Abstract: We study the decentralized consensus and stochastic optimization problems under compressed message sharing over a fixed directed graph. Our goal is to minimize the sum of n functions, each accessible to a single node, where local communications are subject to a directed network. In this work, we propose an iterative push-sum algorithm with compressed message sharing to reduce the communication overhead on the network. Contrary to existing literature, we allow for arbitrary compression ratios in the communicated messages. We provide explicit convergence rates for the stochastic optimization problem on smooth functions that are either (i) strongly convex, (ii) convex, or (iii) non-convex. Finally, we provide numerical experiments to illustrate the arbitrary compression and communication efficiency of our algorithm.

Keywords: Optimization, Optimization algorithms
Abstract: In the gradient descent method, one often focus on the convergence of the sequence generated by the algorithm, but less often on the deviation of these points from the solutions of the original continuous-time differential equation (gradient flow). This also happens when discretizing other ordinary differential equations. In the case of a discretization by explicit Euler’s method with a constant step h, we provide here sufficient conditions, in terms of strong monotonicity and co-coercivity, for the deviation between discrete and continuous solutions to tend asymptotically towards zero. This analysis could shed new light on some applications of the gradient descent algorithm.

Keywords: Optimization algorithms, Agents networks, Stability of nonlinear systems
Abstract: This paper proposes anytime feasible networked dynamics to solve resource allocation problems over time-varying multi-agent networks. The state of agents represents the assigned resources while their total (equal to demand) is constant. The idea is to optimally allocate the resources among the group of agents by minimizing the overall cost subject to fixed sum of resources. Each agent’s information is local and restricted to its own state, cost function, and the ones from its immediate neighbors. This work provides a fast convergent solution (compared to linear dynamics) while considering more-relaxed uniform network connectivity and (logarithmic) quantized communications among agents. The proposed dynamics reaches optimal solution over switching (sparsely-connected) undirected networks as far as their union over some bounded non-overlapping time-intervals has a spanning tree. Moreover, we prove anytime-feasibility of the solution, uniqueness, and convergence to the optimal value irrespective of the specific nonlinearity in the proposed dynamics. Such general proof analysis applies to many similar 1st-order allocation dynamics subject to strongly sign-preserving nonlinearities, e.g., actuator saturation in generator coordination. Further, anytime feasibility (despite the nonlinearities) ensures that our solution satisfies the fixed-sum resources constraint at all times.

Keywords: Observers for linear systems, Distributed control, Adaptive control
Abstract: In this paper, the output synchronization (OS) problem of heterogeneous linear multi-agent systems (MASs) with input delays is addressed. Agents may have different state dimensions and different dynamics. A finite-time observer (FO) is firstly proposed to estimate the uncertain leader's system dynamics. Then, based on the above FO, an adaptive observer (AO) is designed to estimate leader's state information. Thirdly, a novel state predictor is proposed to tackle the input delay effect based on the above AO and output regulation theory. After that, a third observer is designed to estimate the above state predictor so that the controller can be implemented in reality. The stability analysis is performed via Lyapunov stability theory with sufficient conditions derived in terms of an algebraic Riccati equation. The main achievement of this work is the construction of an observer-based fully distributed controller (FDC) which relies on local information only and does not require knowledge of the leader's dynamics or global graph information. As a result, such an approach can be implemented to large-scale systems. Finally, the effectiveness of the proposed FDC is verified via simulations and the influence of the system's graph structure on the convergence rate of the FDC is discussed.

Keywords: Observers for linear systems, Linear systems, Filtering
Abstract: The privacy aspect of state estimation algorithms has been drawing high research attention due to the necessity for a trustworthy private environment in cyber-physical systems. These systems usually engage cloud-computing platforms to aggregate essential information from spatially distributed nodes and produce desired estimates. The exchange of sensitive data among semi-honest parties raises privacy concerns, especially when there are coalitions between parties. We propose two privacy-preserving protocols using Kalman filters and partially homomorphic encryption of the measurements and estimates while exposing the covariances and other model parameters. We prove that the proposed protocols achieve satisfying computational privacy guarantees against various coalitions based on formal cryptographic definitions of indistinguishability. We evaluate the proposed protocols to demonstrate their efficiency using data from a real testbed.

Keywords: Observers for nonlinear systems, Uncertain systems
Abstract: Robust observer design is developed for a normal form system with a prescribed upper bound on the observer convergence time irrespective of initial state values and matched uniformly bounded disturbances. The design recasts a finite-time observer, especially developed for a system composed of diagonal and over-diagonal forms, and it is based on the scaling technique. The proposed observer operates with time-varying gains, uniformly bounded on an infinite horizon, thereby yielding an attractive implementation opportunity compared to the original prescribed-time observer design by Holloway and Krstic [1] with time-varying gains, which escape to infinity as time goes to the prescribed time instant. Capabilities of the present observer are supported in a numerical study, performed for a perturbed triple integrator, and are compared to those of the existing prescribed-time observer.

Keywords: Observers for nonlinear systems, Model/Controller reduction
Abstract: In this work, we consider the state estimation problem for a class of non-autonomous Persidskii systems. This paper presents conditions on the existence and stability of a nonlinear observer based on the invariant manifold approach. The conditions are formulated using Linear Matrix Equalities (LME) and Inequalities (LMI). Two interesting applications of the result are presented: a reduced-order observer (e.g., an observer for unmeasured states) and regression, both in linear and nonlinear settings. An example to demonstrate the efficiency of results is provided.

Keywords: Linear systems, H2/H-infinity methods, Observers for linear systems
Abstract: This paper investigates the problem of static output feedback H∞ control for singular continuous-time fractional-order systems with the fractional-order derivative α belonging to 1 ≤ α < 2. All required conditions in this paper are characterized in the form of strict linear matrix inequalities whose feasible solutions can be obtained easily and directly. Finally, some numerical examples are provided to illustrate the effectiveness of the obtained theoretical results in the paper.

Keywords: Observers for nonlinear systems
Abstract: In this paper, a range based localization method is proposed for position estimation in three dimensions with only three measurements. The proposed scheme is based on a modification of the system to make it globally observable and a change of coordinates which allows to construct a continuous time observer. The obtained continuous time observer is further modified to be able to deal with aperiodic and asynchronous measurements. The performances of the proposed approach are illustrated with simulations. It is shown that the proposed observer performs well even in presence of noise.

Keywords: Observers for nonlinear systems, Stability of nonlinear systems, Lyapunov methods
Abstract: This paper proposes an analytical expression for a nonlinear mapping between steady-state solutions of certain types of nonlinear interconnected systems. This mapping is found using tools from the theory of output regulation for systems presented in lower-triangular or upper-triangular canonical forms. Next, this mapping helps design an excitation input and a corresponding reduced-order observer for interlaced systems, a combination of both upper- and lower-triangular subsystems. A proposed global observer is proved to be robust to additive disturbance and measurement noise by applying the Lyapunov function method. An example involving a massspring system demonstrates the efficiency of our approach.

Keywords: Autonomous systems, Maritime
Abstract: Automating docking operations of ships requires at least two robust and precise localization systems for reaching the typically required safety and redundancy levels. Global navigation satellite systems are typically chosen as one of these, while the second needs to be independent and preferably use another measurement principle. Optical sensors are versatile, low-cost, and assumed to provide sufficient localization range. Used to simultaneously locate the vessel and map the harbor environment, this technology is believed to offer the necessary properties to complement satellite systems in a robust manner. In this paper, a monocular camera together with the state-of-the-art ORB-SLAM3 algorithm is used for localization. Umeyama’s method is used to create an initialization procedure to determine the unknown scale factor encountered in monocular camera odometry and to find the transformation between the camera and a world-fixed coordinate frame. The proposed system is validated using data recorded on a commercial high-speed passenger ferry in nominal operation. The results indicate a localization range of more than 200 m. The mean absolute position error is less than 0.5 m with an estimated heading error of 0.5° in favorable weather conditions.

Keywords: Autonomous systems, Optimization algorithms, Nonlinear system identification
Abstract: This paper proposes an optimization-based approach to predict trajectories of autonomous race cars. We assume that the observed trajectory is the result of an optimization problem that trades off path progress against acceleration and jerk smoothness, and which is restricted by constraints. The algorithm predicts a trajectory by solving a parameterized nonlinear program (NLP) which contains path progress and smoothness in cost terms. By observing the actual motion of a vehicle, the parameters of prediction are updated by means of solving an inverse optimal control problem that contains the parameters of the predicting NLP as optimization variables. The algorithm therefore learns to predict the observed vehicle trajectory in a least-squares relation to measurement data and to the presumed structure of the predicting NLP. This work contributes with an algorithm that allows for accurate and interpretable predictions with sparse data. The algorithm is implemented on embedded hardware in an autonomous real-world race car that is competing in the challenge Roborace and analyzed with respect to recorded data.

Keywords: Autonomous systems, Transportation systems
Abstract: This paper investigates an integrated traffic environment modeling and model predictive control (MPC) system to realize interaction-aware dynamic motion planning of an autonomous vehicle with multiple surrounding vehicles. The interaction-aware interacting multiple model Kalman filter (IAIMM-KF) from the literature is used to hierarchically predict maneuvers and trajectories of surrounding vehicles and to compute safe targets for the ego vehicle. The targets are terminal speed and reference lane, which are moving targets as they are updated at each time step. Then, an MPC controller is designed for the ego vehicle to generate an optimal trajectory by following the moving targets and including the prediction results to formulate collision-free constraints. The proposed interaction-aware planning method has a proactive planning ability and can avoid collisions by non-local replanning. The strengths and effectiveness of the approach are verified in challenging highway lane-change simulation scenarios.

Keywords: Autonomous systems, V&V of control algorithms, Optimal control
Abstract: We propose control barrier functions (CBFs) for a family of dynamical systems to satisfy a broad fragment of Signal Temporal Logic (STL) specifications, which may include subtasks with nested temporal operators or conflicting requirements (e.g., achieving multiple subtasks within the same time interval). The proposed CBFs take into account the actuation limits of the dynamical system as well as a feasible sequence of subtasks, and they define time-varying feasible sets of states the system must always stay inside. We show some theoretical results on the correctness of the proposed method. We illustrate the benefits of the proposed CBFs and compare its performance with the existing methods via simulations.

Keywords: Autonomous systems, Agents and autonomous systems, Agents networks
Abstract: In recent years, the number of autonomous vehi- cles used in manufacturing systems and warehouses is increas- ing and distributed control algorithms are necessary to manage their paths and velocities. This paper considers a consensus problem in a leaderless network of vehicles (agents) that have to reach a common velocity while forming a uniformly spaced string. The stability and convergence speed of such second order consensus problem is studied by the authors in a recent work. In this paper, we address an issue related to its practical implementation such as the communication time delays among the agents. Necessary and sufficient conditions are derived to guarantee consensus in the presence of a constant delay and a modified control law is proposed to reach the consensus even if the delay makes unstable the system. Finally, a simulation campaign is conducted to verify the convergence properties of the proposed consensus protocol.

Keywords: Autonomous systems, Hybrid systems, V&V of control algorithms
Abstract: In this paper, we consider the verification of approximate infinite-step opacity for discrete-time control systems. Relying on finite abstraction techniques for solving this problem requires discretization of the state and input sets, which requires significant computational resources. Here, we propose a discretization-free approach in which we formulate opacity as a safety property over an appropriately constructed augmented system, and seek to verify it by finding suitable barrier certificates. Within our proposed scheme, lack of opacity is also verified by posing it as a reachability property over the augmented system. The main result of this paper offers a discretization-free approach to verify (lack of) infinite-step opacity in discrete-time control systems. We also discuss other notions of opacity, and their relations to one another. We particularly study the conditions under which verifying one form of opacity for a system also implies other forms. Finally, we illustrate our theoretical results on two numerical examples, where we utilize sum-of-squares programming to search for polynomial barrier certificates. In these examples, we verify the infinite-step, and current-step opacity for a vehicle by checking whether its position is concealed from possible outside intruders.

Keywords: Autonomous systems, Linear parameter-varying systems, Automotive
Abstract: This paper considers a path tracking control problem for an autonomous vehicle under the constraint to maintain a safe distance from the vehicle preceding in the same lane. The control design problem is solved on the basis of a mathematical model that jointly describes the lateral and longitudinal dynamics. Specifically, an H_{infty} optimal controller is designed aimed at minimizing the orientation and position tracking errors with respect to a given reference path. Moreover, the controller is able to ensure that the vehicle can both follow a speed profile and automatically adjusts the vehicle speed to maintain a proper distance from the preceding vehicle. The lateral dynamics is regulated by a 2-DOF feed-forward/feed-back controller, where the feedforward part is implemented by exploiting an inverse kinematic model. On the other side, the feedback action, along with the action of the controller in charge to regulate the longitudinal dynamics, are gain-scheduling controllers synthesized through the linear parameter varying H_{infty} control design theory. In particular, a parameter-depending quadratic Lyapunov function is used to reduce the conservativeness of the solution and a suitable convex LMI optimization problem is formulated for the synthesis. The vehicle is modeled as a full car model with seven degrees of freedom with the tires described by the classical Pacejka's Magic Formula. In order to show the performance of the proposed control strategy, simulations have been performed by considering an autonomous vehicle driving into an urban scenario and by taking advantage of the co-simulation environment developed in Matlab/Simulink.

Keywords: Adaptive control, Adaptive systems, Uncertain systems
Abstract: This paper proposes a novel multi-model adaptive identification (MMAI) algorithm for discrete-time linear time invariant (LTI) uncertain systems with tunable performance. The uncertain plant parameter vector is assumed to belong to a known convex hull of a finite number of vertices; these vertices are considered as initial choice of vertices of an adaptive model parameter set. The estimated parameter, corresponding to the uncertain plant parameter, is computed at every instant as a convex combination of the model set vertices. To update the vertices of the adaptive model parameter set, a switched adaptive update law is proposed, along with a novel discrete-time initial excitation (IE) condition, which is imposed on the regressor signal. The proposed discrete-time IE condition is online verifiable and is milder than persistence of excitation (PE) condition, required for parameter convergence in classical adaptive estimation routines. The switched adaptive law guarantees exponential convergence of the vertices of the model set as well as the estimated parameter, to the true plant parameter, provided the regressor signal satisfies the IE condition. The properties of the designed MMAI strategy are validated through suitable simulation examples.

Keywords: Adaptive control, Aerospace, Lyapunov methods
Abstract: In a scenario of always more complex and demanding space missions, enhanced attitude control systems play a key role for satellite design and capabilities improvement. In this paper, an adaptive and robust solution to the pointing angle tracking problem under mixed sources of disturbances is presented. A valid simplified parametric model, of a single axis satellite dynamics, is firstly introduced and discussed, to the objective of effectively accounting for propellant slosh and flexible appendices torque perturbations. The control problem is subsequently solved with a model reference adaptive control approach, namely the Minimal Control Synthesis (MCS). Theoretical fundamentals, architecture implementation and final tuning are reported. A subset of the extensive validation campaign performed is presented and discussed to demonstrate the impressive robustness and performance level reached by the proposed scheme. Limitations and future perspective to tackle them conclude the paper.

Keywords: Adaptive control, Nonlinear system theory, Hybrid systems
Abstract: Traffic congestion has dire economic and social impacts in modern metropolitan areas. To address this problem, in this paper we introduce novel types of model-free transactive controllers designed to manage vehicle traffic in highway networks for which precise mathematical models of their dynamics are absent. Specifically, we consider a highway system with managed lanes on which dynamic tolling mechanisms can be implemented in real-time using measurements from the roads. We present three incentive-seeking feedback controllers capable of finding in real-time optimal economic incentives that persuade highway users to follow a suitable driving behavior that minimizes a predefined performance index. The controllers are agnostic to the exact model of the highway, and they are also able to guarantee fast convergence to the optimal toll by leveraging non-smooth and hybrid mechanisms combining continuous-time dynamics with discrete-time dynamics. We provide numerical examples to illustrate the advantages of the different presented techniques.

Keywords: Adaptive systems, Autonomous systems, Optimization
Abstract: Devising optimal operating strategies for a compressor station relies on the knowledge of compressor characteristics. As the compressor characteristics change with time and use, it is necessary to provide accurate models of the characteristics that can be used in optimization of the operating strategy. This paper proposes a new algorithm for online learning of the characteristics of the compressors using Gaussian Processes. The performance of the new approximation is shown in a case study with three compressors. The case study shows that Gaussian Processes accurately capture the characteristics of compressors even if no knowledge about the characteristics is initially available. The results show that the flexible nature of Gaussian Processes allows them to adapt to the data online making them amenable for use in real-time optimization problems.

Keywords: Adaptive systems, Observers for linear systems, Uncertain systems
Abstract: Design of fast and efficient algorithms for adaptive observers, i.e., joint estimation of parameters and internal states of a dynamical system is vital for various applications. In this paper, a recently proposed initial excitation-based adaptive observer is modified to use a matrix regressor, instead of a vector regressor, for fast on-line estimation. A novel concept of strong initial excitation condition (S-IE) is formulated, leading to a switched estimation scheme, which is shown to be milder than already existing excitation conditions like persistence of excitation (PE), strong persistent excitation (S-PE) and initial excitation (IE). The proposed algorithm enables exponential convergence by a user-assignable convergence rate. Strategic introduction of estimator for the unknown initial conditions of the state variables in conjunction with the parameter estimator ensures uniformly global exponential stability (UGES) of the estimation error dynamics. The proposed adaptive observer scheme is validated using a linear plant dynamics along with an electro-chemical model of Lithium-ion (Li-ion) battery for the estimation of state of charge (SoC) and state of health (SoH).

Keywords: Intelligent systems, Mechatronics, Adaptive control
Abstract: An artificial muscle is well known to have strong asymmetric hysteresis characteristics, which depend on the load applied to the muscle. It is therefore difficult to achieve a high control performance and robustness to various loads. In a previous study, model-free adaptive control (MFAC), which is a data-driven control method, was applied to muscles, and a high tracking control performance was achieved. However, MFAC requires numerous design parameters that are extremely time-consuming to tune. To solve these problems, this study considers the tuning of the design parameters of the MFAC by introducing virtual reference feedback tuning, which is a data-driven control method. In addition, control experiments with five loads were conducted to verify the robustness of the load. The experimental results show that the proposed method achieves a high tracking control performance without an overshoot and is highly robust to loads while reducing the time-consuming routine for parameter tuning and modelling.

Keywords: Output feedback, Constrained control, Adaptive control
Abstract: The dilemma between transient behavior and accuracy in tracking control arises in both theoretical research and engineering practice and funnel control has showed great potential in solving that problem. Apart from the controlled system, the performance of funnel control strongly depends on the reference signal and the choice of prescribed funnel boundary. In this paper, we will present a new form of funnel controller for systems with control saturation. Compared to former research, the new controller is more reliable, and the closed-loop system can even achieve asymptotic tracking. Besides that, a new concept called constrained funnel boundary is introduced. Together with the new controller and the constrained funnel boundary, the application range of funnel control is extended significantly.

Keywords: Genetic regulatory systems, Biomolecular systems, Biological systems
Abstract: Recent work on engineering synthetic cellular circuitry has shown that non-regulatory interactions brought about through competition for shared gene expression resources, such as ribosomes, can result in degraded performance (where circuit behaviour deviates from design specifications) or even failure (qualitatively different functionality). Numerous feedback control strategies have been proposed to decouple co-expressed genes in simple genetic circuits; ranging from feedback within the circuit, resource allocation schemes and growth-based feedback. In this work, we utilise a whole cell mathematical model, which captures key gene expression trade-offs, to compare these control strategies for their ability to ameliorate the impact of resource limitations, maintain growth and assess their robustness to host uncertainty and environmental variation.

Keywords: Biological systems, Optimal control, Modeling
Abstract: Since early 2020, the world has been dealing with a raging pandemic outbreak: COVID-19. A year later, vaccines have become accessible, but in limited quantities, so that governments needed to devise a strategy to decide which part of the population to prioritize when assigning the available doses, and how to manage the interval between doses for multi-dose vaccines. In this paper, we present an optimization framework to address the dynamic double-dose vaccine allocation problem whereby the available vaccine doses must be administered to different age-groups to minimize specific societal objectives. In particular, we first identify an age-dependent Susceptible-Exposed-Infected-Recovered (SEIR) epidemic model including an extension capturing partially and fully vaccinated people, whereby we account for age-dependent immunity and infectiousness levels together with disease severity. Second, we leverage our model to frame the dynamic age-dependent vaccine allocation problem for different societal objectives, such as the minimization of infections or fatalities, and solve it with nonlinear programming techniques. Finally, we carry out a numerical case study with real-world data from The Netherlands. Our results show how different societal objectives can significantly alter the optimal vaccine allocation strategy. For instance, we find that minimizing the overall number of infections results in delaying second doses, whilst to minimize fatalities it is important to fully vaccinate the elderly first.

Keywords: Automata, Optimal control, Stochastic control
Abstract: Aggressive cancers are typically incurable because of drug resistance development. We model cancer growth and adaptive genetic response in a cell-based (CB) setting, displaying how optimal maximum tolerated dose administration protocols of two drugs counteract drug resistance and turn it into an exploitable weakness. Our CB model is a spatial extension of the population-based model proposed by Orlando et al., where a homogeneous population of cancer cells evolves according to a fitness landscape. To make a first feasibly test of the optimal drug administration control problem in the CB framework, we add only the elements we consider most relevant for describing cancer growth and evolution: phenotypic heterogeneity, spatial competition, and drugs diffusion, as well as realistic administration protocols. We calibrate our model on Orlando et al.'s one and find that dynamical protocols switching between the two drugs minimize the cancer size at the end of (or at mid-points during) treatment. These results differ from those of Orlando and colleagues, which suggest static protocols under generalizing and neutral allocation trade-offs.

Keywords: Biomedical systems, Hybrid systems, Identification for hybrid systems
Abstract: A hybrid model for biomedical time series comprising a continuous second-order linear time-invariant system driven by an input sequence of positively weighted Dirac delta functions is considered. The problem of the joint estimation of the input sequence and the continuous system parameters from output measurements is investigated. A solution that builds upon and refines a previously published least-squares formulation is proposed. Based on a thorough analysis of the properties of the least-squares solution, improvements in terms of accuracy and ease of use are achieved on synthetic data, compared to the original algorithm.

Keywords: Biomedical systems, Observers for nonlinear systems
Abstract: Cancer is one of the major causes of death all over the word. Research dedicated to understanding and combating cancer continues to attract the interest of scientists. In this paper, we investigate a system of nonlinear ordinary differential equations describing a tumor growth model accounting for angiogenic stimulation and inhibition. This model has a high Lipschitz constant which renders the implementation of the standard high-gain observer very complicated. To overcome this problem, we propose in this paper to use a HG/LMI-based observer strategy to estimate the tumor volume. By combining the high-gain methodology with linear matrix inequalities based technique, such an observer allows reducing significantly the high-gain tuning parameter which improves highly the performances with respect to measurement noise, and avoids the peaking phenomenon. These two advantages, which are the weaknesses of the standard high-gain observer, play an important role in very sensitive applications like tumor growth control. The numerical simulations show the superiority of the HG/LMI-based observer compared to other observers in the literature.

Keywords: Biological systems, Markov processes, Model/Controller reduction
Abstract: Epigenetic cell memory (ECM), the inheritance of gene expression patterns without changes in genetic sequence, is a critical property of multi-cellular organisms. Chromatin state, as dictated by histone covalent modifications, has recently appeared as a mediator of ECM. In this paper, we conduct a stochastic analysis of the histone modification circuit that controls chromatin state to determine key biological parameters that affect ECM. Specifically, we derive a one-dimensional Markov chain model of the circuit and analytically evaluate both the stationary probability distribution of chromatin state and the mean time to switch between active and repressed chromatin states. We then validate our analytical findings using stochastic simulations of the original higher dimensional circuit reaction model. Our analysis shows that as the speed of basal decay of histone modifications decreases compared to the speed of autocatalysis, the stationary probability distribution becomes bimodal and increasingly concentrated about the active and repressed chromatin states. Accordingly, the switching time between active and repressed chromatin states becomes larger. These results indicate that time scale separation among key constituent processes of the histone modification circuit controls ECM.

Keywords: Biological systems, Neural networks, Modeling
Abstract: We present an original model of lateral inhibition based on a network of heterogeneous Hodgkin-Huxley neurons. We consider two types of neurons: excitatory Regular Spiking with Adaptation neurons and inhibitory Fast Spiking neurons and two types of synapses: chemical and electrical. We design networks of lateral inhibition based on a previously introduced accurate and scalable feedback structure. In addition, we demonstrate the effect of lateral inhibition when the network receives input current of different intensities that conceptually corresponds to different levels of brightness received by neurons belonging to the visual system. Finally, we prove the accuracy of the lateral inhibition model by simulating a very common visual illusion.

Keywords: Identification, Biological systems
Abstract: We consider in this paper a general SEIRS model describing the dynamics of an infectious disease including latency, waning immunity and infection-induced mortality. We derive an infinite system of differential equations that provides an image of the same infection process, but counting also the reinfections. Existence and uniqueness of the corresponding Cauchy problem is established in a suitable space of sequence valued functions, and the asymptotic behavior of the solutions is characterized, according to the value of the basic reproduction number. This allows to determine several mean numbers of reinfections related to the population at endemic equilibrium. We then show how using jointly measurement of the number of infected individuals and of the number of primo-infected provides observability and identifiability to a simple SIS model for which none of these two measures is sufficient to ensure on its own the same properties.

Keywords: Medical signal processing, Modeling, Identification
Abstract: Measuring the blood glucose (BG) concentrations for people with diabetes is essential to achieve a better glycemic control either by medical professionals or by using feedback control algorithms. Continuous Glucose Monitoring (CGM) devices provide indirect measurements of the BG each 1--5 minutes. However, CGM devices suffer from correlated measurement errors and calibration errors. Detailed models for the errors of CGM devices already exist in the literature. Nonetheless, the identification of these models requires data from multiple CGM devices at once and accurate reference blood glucose measurements obtained clinically. This fact makes these models difficult to be subject-specific during typical treatment since diabetic subjects only use one CGM device with 3--4 finger pricking blood glucose measurements per day. In this paper, a methodology to obtain subject-specific CGM error models using Temporal Gaussian Processes (TGP) in their state space form is introduced. Three different TGPs are proposed and a strategy based on a particle Markov Chain Monte Carlo (MCMC) is used to perform regression and fit parameters for the models. The strategy is tested against data generated from virtual subject using detailed CGM error measurement models which were fitted with more than one CGM device and detailed clinical data from the literature. The results demonstrated the ability for TGP models with the proposed particle MCMC strategy to obtain subject-specific CGM error models using data available during the typical life of diabetic subjects.

Keywords: Identification for control, Linear systems, Mechatronics
Abstract: Dual Youla-Kucera plant model parametrization is very useful for describing model uncertainties. Therefore it is interesting to develop recursive identification algorithms for identification of these type of plant model structures in closed loop operation for potential use in iterative tuning or adaptive control. Closed loop output errors type recursive algorithms are developed specifically for this type of model structure. The algorithms assure global asymptotic stability in the deterministic environment and unbiased parameter estimation in the presence of noise when the plant model is in the model set. The algorithms will be applied for the identification in closed loop of a test bench for active noise control.

Keywords: Identification, Modeling, Biomedical systems
Abstract: This paper examines the problem of modeling the dynamics of the filling, drainage, and pressurization of the peritoneal cavity of a laboratory animal during perfusion. The paper is motivated by the potential of the peritoneal perfusion of an oxygenated perfluorocarbon (PFC) to provide a pathway for gas exchange in patients suffering from respiratory failure. Modeling cavity mechanics is important for avoiding excessive intracavity pressures that could potentially cause abdominal compartment syndrome during perfusion. Previous research in the literature examines elastic cavity behavior, but the problem of experimentally identifying models that couple this behavior with suction-assisted discharge remains relatively unexplored. Towards this goal, we performed large animal (namely, swine) experiments where we measured variables including peritoneal intracavity pressure, suction pressure, and PFC inflow. A simple state-space model fits data from the above experiment well. This model helps elucidate important preliminary insights into: (i) the role of active suction in facilitating discharge, (ii) the stiffening of the peritoneal cavity with perfusion, (iii) the linearity of cavity discharge behavior, (iv) the potential need to examine the impact of paralytics on cavity pressure dynamics as future work.

Keywords: Identification, Nonlinear system identification, Identification for control
Abstract: The problem of parameter estimation for salient permanent magnet synchronous motors (PMSMs) is considered in this paper. It is assumed that stator currents and voltages and rotor position are measurable signals and the only known motor parameter is the number of pole pairs. The proposed estimation algorithm uses dynamic operators applied to the measurable signals to get a linear regression model with unknown motor parameters. As proven, scalar estimators that are designed for each unknown motor parameter ensure finite time convergence of estimation errors to zero. The main benefit of the proposed approach is to relax the stringent assumptions imposed on regression functions such as condition of persistent excitation (PE) and non-square integrability that are often difficult to verify in practice.

Keywords: Identification for control, Fault estimation, Linear systems
Abstract: The problems of identifying the parameters (frequency, damping factor, amplitude and phase) of exponentially damped sinusoidal signals with constant parameters are considered. The signal is represented as the output of a linear generator, the parameters of the sinusoidal signal (amplitude, phase, damping factor and frequency) are unknown. The main idea is to apply the Jordan waveform and delay to parameterize the signal and obtain a linear regression model. The performance of the algorithms considered in the article is illustrated by computer modeling.

Keywords: Identification, Linear systems
Abstract: When recorded signals are corrupted by noise on both input and output sides, standard identification methods give biased parameter estimates, due to the presence of input noise. This paper discusses in what situations such a bias is large and, consequently, when the errors-in-variables identification methods should preferably be used.

Keywords: Distributed parameter systems, Delay systems, Output feedback
Abstract: This paper studies the boundary output feedback stabilization of reaction-diffusion PDEs in the presence of an arbitrarily long distributed input delay. The boundary control applies at the right boundary through a Robin boundary condition while the system output is selected as the left boundary Dirichlet trace. The actual control input applies to the boundary via a distributed delay spanning over a finite time interval. The proposed control strategy leverages a predictor feedback relying on Artstein's reduction method and is coupled with a finite-dimensional observer. Provided a structural controllability assumption, sufficient stability condition are derived and are shown to be always feasible provided the order of the observer is selected to be large enough.

Keywords: Biomolecular systems, Modeling, Distributed parameter systems
Abstract: Recently, molecular communication (MC) has been studied as a micro-scale communication between cells or molecular robots. In previous works, the MC channels in unbounded environment was analyzed. However, many of the experimentally implemented MC channels are surrounded by walls, thus the boundary condition should be explicitly considered to analyze the dynamics of MC channels. In this paper, we propose a framework to analyze the frequency response of one-dimensional MC channels based on a diffusion equation with a boundary. In particular, we decompose the MC channel into the diffusion system and the boundary system, and show the relation between the cut-off frequency of the MC channel and the communication distance based on the transfer function. We then analyze the frequency response of a specific MC channel and reveal that the boundary can restrict the communication bandwidth of the MC channel.

Keywords: Reduced order modeling, Distributed parameter systems, Process control
Abstract: Advection-Diffusion-Reaction (ADR) Partial Differential Equations (PDEs) appear in a wide spectrum of applications such as chemical reactors, concentration flows, and biological systems. A large number of these applications require the solution of ADR equations involving time-varying coefficients, where analytical solutions are usually intractable. Numerical solutions on the other hand require fine discretization and are computationally very demanding. Consequently, the models are normally not suitable for real-time monitoring and control purposes. In this contribution, a reduced order modeling method for a general ADR system with time-varying coefficients is proposed. Optimality of the reduced order model regarding the reduction induced error is achieved by using an H2-norm reduction method. The efficacy of the method is demonstrated using two test cases. Namely, a case for an ADR with arbitrary dynamics varying coefficients and a second case including the modeling of an exemplary water quality distribution path with randomly generated demand. The reduced order models are evaluated against high fidelity simulations using MATLAB's finite element method PDE toolbox. It is shown that the reduction can achieve a significant computational speedup allowing for the usage of the model for real-time applications with sampling times in milliseconds range. Moreover, the constructed ROM is shown to achieve high prediction accuracy with the normalized mean square error below 2.3 % for a real-world water quality simulation test case.

Keywords: Distributed parameter systems, Output regulation, Lyapunov methods
Abstract: This paper is about the stabilization of a cascade system composed by an infinite-dimensional system, that we suppose to be exponentially stable, and an ordinary differential equation (ODE), that we suppose to be marginally stable. The system is controlled through the infinite-dimensional system. Such a structure is particularly useful when applying the internal model approach on infinite-dimensional systems. Our strategy relies on the forwarding method, which uses a Lyapunov functional and a Sylvester equation to build a feedback-law. Under some classical assumptions in the output regulation theory, we prove that the closed-loop system is globally exponentially stable.

Keywords: Algebraic/geometric methods, Autonomous systems, Hybrid systems
Abstract: This paper discusses Routh reduction for simple hybrid forced mechanical systems. We give general conditions on whether it is possible to perform symmetry reduction for a simple hybrid Lagrangian system subject to non-conservative external forces, emphasizing the case of case of cyclic coordinates. We illustrate the applicability of the symmetry reduction procedure with an example and numerical simulations.

Keywords: Delay systems, Linear systems
Abstract: Halanay's inequality is a powerful and widely used tool to prove asymptotic convergence of functions that arise in the study of systems with delays or continuous-discrete features. In its standard form, it applies to scalar valued functions that satisfy decay conditions, with overshoots depending on suprema of the functions over suitable intervals. Then it provides exponential decay estimates on the scalar functions. Here, we provide vector versions of Halanay's inequality, and of the so-called trajectory based approach, both yielding input-to-state stability (or ISS) inequalities. Our proofs of the inequalities use the theory of positive systems. We apply our results to prove ISS for interval observers.

Keywords: Delay systems, Uncertain systems
Abstract: We propose a new output event-triggered control design for linear discrete-time systems with constant arbitrarily long input delays, using delay compensating subpredictors. We prove input-to-state stability of the closed loop system, using framers and the theory of positive systems. A novel feature of our approach is our use of matrices of absolute values, instead of Euclidean norms, in our discrete-time event triggers for our delay compensating control design. We illustrate our approach using a model of the BlueROV2 marine vehicle, where our new event triggers lead to a smaller number of control recomputation times as compared with standard event triggers that were based on Euclidean norms, without sacrificing on settling times or on other performance metrics.

Keywords: Control over networks, Sampled data control, Stability of hybrid systems 
Abstract: In the first part of this session we will start by motivating the need for event-triggered control techniques. We will then introduce the basic techniques in the literature, and present the main features and subtleties of the analysis and design problems in event-triggered control.

Keywords: Control over networks, Sampled data control, Stability of hybrid systems
Abstract: See Part 1.

Keywords: Control over networks, Sampled data control, Stability of hybrid systems
Abstract: See Part 1.

Keywords: Control over networks, Sampled data control, Stability of hybrid systems
Abstract: We will explain how event-triggered controlled systems can be modeled as hybrid systems, for which the required background will be recalled. Then, we will show how a hybrid small-gain theorem can be used to analyze various seemingly disparate techniques of the literature in a unified way. We will then further demonstrate the strength of this ``small-gain'' viewpoint by systematically modifying existing techniques, thereby potentially reducing the number of transmissions as illustrated on a numerical example.

Keywords: Control over networks, Sampled data control, Stability of hybrid systems
Abstract: See Part 1.

Keywords: Control over networks, Sampled data control, Stability of hybrid systems
Abstract: We will discuss several important aspects of event-triggered control, including characterizations of the amount of transmissions generated by event-triggered controllers, as well as conditions under which event-triggered control can outperform the traditional periodic sampling paradigm. We will also briefly overview results available for other control problems than the stabilisation of the origin. We will finally present some experimental validations of event-triggered controllers.

Keywords: Control over networks, Sampled data control, Stability of hybrid systems
Abstract: See Part 1.

Keywords: Control over networks, Sampled data control, Stability of hybrid systems
Abstract: See Part 1.

Keywords: Fault detection and identification, Fault tolerant systems
Abstract: The latest studies confirm that considerable changes in freshwater resources have been occurring across the globe, indicating a future in which already limited water resources will become even more precious. According to the World Economic Forum, water crises is one of the top global risks in terms of impact. On the other hand, the continuous expansion of urban footprint means that an estimated 70% of the world’s population will live in urban areas by 2050. The dramatic increase in water demands resulting from this unprecedented urbanization, together with increasingly uncertain climate conditions, indicate the need for a holistic, intelligent decision-making framework for managing water infrastructures in the cities of the future. Consequently, there is a need for a new approach for designing the next generation of urban drinking water systems that is applicable not only to planning and management of mature water infrastructure systems such as those found in developed countries, but also to developing countries where the fastest population growth is predicted over the next 50 years. From a system engineering perspective, urban drinking water networks are complex, large-scale systems designed to supply clean water to industrial and domestic users. Some of the key water challenges include water losses, ensuring water quality, energy efficiency, and safety and security of water resources. Recent advances in information and communication technologies have facilitated the modernization of water systems with the installation of sensors, actuators, data processing units and wireless communications, which enables the collection of real-time data related to water systems. The objective of this presentation is to provide an overview of current advances in smart water systems from a systems and control perspective. Several results on monitoring, control and fault tolerance of water distribution networks will be presented, and directions for future research will be discussed.

Keywords: Traffic control, Emerging control applications, Predictive control for nonlinear systems
Abstract: In this paper we propose a model for mixed traffic composed of few Connected and Automated Vehicles (CAVs) in the bulk flow. We rely on a multi-scale approach, coupling a Partial Differential Equation describing the overall traffic flow and Ordinary Differential Equations accounting for CAV trajectories, which act as moving bottlenecks on the surrounding flux. In our framework, CAVs are allowed to overtake (if on different lanes) or merge (if on the same lane). Controlling CAV desired speeds allows to act on the system to minimize any traffic density dependent cost function. More precisely, we apply Model Predictive Control to reduce fuel consumption in congested situations. In particular, we observe how the CAV number impacts the result, showing that low penetration rates are sufficient to significantly improve the selected performance indexes. The outcome of this paper supports the attractive perspective of exploiting a small number of controlled vehicles as endogenous actuators to regulate traffic flow on road networks.

Keywords: Traffic control, Identification for control, Transportation systems
Abstract: Connected and Automated Vehicles (CAVs) are likely to have a large impact on the traffic in the near future. Assuming we are able to communicate some commands directly to them, it is of interest to know how CAVs can be used for traffic control. In order to achieve this, we need to understand how such controls affect the rest of the traffic. In this work, we study the influence of a CAV acting as a moving bottleneck, using the CAV's speed as a control input. We discuss the interpretation of the microscopic traffic data in the macroscopic framework, and propose nonparametric methods for learning the micro-macro model describing the interaction between the CAV and the surrounding traffic. We use only the local traffic data in the vicinity of the CAV, and design simple, targeted data collection experiments. This learned model is then used to predict the evolution of the traffic, and the predictions are compared with corresponding data from microscopic simulations.

Keywords: Transportation systems, Traffic control, Cooperative control
Abstract: Signalized intersections in arterial roads result in persistent vehicle idling and excess accelerations, contributing to fuel consumption and CO2 emissions. There has thus been a line of work studying eco-driving control strategies to reduce fuel consumption and emission levels at intersections. However, methods to devise effective control strategies across a variety of traffic settings remain elusive. In this paper, we propose a reinforcement learning (RL) approach to learn effective eco-driving control strategies. We analyze the potential impact of a learned strategy on fuel consumption, CO2 emission, and travel time and compare with naturalistic driving and model-based baselines. We further demonstrate the generalizability of the learned policies under mixed traffic scenarios. Simulation results indicate that scenarios with 100% penetration of connected autonomous vehicles (CAV) may yield as high as 18% reduction in fuel consumption and 25% reduction in CO2 emission levels while even improving travel speed by 20%. Furthermore, results indicate that even 25% CAV penetration can bring at least 50% of the total fuel and emission reduction benefits.

Keywords: Transportation systems, Game theoretical methods
Abstract: Road networks are used inefficiently when drivers selfishly choose their own routes rather than being directed on routes to minimize the social cost, which is the total travel delay experienced by all users. In the setting of mixed autonomy, when there are multiple types of vehicles which affect congestion differently, this inefficiency can increase greatly beyond the setting of only a single vehicle type. In the single vehicle type setting, well-known optimal tolls completely eliminate this inefficiency. However, the difficult qualities of mixed autonomy makes it such that optimal tolls are unknown in this setting, absent very restrictive assumptions. In this work we provide the first optimal tolls for the mixed autonomous setting in general networks with multiple source-destination pairs, where road latencies are affine functions of vehicle flow. These optimal tolls are differentiated, meaning that the different vehicle types must be given different toll values. Administering differentiated tolls may be difficult -- to address this, we also provide anonymous tolls which may reduce inefficiency in this setting. We quantify the limits of such tolls and compare the performance of our tolls to these bounds.

Keywords: Machine learning, Transportation systems, Traffic control
Abstract: This paper introduces an equity emphasized recovery planning method for urban traffic networks based on a data driven approach. An integrated evaluation index is proposed to assess equity in territorial accessibility during hazards recovery, which brings the variance in accessibility between communities as a penalty term into the overall accessibility. Taking the improvement of the integrated index as the reward function, the equity emphasized recovery control strategy is designed with a reinforcement learning algorithm to determine the recovery priority of the affected links. To test the performance of the proposed approach, a simulation environment with reference to the San Francisco Bay Area was constructed. Experiment results indicate that, compared with the explicit strategies, the proposed recovery strategy is able maintain a more equitable approach during the reconstruction process.

Keywords: Traffic control, Transportation systems, Linear parameter-varying systems
Abstract: In this paper, we study a mixed traffic system moving along a single-lane open-road. This platoon includes a number of human-driven vehicles (HDVs) together with one connected and automated vehicle (CAV). The dynamics of HDVs are assumed to follow the optimal velocity model (OVM), and the acceleration of the single CAV is directly controlled by a static output-feedback controller. Due to different traffic conditions, the desired velocity of the platoon can change over time. Moreover, there are multiple system parameters that are uncertain. The ultimate goal of this work is to present a gain-scheduled robust control strategy that, with a varying desired speed, stabilizes the traffic flow in the presence of undesired disturbances and parametric uncertainties. In this direction, a gain-scheduled H_{infty} static output-feedback controller is designed, and its efficiency is illustrated through numerical simulations.

Keywords: Constrained control, Agents and autonomous systems, Traffic control
Abstract: Ensuring safety of automated vehicle (AV) con- trol systems in multi-lane mixed-autonomy traffic is challeng- ing. Control barrier functions (CBFs) represent a promising approach in which control inputs are filtered to guarantee forward-invariance of satisfaction of desired safety properties. This allows for balancing safety with performance, such as in the context of data-driven adaptive cruise control systems, which may otherwise be difficult to assure safety for. In real-world deployments to mixed-autonomy multi-lane traffic, however, ex- ternal disturbances such as cut-in events can generate violations to the satisfaction of safety properties which would otherwise be met, such as maintaining a minimum time gap between vehicles. This work extends the design of CBFs for AVs by explicitly considering the effect of cut-in events. We show that a commonly proposed CBF designed to maintain time-gap cannot guarantee collision avoidance in the event of a cut-in. We show that when paired with a secondary CBF designed to maintain a positive space-gap through the use of higher-order CBFs via simple switching logic that both collisions can be avoided when cut-ins occur, and that over time the desired time-gap will be restored. Additionally, we present criteria for pole placement and string-stability of the AV when choosing CBF parameters. A series of numerical experiments are presented to demonstrate the main results.

Keywords: Optimization algorithms, Optimization, Predictive control for nonlinear systems
Abstract: This paper presents alpaqa, an open-source C++ implementation of an augmented Lagrangian method for nonconvex constrained numerical optimization, using the first-order PANOC algorithm as inner solver. The implementation is packaged as an easy-to-use library that can be used in C++ and Python. Furthermore, two improvements to the PANOC algorithm are proposed and their effectiveness is demonstrated in NMPC applications and on the CUTEst benchmarks for numerical optimization.

Keywords: Optimization, Predictive control for nonlinear systems, Machine learning
Abstract: Real-time optimization (RTO) is a model-based technique to drive process operation towards its optimal condition according to an objective function while respecting constraints. However, whenever the used model presents structural uncertainty, the classic two-step approach fails to find the true optimum condition. Recently, a Modifier Adaption (MA) that uses Gaussian Process (GP) and trust-region concepts has been proposed in order to overcome plant-model mismatch without the need to estimate plant gradients. Based on these premises, this paper extends the methodology by proposing an Output MA (MAy) based on GP. The benefit of using output modifiers is its versatility to be applied in other model-based techniques, since it corrects the model instead of correcting a specific optimization problem. Therefore, we also propose a Hammerstein Nonlinear Model Predictive Control (NMPC) that incorporates the MAy-GP correction term into the predictive controller. The results showed that the performance of MAy-GP is similar to the MA-GP, also showing the ability to drive the plant towards its optimum even in the presence of structural uncertainty. In addition, the proposed NMPC with MAy-GP correction terms was successful, due to the correction in the static term of the Hammerstein model.

Keywords: Optimization, Optimization algorithms, Machine learning
Abstract: We study the acceleration of the Local Polynomial Interpolation-based Gradient Descent method (LPI-GD) recently proposed for the approximate solution of empirical risk minimization problems (ERM). We focus on loss functions that are strongly convex and smooth with condition number sigma. We additionally assume the loss function is eta-Hölder continuous with respect to the data. The oracle complexity of LPI-GD is tilde{O}left(sigma m^d log(1/varepsilon)right) for a desired accuracy varepsilon, where d is the dimension of the parameter space, and m is the cardinality of an approximation grid. The factor m^d can be shown to scale as O((1/varepsilon)^{d/2eta}). LPI-GD has been shown to have better oracle complexity than gradient descent (GD) and stochastic gradient descent (SGD) for certain parameter regimes. We propose two accelerated methods for the ERM problem based on LPI-GD and show an oracle complexity of tilde{O}left(sqrt{sigma} m^d log(1/varepsilon)right). Moreover, we provide the first empirical study on local polynomial interpolation-based gradient methods and corroborate that LPI-GD has better performance than GD and SGD in some scenarios, and the proposed methods achieve acceleration.

Keywords: Optimization algorithms, Optimization
Abstract: This paper describes the design of a safeguarding scheme for Anderson acceleration to improve its practical performance and stability when used for first-order optimisation methods. We show how the combination of a non-expansiveness condition, conditioning constraints, and memory restarts integrate well with solver algorithms that can be represented as fixed point operators with dynamically varying parameters. The performance of the scheme is demonstrated on seven different QP and SDP problem types, including more than 500 problems. The safeguarded Anderson acceleration scheme proposed in this paper is implemented in the open-source ADMM-based conic solver COSMO.

Keywords: Optimization algorithms, Optimization, Linear time-varying systems
Abstract: In this manuscript, we propose a dynamic weighted average consensus algorithm as the solution to the optimal resource allocation problem with time-varying cost functions and resources. The objective is to minimize a global cost which is the summation of local quadratic time-varying cost functions by allocating time-varying resources. A reformulation of the original problem is developed and is solved distributively only using local interactions over an undirected connected graph. Local state trajectories converge to a neighborhood of the optimal trajectory which is bounded. This bound is computed with respect to cost parameters and topology properties. We also show that the trajectories are feasible at all times, meaning that the resource allocation equality constraint is met over the entire execution. As in most smart grid systems where agents may shut down or stop responding, i.e., cannot generate any load, our algorithm takes this into account and tracks the real-time optimal trajectory. To show the performance of the algorithm, simulations of a resource allocation problem are illustrated in the numerical example section.

Keywords: Optimization algorithms, Constrained control, Predictive control for linear systems
Abstract: This paper introduces the quadratically-constrained quadratic programming (QCQP) framework recently added in HPIPM alongside the original quadratic-programming (QP) framework. The aim of the new framework is unchanged, namely providing the building blocks to efficiently and reliably solve (more general classes of) optimal control problems (OCP). The newly introduced QCQP framework provides full features parity with the original QP framework: three types of QCQPs (dense, optimal control and tree-structured optimal control QCQPs) and interior point method (IPM) solvers as well as (partial) condensing and other pre-processing routines. Leveraging the modular structure of HPIPM, the new QCQP framework builds on the QP building blocks and similarly provides fast and reliable IPM solvers.

Keywords: Optimization algorithms, Modeling, Statistical learning
Abstract: Metamodeling has become a common approach to replace costly and time-consuming physical experiments or computer experiments (e.g., numerical simulation, training of AI models) by an easy-to-evaluate metamodel, which is trained on samples of the experiment. Since model accuracy depends significantly on the choice of sample points, these are in many cases determined using adaptive sampling methods. The cost of conducting an experiment often depends decisively on the choice of its parameters. However, only few strategies for selecting the sample points have been proposed, that take into account parameter-dependent costs. In this work, we introduce a novel Voronoi-based cost-aware adaptive sampling algorithm for global metamodeling that is independent of the choice of sampling strategy and metamodel. The method is evaluated on a variety of randomly generated black-box and cost functions, where it has shown to vastly outperform existing sampling strategies.

Keywords: Distributed estimation over sensor nets, Linear systems, Fault detection and identification
Abstract: In this paper, we study stateless and stateful physics-based anomaly detection scenarios via distributed estimation over sensor networks. In the stateful case, the detector keeps track of the sensor residuals (i.e., the difference of estimated and true outputs) and reports an alarm if certain statistics of the recorded residuals deviate over a predefined threshold, e.g., χ2 (Chi-square) detector. Instead, only instantaneous deviation of the residuals raises the alarm in the stateless case without considering the history of the sensor outputs and estimation data. Given (approximate) false-alarm rate for both cases, we propose a probabilistic threshold design based on the noise statistics. We show by simulation that increasing the window length in the stateful case may not necessarily reduce the false-alarm rate. On the other hand, it adds unwanted delay to raise the alarm. The distributed aspect of the proposed detection algorithm enables local isolation of the faulty sensors with possible recovery solutions by adding redundant observationally-equivalent sensors. We, then, offer a mechanism to design Q-redundant distributed observers, robust to failure (or removal) of up to Q sensors over the network.

Keywords: Fault tolerant systems, System reconfiguration, Process control
Abstract: In this work a fault tolerant control (FTC) based on virtual actuators is applied to an industrial drum boiler model. The objective was to implement a simple model based active FTC to a linear model of the drum boiler. For the case analyzed, the virtual actuator methodology is not able to make the faulty system behave identically to the nominal one. However, an approximate solution using genetic algorithms is shown. The algorithm is tested by simulating actuator faults and a robust system response is obtained.

Keywords: Fault tolerant systems, Decentralized control, Agents and autonomous systems
Abstract: The problem of fault-tolerant formation control for a team of wheeled robots is addressed. The multi-agent network is represented as the passive interconnection of port-Hamiltonian systems, and trajectory-tracking is achieved by using passivity arguments. Two fault-tolerant control strategies have been proposed, depending on the fault severity: a soft one, consisting in lowering the control burden, and a hard one, corresponding to formation reconfiguration

Keywords: Model/Controller reduction, Fault detection and identification, Optimization algorithms
Abstract: This paper presents a novel optimization-based method for multi-model affine abstraction (i.e., for simultaneous model reduction of multiple models), which solves for the existence of affine abstractions of a pair of different nonlinear systems with guarantees of model discrimination with the minimum detection time T under worst-case uncertainties and approximation errors. Our approach combines mesh-based affine abstraction methods with T-distinguishability analysis in the literature into a bilevel bilinear optimization problem. Then, to obtain a tractable solution, we leverage robust optimization techniques and a suitable change of variables to obtain a sufficient linear program (LP). Finally, the efficacy of proposed methods is illustrated by several numerical examples.

Keywords: Fault tolerant systems, Linear systems, Emerging control theory
Abstract: Actuator malfunctions may have disastrous consequences for systems not designed to mitigate them. We focus on the loss of control authority over actuators, where some actuators are uncontrolled but remain fully capable. To counteract the undesirable outputs of these malfunctioning actuators, we use real-time measurements and redundant actuators. In this setting, a system that can still reach its target is deemed resilient. To quantify the resilience of a system, we compare the shortest time for the undamaged system to reach the target with the worst-case shortest time for the malfunctioning system to reach the same target, i.e., when the malfunction makes that time the longest. Contrary to prior work on driftless linear systems, the absence of analytical expression for time-optimal controls of general linear systems prevents an exact calculation of quantitative resilience. Instead, relying on Lyapunov theory we derive analytical bounds on the nominal and malfunctioning reach times in order to bound quantitative resilience. We illustrate our work on a temperature control system.

Keywords: Manufacturing processes, Process control
Abstract: Over the last few years, with the increasing worldwide competition, semiconductor industries have had to constantly innovate in order to enhance their performance, productivity and minimize the downtime. Monitoring the state of health of their equipment units is important to avoid machine failures and to plan maintenance actions. For that, a novel approach for health indicator extraction named Significant Points combined to the Least Absolute Shrinkage and Selection Operator (SP-LASSO) is proposed in this paper. It deals with the problem of high dimensional data and the specificity of the health indicator in real industrial cases. The proposed method performs feature selection and health indicator extraction and it is mainly based on LASSO. A numerical application on simulated data illustrates the accuracy of this approach.

Keywords: Algebraic/geometric methods, Lyapunov methods, Autonomous systems
Abstract: An attitude tracking problem for a full-actuated rigid body in 3-D is studied using a model of rotation dynamics on the Lie group SO(3). A generalized homogeneous control is adopted to achieve tracking of a smooth attitude trajectory in a finite-time. The attitude dynamics on the Lie algebra mathfrak{so}(3) is utilized to design the homogeneous control, since SO(3) is not a vector space. A switch control is proposed for achieving global finite-time tracking by combining a global asymptotic control with local homogeneous control. The simulations illustrate the performance of the proposed control algorithm.

Keywords: Biological systems, Autonomous robots, Maritime
Abstract: This paper deals with the use of a passive hydrophone array (PHA) for following non-stationary underwater objects with a miniature biomimetic autonomous underwater vehicle. Current acoustic underwater localization systems are large and expensive. They are not well suited for using on miniature, low-cost underwater vehicles. However, compact size and low cost are necessary when developing autonomous vehicles that can be widely used by recreational and professional divers. As an alternative, a simple hydrophone array is developed for a biomimetic 4-flipper U-CAT AUV (Autonomous Underwater Vehicle). The array is used to measure the bearing of a single acoustic beacon. Two control approaches are proposed for navigating the vehicle towards the beacon. The whole system is validated both in a simulator and in a natural environment using two scenarios: 1) navigating towards a stationary beacon and 2) navigating towards a beacon mounted on a diver. The results show that the proposed control system is a viable approach to be used for diver following.

Keywords: Maritime, Agents and autonomous systems, Stability of nonlinear systems
Abstract: This paper proposes a novel method for formation path following of multiple underactuated autonomous underwater vehicles. The method combines line-of-sight guidance with null-space-based behavioral control, allowing the vehicles to follow curved paths while maintaining the desired formation. We investigate the dynamics of the path-following error using cascaded systems theory and show that the closed-loop system is uniformly semi-globally exponentially stable. We validate the theoretical results through numerical simulations.

Keywords: Adaptive control, Autonomous systems, Linear parameter-varying systems
Abstract: Autonomous driving is achieved by controlling the coupled nonlinear longitudinal and lateral vehicle dynamics. Longitudinal control greatly affects lateral dynamics and must preserve lateral stability conditions, while lateral controllers must take into account actuator limits and ride comfort. This work deals with the coordinated longitudinal and lateral control for autonomous driving. An improved particle swarm optimized PID (PSO-PID) is proposed to handle the task of speed tracking based on nonlinear longitudinal dynamics. An enhanced linear parameter varying model predictive controller (LPV-MPC) is also designed to control lateral dynamics, the latter is formulated with an adaptive LPV model in which the tire cornering stiffness coefficients are estimated by a recursive estimator. The proposed LPV-MPC is enhanced with an improved cost function to provide better performance and stability. Matlab/Carsim co-simulations are carried out to validate the proposed controllers.

Keywords: Agents and autonomous systems, Concensus control and estimation, Agents networks
Abstract: Connectivity maintenance control constitutes an issue of crucial importance for networked multi-agent systems with limited communication range. Towards this direction, we propose a decentralized algorithm to robustly estimate the global connectivity of the communication graph of a multi-agent system, which subsequently feeds a gradient-based controller that guarantees global connectivity maintenance. Finally, we verify the superiority of our algorithm against a well-established related method and validate the efficacy of the control scheme through simulated paradigms.

Keywords: Automotive, Transportation systems, Autonomous systems
Abstract: While internal combustion engine (ICE) vehicles are a major contributor to greenhouse gases, electric vehicles (EVs) have proven their potential to achieve eco-friendly transportation. To date, the usage of EVs is limited mainly due to their short driving range and relatively long charging times as well as their insufficient charging infrastructures, especially outside major cities or in less developed countries. Dynamic charging is also very limited due to high installation and maintenance cost. To address these issues, a novel moving stationary Vehicle to Vehicle (V2V) charging approach is proposed in this paper which has the potential to mitigate the range anxiety associated with EVs. To this end, we employ cooperative adaptive cruise control (CACC) to form platoons of EVs on freeways which allows EVs to wirelessly exchange their desired amount of electric charge. The proposed methodology is threefold; a novel platoon structure, a sophisticated collision-free longitudinal control which regulates a small gap between the supplier and recipient vehicles, and a new lane-changing framework that allows a proper catch-up phase as well as cut in/out maneuvers. The proposed approach is tested in a stretch of a relatively long freeway with on- and off-ramps in SUMO traffic simulator. The results demonstrate about 90% of successful power transfer among vehicles with feasible V2V capability. In addition, the employed approach demonstrates a promising speed and gap control for recipient and supplier vehicles which is indeed essential for an efficient power transfer.

Keywords: Autonomous robots, Robotics, Optimization
Abstract: This paper presents a hybrid robot motion planner that generates long-horizon motion plans for robot navigation in environments with obstacles. We propose a hybrid planner, RRT* with segmented trajectory optimization (RRT*-sOpt), which combines the merits of sampling-based planning, optimization-based planning, and trajectory splitting to quickly plan for a collision-free and dynamically-feasible path. When generating a plan, the RRT* layer quickly samples a semi-optimal path and sets it as an initial reference path. Then, the sOpt layer splits the reference path and performs optimization on each segment. It then splits the new trajectory again and repeats the process until the whole trajectory converges. We also propose to reduce the number of segments before convergence with the aim of further reducing computation time. Simulation results show that RRT*-sOpt benefits from the hybrid structure with trajectory splitting and performs robustly in various robot platforms and scenarios.

Keywords: Nonlinear system theory, Computational methods, Aerospace
Abstract: Applications in many engineering fields require nonlinear modelling and control. However, for nonlinear dynamical systems, designing a control law that makes the system's operating point globally asymptotically stable is challenging. In this paper, for systems whose dynamics can be modelled through polynomial functions, we provide a procedure to design a controller that guarantees global asymptotic stability and a certain cost. To achieve this goal, we use sum of squares programming. In particular, we further develop a previously proposed approach to provide such a control law also for cases where the previous approach failed. The new approach is less restrictive and requires less memory. We apply it to the van der Pol system, the Lorenz system, to a model of a hypersonic flight vehicle, and to satellite attitude control.

Keywords: Nonlinear system theory, Lyapunov methods, Computational methods
Abstract: In recent years data-driven analysis of dynamical systems has attracted a lot of attention and transfer operator techniques, namely, Perron-Frobenius and Koopman operators are being used almost ubiquitously. Since data is always obtained in discrete-time, in this paper, we propose a purely data-driven approach for the design of a stabilizing feedback control law for a general class of discrete-time control-affine non-linear systems. In particular, we use the Koopman operator to lift a control-affine system to a higher-dimensional space, where the control system's evolution is bilinear. We analyze the controllability of the lifted bilinear system and relate it to the controllability of the underlying non-linear system. We then leverage the concept of Control Lyapunov Function (CLF) to design a state feedback law that stabilizes the origin. Furthermore, we demonstrate the efficacy of the proposed method to stabilize the origin of the Van der Pol oscillator and the chaotic Henon map from the time-series data.

Keywords: Lyapunov methods, Nonlinear system theory, Optimization
Abstract: An improved method of control design for rational nonlinear discrete-time systems with input saturation is proposed in this paper. Using Difference-Algebraic Representation (DAR) and parameter-dependent Lyapunov functions, a novel regional stabilization condition in terms of Linear Matrix Inequalities (LMI) is presented. Two optimization problems are addressed to either obtain the largest estimated Domain of Attraction (DoA) or minimize the l2-gain from the energy-bounded disturbance input to the performance output. Numerical examples illustrate the potential of the proposed approach.

Keywords: Nonlinear system identification, Linear parameter-varying systems, Identification for control
Abstract: This paper presents a new model structure and structure selection procedure for the identification of control-oriented affine quasi-LPV (qLPV) models in state-space (SS) representation using Deep Recurrent Neural Networks (RNNs). The proposed model structure is intended to be an alternative to the existing black-box approaches for the case where no state measurements are available and the scheduling variables (and hence the dependence of the time-varying parameters on the scheduling variables) are unknown. Existing identification approaches are not able to incorporate deep neural network (DNN) structures but employ (normalized) radial basis functions (RBFs) to model the dependence of the time-varying parameters on the scheduling variables. This may increase the dimension of the time-varying parameter vector unnecessarily, making parameter estimation more difficult and limiting the model's use for LPV controller synthesis. The proposed identification approach aims to reduce the required dimension of the time-varying parameter vector by using a (Deep) Neural Network (NN) to model the parameter variation. In order to curb the complexity of the resulting nonlinear optimization problem and make the developed model approach useful in real-life applications, a structure selection procedure based on an initialization method developed in is proposed. The performance of the presented approach is demonstrated on a nonlinear system identification problem.

Keywords: Nonlinear system theory, Hybrid systems, Computational methods
Abstract: This paper addresses safety verification of nonlinear systems through invariant set computation. More precisely, our goal is verifying if the state of a given discrete time nonlinear system will keep evolving within a safe region, starting from a given set of initial conditions. To this purpose, we introduce a conformant PieceWise Affine (PWA) abstraction of the nonlinear system, which is instrumental to computing a conservative approximation of its maximal invariant set within the safe region. If the obtained set covers the set of initial conditions, safety is proven. Otherwise, subsequent refinements of the PWA abstraction are performed, either on the whole safe region or on some appropriate subset identified through a guided refinement approach and containing the set of initial conditions. Some numerical examples demonstrate the effectiveness of the approach.

Keywords: Genetic regulatory systems, Observers for nonlinear systems, Lyapunov methods
Abstract: In this paper, we investigate the problem of state estimation for a simple one-gene regulation dynamic process involving end-product activation to rebuild the non-measured concentrations of mRNA and the involved protein. We synthesize a new observer structure following the high-gain methodology by combining the observer proposed in~cite{CDCDyhia2020} based on the system state augmentation approach and the HG/LMI technique presented in~cite{Zemouche19}. The proposed design reduces significantly the value of the tuning parameter and the observer gain along with improving its sensitivity to disturbances and measurement noise. The results are compared with the standard high-gain observer to evaluate the effectiveness of the proposed design.

Keywords: Reduced order modeling, Nonlinear system theory, Model/Controller reduction
Abstract: In this work, we investigate the approximation and model reduction of bilinear dynamical systems having a particular dynamical structure, such as time-delayed terms, second-order, or fractional derivatives. By means of Volterra series expansion, such systems can be characterized using generalized multivariate transfer functions. Our main goal is to construct unstructured reduced-order bilinear models as reduced models by using evaluations of the generalized multivariate transfer functions associated with the original system. To this aim, we propose a new setup for the bilinear Loewner framework that enables Hermitian interpolation conditions. Moreover, inspired by the TF-IRKA method, we propose an iterative procedure that constructs accurate bilinear reduced models. A numerical test case illustrates the practical applicability of the method.

Keywords: Electrical machine control, Optimization, Nonlinear system theory
Abstract: A concept of nonlinear optimal control is introduced for IPMSM drives. The control configuration remains the same as for the conventional control with three PI controllers for the d and q axes currents and velocity but these controllers are replaced by corresponding nonlinear optimal controllers. The proposed controllers include linear (LQR) parts and nonlinear optimal parts emulating the automatic adjustment of the LQR gains based on operating conditions. The nonlinear parts are designed based on Krasovkiy’s optimality criterion leading to an explicit solution of the control design problem. The proposed controller possesses some robustness properties which is explored in simulations. Compared to conventional system, it allows to reduce velocity overshoot and torque oscillations without extending the transient times. control concept is validated using a test rig.

Keywords: Energy systems, Modeling, Optimal control
Abstract: Battery energy storage systems （BESS） have been widely used in grid services. This paper focuses on BESS's participations in grid price arbitrage with considerations of three battery life models that account for different factors for battery degradation. The BESS is optimized for maximizing the lifetime value using dynamic programming algorithms. Four cases are studied: case 1 – no battery degradation accounted; case 2 – an empirical battery degradation model using charge-discharge rate (Crate); case 3 – a semi-empirical model using Crate and the cell temperature; and case 4 – a new model that considers the influence of the battery's state of energy, Crate, and current state of health (SOH) on the battery aging rate. Taking the results of case 1 as a comparison benchmark, the battery SOH and power fluctuations of cases 2, 3, and 4 are much smoother, and the number of charge and discharge cycles during operation is reduced by 51.14%, 15.07%, and 66.23% respectively. In case 2, 3, and 4, after one year of continuous operation, the battery SOH decreases from 0.95 to 0.898, 0.89, and 0.9, respectively, leading to varied battery profits. These insights indicate the importance of selecting an appropriate battery degradation modelling in the controller design for BESS applications.

Keywords: Energy systems, Distributed control, Predictive control for linear systems
Abstract: This paper proposes a resilient distributed energy management algorithm able to cope with different types of faults in a DC microgrid system. A distributed optimization method allows to solve the energy management problem without sharing any private data with the network and reducing the computational cost for each agent, with respect to the centralised case. A distributed MPC scheme based on distributed optimization is used to cope with uncertainty that characterizes the microgrid operation. In order to be resilient to faults that limit the amount of power available to consumers, we propose to adaptively store an amount of power in the storage systems to support the loads. A soft constraint on the minimum energy stored in each battery is introduced for feasibility and to cope with persistent faults. The effectiveness of the method is proved by extensive simulation results considering faults on three types of components: renewable generator, distribution grid and communication network.

Keywords: Energy systems, Machine learning, Neural networks
Abstract: This article reports a new state of health (SOH) estimation method for lithium-ion batteries using machine learning. Practical problems with cell inconsistency and online implementability are addressed using a proposed individualized estimation scheme that blends a model migration method with ensemble learning. A set of candidate models, based on slope-bias correction (SBC) and radial basis function neural networks (RBFNNs), are first trained offline by choosing a single-point feature on the incremental capacity curve as the model input. For online operation, the prediction errors due to cell inconsistency in the target new cell are next mitigated by a proposed modified random forest regression (mRFR) for high adaptability. The results show that compared to prevailing methods, the proposed SBC-RBFNN-mRFR-based scheme can achieve considerably high SOH estimation accuracy with only a small amount of early data and online measurements are needed for practical operation.

Keywords: Energy systems, Optimal control, Maritime
Abstract: Motivated by the ubiquitous presence of input uncertainty in the wave energy control problem, we propose, in this paper, a robust energy-maximising framework which explicitly considers potential wave excitation force deviations in the computation of the optimal control law, while systematically respecting state and input constraints. In particular, this is achieved by a suitable moment-based characterisation for the input uncertainty, taking into consideration an appropriate convex uncertainty set. The concept of moments is combined with well-known robust optimisation principles, by proposing a worst-case performance approach. We show that this novel moment-based robust optimal control framework always admits a unique global energy-maximising solution, hence leading to a computationally efficient robust solution. The performance of the proposed controller is illustrated by means of a case study, considering a heaving point absorber WEC.

Keywords: Electrical machine control, Predictive control for nonlinear systems, Identification for control
Abstract: In this work, a data-driven model predictive control (MPC) approach for the current control of synchronous machines is presented. The emph{model} of the motor is represented via a long-short term memory (LSTM) neural network (NN). The model is obtained purely from collected data and doesn't include any physical knowledge. As an online optimization using the obtained data-driven model is not easily implementable in the available sampling time, the neural model is used to solve an MPC problem offline. Finally, the control policy is learned via another computationally implementable NN that runs in real-time as a current controller. The proposed data-driven MPC controller is tested experimentally, and is bench-marked against MPC schemes that incorporate the well-known physically-based first-principles linear and nonlinear model of the machine.

Keywords: Constrained control, Energy systems, Optimization
Abstract: The development of safe and reliable battery fast-charging protocols is one of the key technologies needed to reduce our over-reliance on fossil fuel energy storage. Existing charging algorithms are typically generated by discretising a lumped parameter battery model in time and then solving a constrained numerical optimisation problem. One of the main limitations with this approach is that its performance is reliant upon the granularity of the time discretisation; too coarse a discretisation gives an inaccurate solution where the constraints may be violated between time steps; too fine a discretisation gives a rapid growth in computational complexity. To overcome this trade-off, a discretisation-free approach to solving the optimisation problem is proposed. By exploiting recent developments in the field of measure-moment theory, the method is able to generate discontinuous charging profiles from numerical solutions of a convex optimisation problem, which intrinsically satisfy the physical conservation laws of the model. Numerical examples highlight the potential of the approach for efficiently generating fast-charging protocols.

Keywords: Modeling, Agents networks, Agents and autonomous systems
Abstract: Recently a model for the interplay between homophily-based appraisal dynamics and influence-based opin- ion dynamics has been proposed. The model explores for the first time how the opinions of a group of agents on a certain number of issues/topics is influenced by the agents’ mutual appraisal and, conversely, the agents’ mutual appraisal is updated based on the agents’ opinions on the various issues, according to a homophily model. In this paper we show that a simplified (and, in some situations, more feasible) version of the model, that accounts only for the signs of the agents’ appraisals, provides an equally accurate model of the opinion dynamics in small networks. The equilibria reached by this model correspond, almost surely, to situations in which the agents’ network is complete and structurally balanced. On the other hand, we ensure that such equlibria can always be reached in a finite number of steps and, differently from the original model, we rule out other types of equilibria that correspond to disconnected social networks.

Keywords: Modeling, Feedback linearization, Linear systems
Abstract: We present a novel model approximation scheme, named "wienerization", of nonlinear time invariant systems that uses the systems' nonlinear equilibria maps as the static nonlinearities of Wiener models. We show conditions for when controllers based on such wienerization schemes exploiting correct system equilibria maps lead to asymptotically vanishing tracking errors. Furthermore, we show that this modeling scheme can be used to improve closed loop control of nonlinear systems without the need for modifying a preexisting linear controller nor re-tuning it, but rather simply adding opportune nonlinear static maps to the reference and output signals. To this end, we demonstrate a novel controller retrofitting scheme that may be implemented following an arbitrarily smooth transition from the original controller to the modified, improved one.

Keywords: Modeling, Model validation, Nonlinear system identification
Abstract: An original simplified 6-DOF on-ground aircraft model is developed, with the intent of providing an efficient means for novel braking control law synthesis. The proposed modeling approach consists of including the entire set of contributing dynamics, during aircraft high-speed braking, scaling the level of complexity and fidelity in function of the dominance of each physical phenomenon described. To this end, simplistic aerodynamic and propulsive models are coupled with a very detailed landing gear mathematical description, including shock-absorber, wheel-ground interaction and tire dynamics. Different modeling techniques are employed, such as analytical, parametric or tabulated models, depending on the subsystem nature and the related available data. The complete aircraft on-ground model is finally validated against a fully nonlinear multi-body simulator available at Safran Landing Systems. The proposed approach demonstrates high fidelity behavior, being capable of capturing critical phenomena such as load transfer between landing gear struts and wheel longitudinal-vertical dynamic coupling, while remaining simple and suitable for a control design purpose. Pertinent results are provided and discussed at the end of the paper.

Keywords: Modeling, Nonlinear system theory, Model/Controller reduction
Abstract: In this paper we study a class of compartmental models with bounded capacities, called generalized ribosome flow models and show that they are formally kinetic, i.e. we can assign a chemical reaction network (CRN) to the system based on the compartmental structure which realizes the nonlinear dynamics. We decompose the model into two dual subsystems both having positive linear first integrals representing conservation. Based on the dynamics of the reaction network we construct a port-controlled Hamiltonian representation of the system in the original and also in the reduced state spaces with clear connection between the structure matrices and the compartmental graph topology. We finally demonstrate that the system is l1-contractive both in the original and in the reduced state spaces. The generality of our framework ensures the the results are valid for a wide class of reaction rate functions used in the CRN representation.

Keywords: Modeling, Robotics, Autonomous robots
Abstract: In this paper, we present a torpedo-shaped novel autonomous underwater vehicle (AUV) with vectored tunnel thrusters. Traditional torpedo-shaped AUVs are incapable of hovering and lateral motion. We mitigate these deficiencies in the proposed design leading to greater maneuverability in constrained environments. We put forth the mathematical model of the vehicle and develop an appropriate control allocation. A simple PD controller is then implemented for pose stabilization which shows satisfactory performance in simulations.

Keywords: Modeling
Abstract: This paper proposes a methodology to forecast the number of hospital beds required by COVID-19 inpatients in mild and in critical conditions. To that end, a compartmental model is extended to include the number of critical inpatients, which require hospitalization in intensive care units (ICUs). The model parameters are tailored by using a data-driven approach and a computational methodology for numerical optimization. A multi-objective cost function is adopted, representing the match between the model output and the observed data for four variables, namely the total number of cases, demises, hospitalizations and ICU beds. Results for different regions of the Brazilian state of Sao Paulo are presented. The results show that the model represents well the training data and is able to predict the required health system resources.

Keywords: Decentralized control, Modeling, Electrical power systems
Abstract: This paper is concerned with sequential distributed development of multiple retrofit controllers managed by respective local controller designers. Retrofit control is a modular control approach for stable network systems. From the standpoint of a single controller designer, the subsystems managed by all other designers can be regarded as an unknown environment, an approximate model of which is supposed to be accessible in local controller design. This paper shows that, for the identification of approximate environment models, all environments for controller designers are invariant with respective to the design parameters in all retrofit controllers.　Furthermore, it is also shown that local control performance in response to a local disturbance is improved as improving the accuracy of a corresponding approximate environment model. The significance of the theoretical results is verified through an example of power systems control.

Keywords: Sampled data control, Distributed control, Decentralized control
Abstract: Recent work in data-driven control has led to methods that find stabilizing controllers directly from measurements of an unknown system. However, for multi-agent systems we are often interested in finding controllers that take their distributed nature into account. For instance, the full state might not be available for feedback at every agent. In order to deal with such information, we consider the problem of finding a feedback controller with a given block structure based on measured data. Moreover, we provide an algorithm that, if it converges, leads to a maximally sparse controller.

Keywords: Sampled data control, Constrained control, Hybrid systems
Abstract: This work addresses the problem of stabilization of linear systems under saturating aperiodic sampled-data control. Considering an impulsive model, a hybrid system formalism is applied to obtain local stability conditions for the origin of the closed-loop system. Thus, considering the class of quadratic clock (timer) dependent Lyapunov function candidates, conditions in the form of matrix inequalities are proposed to design the parameters of the control law. An LMI-based optimization problem aiming at computing the control law to maximize an estimate of the region of attraction of the origin (RAO) is therefore proposed.

Keywords: Sampled data control, Feedback linearization, Algebraic/geometric methods
Abstract: In this paper, we consider the problem of imposing Virtual Holonomic Constraints to mechanical systems in Euler-Lagrangian form under sampling. An exact solution based on multi-rate sampling of order two over each input channel is described. The results are applied to orbital stabilization of the pendubot with illustrative simulations.

Keywords: Sampled data control, Lyapunov methods, Energy systems
Abstract: The paper deals with the sampled-data asymptotic stabilization of the Acrobot at its upward equilibrium. The proposed controller results from the action of an Input-Hamiltonian-Matching (IHM) strategy that shapes the closed-loop energy combined with a Damping Injection (DI) feedback designed on the sampled-data equivalent model. Simulations show the effectiveness of the proposed controller.

Keywords: Electrical machine control, Adaptive control, Sampled data control
Abstract: In this paper the data driven control technique called Model-Free Adaptive Control approach is applied to a Permanent Magnet Synchronous Motor using two different double-loop control configurations. A careful theoretical analysis is reported, proving the boundedness of the tracking error and the Bounded-Input-Bounded-Output stability of the closed-loop system in both cases. A novelty here presented is considering time-varying reference variables for the motor. Simulation tests are also reported supporting the effectiveness of the proposed approaches. This work is a preliminary study, to be completed with intensive experimental tests and a comparative analysis with available control techniques.

Keywords: Reduced order modeling, Sampled data control, Stability of hybrid systems
Abstract: Virtually all industrial control systems are implemented digitally by a sample-and-hold device. The design and performance analysis of sampled-data control systems for large-scale systems, represented by a set of high-dimensional differential equations, is challenging. Therefore, commonly, the complex high-dimensional system (plant) model is reduced to 1) support computationally efficient performance analysis for a given digital controller or 2) support (sampled-data) controller synthesis. This paper proposes a novel approach for reduction of such systems taking the sampled-data effects into account in the reduction process. In addition, we propose a condition under which stability of the closed-loop sampled-data control system is preserved in the reduction. This easy-to-check condition depends on 1) reduction error on the continuous-time plant model, 2) the sampling interval, 3) the controller gain. Finally, under this condition also an error bound for the reduced-order system is provided. The proposed methodology is illustrated by numerical examples of a controlled structural dynamical system.

Keywords: Adaptive control, Sampled data control, Uncertain systems
Abstract: In this paper, a time discretization of the Modified Linear Extended State Observer (MLESO) is proposed to improve performance for large sample times. Stemming from Model Reference Adaptive Control (MRAC) and L1-adaptive control schemes, the goal of MLESO control is to approximately track a reference model. This is achieved by interpreting all deviations from the reference model as an effect of disturbances, estimating the disturbances, and rejecting the estimates by using a suitable control law. Fast sampling, which is favorable for accurate estimation and compensation, is often limited by hardware. To reduce the required sample time, the continuous-time MLESO is discretized in this paper. After discretizing the plant, state-space dead-beat methods are used to derive a discrete-time control law. Parts of the method can be used for general discrete-time dynamic disturbance rejection. The effectiveness of the proposed discretization scheme is demonstrated in simulation on a hydraulic system.

Keywords: Predictive control for nonlinear systems, Optimal control
Abstract: This paper investigates the use of event-triggered model predictive control (MPC) to enhance the performance of an enumeration-based MPC for a DC-DC boost converter. Enumeration-based MPC utilizing a discrete-time switched model has been used to track a converter’s output voltage to a given reference. The optimization process of enumeration-based MPC relies on evaluating the cost function for every enumerated sequence, which leads to a great number of computations. For time-triggered MPC, this optimization process is repeated at every control loop. To reduce online computations while maintaining comparable control performance, event-triggered MPC only runs the optimization problem when an event is triggered. The control performance of both time-triggered and event-triggered MPC are simulated and compared, with the advantage of using event-triggered MPC being clearly demonstrated.

Keywords: Predictive control for nonlinear systems, Variational methods, Modeling
Abstract: A multirate nonlinear model predictive control (NMPC) strategy is proposed for systems with dynamics and control inputs evolving on different timescales. The proposed multirate formulation of the system model and receding horizon optimal control problem allows larger time steps in the prediction horizon compared to single-rate schemes, providing computational savings while ensuring recursive feasibility. A multirate variational model is used with a tube-based successive linearization NMPC strategy. This allows either Jacobian linearization or linearization using quadratic and linear Taylor series approximations of the Lagrangian and generalized forces respectively, providing alternative means for computing linearization error bounds. The two approaches are shown to be equivalent for a specific choice of approximation points and their structure-preserving properties are investigated. Numerical examples are provided to illustrate the multirate approach, its conservation properties and computational savings.

Keywords: Predictive control for nonlinear systems, Machine learning, Optimization algorithms
Abstract: In the last years, Nonlinear Model Predictive Control (NMPC) has widely spread both in research and industrial contexts. NMPC reliability, however, depends on the accuracy of the model description, that can be subject to considerable uncertainty. On the other hand, advances in the machine learning techniques led to a renewed interest in data-driven control strategies for complex systems, defining a novel research field, namely Learning-based NMPC. In this manuscript, we present an open-source toolbox that combines MATMPC (a MATLABbased Fast NMPC solver) and gpr-pytorch (a Python library for Gaussian Process (GP) regression) to define an off-the-shelves framework for implementation of GP-based Learning-based NMPC. Starting from a nominal model and model mismatch data or input-output measurements, the toolbox allows to train the GP either as model mismatch estimator or black-box model and get automatically the information needed for the NMPC problem. An example of usage in an experimental scenario for the control of a Furuta pendulum is presented, and a Monte Carlo analysis is carried out in a simulative scenario.

Keywords: Predictive control for nonlinear systems, Optimal control
Abstract: We propose a model predictive control (MPC) approach for minimising the social distancing and quarantine measures during a pandemic while maintaining a hard infection cap. To this end, we study the admissible and the maximal robust positively invariant set (MRPI) of the standard SEIR compartmental model with control inputs. Exploiting the fact that in the MRPI all restrictions can be lifted without violating the infection cap, we choose a suitable subset of the MRPI to define terminal constraints in our MPC routine and show that the number of infected people decays exponentially within this set. Furthermore, under mild assumptions we prove existence of a uniform bound on the time required to reach this terminal region (without violating the infection cap) starting in the admissible set. The findings are substantiated based on a numerical case study.

Keywords: Predictive control for nonlinear systems, Output feedback, Chemical process control
Abstract: This paper proposes a computationally efficient predictive control law for non-linear systems, that is one that can easily be coded and implemented on low cost hardware. Moreover, it has a secondary core benefit that the core tuning parameter reduces to a single choice which is: how much faster than open-loop would you like the closed-loop to converge? Simulations demonstrate that for some non-linear systems this is a cheap and simple way of ensuring effective feedback, with constraint handling.

Keywords: Predictive control for nonlinear systems, Autonomous systems, Maritime
Abstract: This paper presents the docking control of an autonomous vessel using the nonlinear Model Predictive Path Integral (MPPI) approach. This algorithm is based on a path integral over stochastic trajectories and can be parallelized easily. The controller parameters are tuned offline using knowledge of the system and simulations, including nonlinear state and disturbance observer. The cost function implicitly contains information regarding the surrounding of the docking position. This approach allows continuous optimization of the trajectory with respect to the system state, disturbance state and actuator dynamics. The control strategy has been tested in full-scale experiments using the solar research vessel Solgenia. The investigated MPPI controller has demonstrated excellent performance in both, simulation and real-world experiments. This paper addresses the question of how the MPPI algorithm can be applied to dock a fully-actuated vessel and what benefits its application achieves.

Keywords: Predictive control for nonlinear systems, Feedback linearization, Optimal control
Abstract: We show how the solution to NMPC problems for a special type of input-affine discrete-time systems can be obtained by reformulating the underlying non-convex optimal control problem in terms of a finite number of convex subproblems. The reformulation is facilitated by exact (input-state) linearization, which is shown to provide beneficial properties for the treated class of systems. We characterize possible types of the resulting convex subproblems and illustrate our approach with three numerical examples.

Keywords: Automotive, Intelligent systems
Abstract: This paper presents a method that adjusts the operation of advanced driver-assistance systems (ADAS) to a specific location and driver. The method uses crowd- sourced data collected from multiple drivers in multiple locations/environments to predict the vehicle behavior of an individual driver in a previously unseen location/environment. This prediction can in turn be used for adapting the calibration of ADAS to the specific location/environment, as well as to the individual driver. We describe an algorithm for making predictions, which uses probabilities and quantile functions of empirical cumulative distribution functions to relate an individ- ual driver to the population. The paper reports a simulation study in SUMO (Simulation of Urban MObility), where an emergency braking system is adapted to individual drivers and to different road surface conditions. The results show that the algorithm is quickly able to make accurate predictions and consequently adjust ADAS to the specific location and driver.

Keywords: Fuzzy systems, Behavioural systems, Biomedical systems
Abstract: Driver Monitoring Systems are gaining increasing interest from car-makers, being featured by technologies capable to improve road safety levels and prevent accidents. Currently, these systems are mostly passive, based on distance traveled and travel time. The market trend is moving towards active systems, featured by popular approaches, such as behavioral and physiological. This article proposes a Driver Monitoring System to detect driver’s drowsiness with a hybrid technique that uses both approaches. The first performs a driver’s face analysis by monitoring ocular variables; the second is based on biometric signals, such as Heart Variability and Galvanic Skin Response. A distributed hardware/software architecture has been designed in order to develop real-time data acquisition and a fuzzy-based inference. Comparison tests between the hybrid approach and the two single ones showed the benefits of the proposed method in estimating driver’s drowsiness, considering a subjective estimation conducted during driving sessions with the CARLA simulator

Keywords: Automotive, Identification, Model validation
Abstract: Driving behavior prediction plays an increasingly important role in the development of autonomous driving functions and Advanced Driver Assistance Systems (ADAS). The more precise a prediction is, and the longer the temporal horizon it covers, the easier it becomes to correctly assess a situation and effectively perform a driving or intervention task. In most cases, this is relevant above all for maintaining safety. While on highways and in cities, it is primarily the surrounding traffic that influences the behavior of vehicles, on country roads it is often necessary for a driver to adapt the driving behavior to the road topology, making it challenging for behavior prediction. Therefore, we build a prediction model that uses the characteristics of the road to estimate the future range where vehicles will be located on country roads. The magnitude of the influence of different factors on the prediction is investigated, and the concept is validated in terms of accuracy and conservativeness. We show in simulation that the method can improve both comfort and fuel economy for an Adaptive Cruise Controller (ACC) compared to other prediction approaches.

Keywords: Reduced order modeling, Linear parameter-varying systems, Automotive
Abstract: A Restricted-Structure Non-linear Generalized Minimum Variance (RS-NGMV) algorithm is applied to a scalar quasi Linear Parameter-Varying (qLPV), or State-Dependent (SD), Electric Vehicle speed tracking control problem. The model represents the longitudinal vehicle dynamics with disturbance factors from the road and the environment such as road inclination, aerodynamic drag and the rolling resistance forces. The control problem is based on the longitudinal speed tracking under the impact of these disturbances with an emphasis on the inclination. The simulation studies consider constant speed, UDDS and HWFET drive cycle scenarios as the reference speed profiles. The Restricted-Structure (RS) controller is of low order and uses NGMV optimization to calculate the feedback gains. The results show that RS-NGMV is efficient in dealing with disturbances and parameter variations, and battery State of Charge (SOC) results are also presented.

Keywords: Automotive, Observers for linear systems, Robust control
Abstract: This paper focuses on the control of longitudinal self-driving racecar models with model uncertainty and proposes a robust solution that comprises an online disturbance estimator and a nonlinear compensation control feedback law. By modeling all uncertainty with respect to a nominal model as suitably disturbance signals and afterward exploiting unknown-input state observer theory, a lean and fast estimator is derived for the racecar model. The estimator does not require a priori knowledge of the uncertainty. Closed-loop stability of the proposed controller ensuring the asymptotic reconstruction of the system state and disturbance inputs as well as asymptotic tracking of desired longitudinal speed is proved. Simulations are presented to exemplify the functioning of the proposed solution and show its validity.

Keywords: Automotive
Abstract: Electric vehicles are becoming widely adopted; besides environmental advantages, electric power trains present peculiar characteristics that pave the way for reconsidering classical vehicle dynamics control (emph{e.g.} ABS, TC) and improving their performance. In this paper, we propose an ABS/TC system for full electric vehicles with 4-wheel drive. The regulator is derived from a nonlinear brake-based slip control. We modelled the wheel dynamics with an augmented single-corner model that includes the transmission, a crucial element. The controller has been tuned in simulation on the identified grey-box model and validated with an instrumented vehicle on ice and snow. It shows good performance relieving the driver of limiting the slips. The wheels are kept controlled and the magnitude of the average acceleration is increased with respect to professional driver performance.

Keywords: Automotive, Safety critical systems, Predictive control for linear systems
Abstract: We present a framework for shared vehicle control between a human driver and an autonomous system. The proposed control strategy, termed Driver-in-the-Loop Contingency Model Predictive Control (MPC), is inspired by the concept of contingency planning and is designed to intervene from the driver under emergency conditions in a manner that is smooth and not overly conservative. Driver-in-the-Loop Contingency MPC relies on the computation of invariant sets, which are used as MPC terminal sets. The proposed method is demonstrated on two longitudinal traffic scenarios: (1) vehicle-following, and (2) an intersection where the cross-traffic has the right of way. We use these examples to demonstrate the safety of the controller as well as the inherent trade-off between smooth intervention and minimal intervention.

Keywords: Predictive control for linear systems, Large-scale systems, Optimization
Abstract: Electron beam stabilization in a synchrotron is a large-scale disturbance rejection problem, with hundreds of inputs and outputs, that is sampled at frequencies higher than 10 kHz. In this feasibility study, we focus on the practical issues of implementing model predictive control (MPC) for the heavily ill-conditioned plant of the electron beam stabilization problem at Diamond Light Source. If the terminal cost matrix of the MPC problem is obtained from the discrete-time algebraic Riccati equation (DARE), the ill-conditioned plant results in an ill-conditioned Hessian. Here, we propose obtaining a stabilizing terminal cost matrix from a modified version of the DARE that includes a constraint on the condition number of the Hessian. We implement our control algorithm on the hardware designated for the imminent Diamond Light Source upgrade, and show that the modified terminal cost matrix allows MPC to be executed at the rate required for synchrotron control. MPC overcomes various problems of standard electron beam stabilization techniques, and we show that the successful implementation can increase the stability of photon beams in synchrotron light sources.

Keywords: Predictive control for linear systems, Power electronics, Switched systems
Abstract: Motivated by increasing precision requirements for switched power amplifiers, this paper addresses the problem of model predictive control (MPC) design for discrete-time linear systems with a finite control set (FCS). Typically, existing solutions for FCS-MPC penalize the output tracking error and the control input rate of change, which can lead to arbitrary switching among the available discrete control inputs and unpredictable steady-state behavior. To improve the steady-state behavior of FCS-MPC, in this paper we design a cost function that penalizes the tracking error with respect to a state and input steady-state limit cycle. We prove that if a suitable terminal cost is added to the FCS-MPC algorithm convergence to the limit cycle is ensured. The developed methodology is validated in direct switching control of a power amplifier for high-precision motion systems, where it significantly improves the steady-state output current ripple.

Keywords: Predictive control for linear systems, Robust control
Abstract: This paper considers linear discrete-time systems subject to possibly low dimensional disturbances, i.e. where the disturbance vector has a lower dimension than the state vector, and introduces a tailored system level parameterization. Based on that a robust model predictive controller is defined. It is shown that if the dimension of the possible disturbances is lower than the state dimension, this leads to a reduction in the variables of the online optimization problem. For the case of linear time-invariant systems, the controller is proven to ensure recursive feasibility, robust constraint satisfaction and input-to-state-stability. As a simulative example, an evasive maneuver of an autonomous car is considered. In the context of region of attraction and computational complexity, the proposed controller shows its benefits in comparison to a state-of-the-art tube-based and system level synthesis approaches.

Keywords: Predictive control for linear systems, Uncertain systems, Stochastic control
Abstract: Model Predictive Control is an extremely effective control method for systems with input and state constraints. Model Predictive Control performance heavily depends on the accuracy of the open-loop prediction. For systems with uncertainty this in turn depends on the information that is available about the properties of the model and disturbance uncertainties. Here we are interested in situations where such information is only available through realizations of the system trajectories. We propose a general scenario-based optimization framework for stochastic control of a linear system affected by additive disturbance, when the dynamics are only approximately known. The main contribution is in the derivation of an upper bound on the number of scenarios required to provide probabilistic guarantees on the quality of the solution to the deterministic scenario-based finite horizon optimal control problem. We provide a theoretical analysis of the sample complexity of the proposed method and demonstrate its performance on a simple simulation example. Since the proposed approach leverages sampling, it does not rely on the explicit knowledge of the model or disturbance distributions, making it applicable in a wide variety of contexts.

Keywords: Predictive control for linear systems, Autonomous systems, Stochastic systems
Abstract: In autonomous driving, it is essential to be able to avoid any type of collision with the environment by appropriate control. Therefore, the distance between vehicle and obstacles needs to be sufficiently large, providing a norm constraint e.g. for optimal control of the vehicle. In general, future positions of dynamic obstacles are highly uncertain and thus predictions are e.g. made using a stochastic model of the obstacle dynamics. We propose an application-independent framework that extends Linear Model Predictive Control to minimize the probability of norm constraint violation in the prediction horizon. Thus, for the autonomous driving application, the probability of collision is minimized. In contrast to Robust Model Predictive Control approaches, the proposed approach can deal with unexpected behavior of the obstacle without loss of feasibility.The applicability of the method is demonstrated in simulation of a vehicle that is successfully avoiding a suddenly emerging pedestrian.

Keywords: Predictive control for linear systems, Distributed control, Large-scale systems
Abstract: A distributed model predictive control scheme is developed for tracking piecewise constant references where the terminal set is reconfigured online, whereas the terminal controller is computed offline. Unlike many standard existing schemes, this scheme yields large feasible regions without performing offline centralized computations. Although the resulting optimal control problem (OCP) is a semidefinite program (SDP), an SDP scalability method based on diagonal dominance is used to approximate the derived SDP by a second-order cone program. The OCPs of the proposed scheme and its approximation are amenable to distributed optimization. Both schemes are evaluated using a power network example and compared to a scheme where the terminal controller is reconfigured online as well. It is found that fixing the terminal controller results in better performance, noticeable reduction in computational cost and similar feasible region compared to the case in which this controller is reconfigured online.

Keywords: Predictive control for linear systems, Identification, Stochastic systems
Abstract: Recently, direct data-driven prediction has found important applications for controlling unknown systems, particularly in predictive control. Such an approach provides exact prediction using behavioral system theory when noise-free data are available. For stochastic data, although approximate predictors exist based on different statistical criteria, they fail to provide statistical guarantees of prediction accuracy. In this paper, confidence regions are provided for these stochastic predictors based on the prediction error distribution. Leveraging this, an optimal predictor which achieves minimum mean-squared prediction error is also proposed to enhance prediction accuracy. These results depend on some true model parameters, but they can also be replaced with an approximate data-driven formulation in practice. Numerical results show that the derived confidence region is valid and smaller prediction errors are observed for the proposed minimum mean-squared error estimate, even with the approximate data-driven formulation.

Keywords: Safety critical systems, Fault detection and identification, Mechatronics
Abstract: Implementations of model predictive control on resource-limited embedded devices usually do not follow theoretical schemes which guarantee stability by design because of real-time and feasibility issues or conservativeness. However, a dependable design is essential for safety-critical systems. This work proposes a dependable model predictive control system architecture, which uses an offline computed set of states declared to be dependable. A method is introduced for computing a classification function of such sets based on binary trees and convex hulls. The approach is tailored for embedded hardware limited in computing capacity and memory. Finally, the approach is applied to a realistic model of the safety-critical magnet control system of the magnetic levitation vehicle Transrapid and deployed on a microcontroller.

Keywords: Safety critical systems
Abstract: The safety assessment of autonomous driving system is a major challenge in the automotive industry and the role of simulation in development and testing of autonomous vehicles has become predominant to significantly reduce the hundreds of millions of miles required to demonstrate the safety performance of such systems. In this paper, a novel methodology is presented to assess automated vehicles safety performance based on a multifactorial analysis of severity indicators in the vicinity of the under test self-driving car. The set of severity indicators includes commonly used time intervals (Inter Vehicular Time, Time to Collision, Time to Steer, ...), distance-based indicators, traffic congestion indicators and a newly developed indicator relying on overlapping geometrical regions. Unsupervised clustering techniques are then used to investigate the correlations, dependence among the whole set of indicators. To address the problem of combining these heterogeneous quantities to derive a global measure of dangerousness for a given scenario, appropriate scaling is performed and various aggregation methods are tested against cut-in, cut-out and cut-through scenarios.

Keywords: Safety critical systems, Iterative learning control, Autonomous systems
Abstract: We propose a method to perform set-based state estimation of an unknown dynamical linear system using a data-driven set propagation function. Our method comes with set-containment guarantees, making it applicable to safety-critical systems. The method consists of two phases: (1) an offline learning phase where we collect noisy input-output data to determine a function to propagate the state-set ahead in time; and (2) an online estimation phase consisting of a time update and a measurement update. It is assumed that known finite sets bound measurement noise and disturbances, but we assume no knowledge of their statistical properties. These sets are described using zonotopes, allowing efficient propagation and intersection operations. We propose a new approach to compute a set of models consistent with the data and noise-bound, given input-output data in the offline phase. The set of models is utilized in replacing the unknown dynamics in the data-driven set propagation function in the online phase. Then, we propose two approaches to perform the measurement update. Simulations show that the proposed estimator yields state sets comparable in volume to the 3σ confidence bounds obtained by a Kalman filter approach, but with the addition of state set-containment guarantees. We observe that using constrained zonotopes yields smaller sets but with higher computational costs than unconstrained ones.

Keywords: Delay systems, Safety critical systems, LMI's/BMI's/SOS's
Abstract: We address the problem of finding an ellipsoid to bound the set of the states that are reachable from the origin in finite time for Lure-type systems with time-varying delays and bounded input. The time-varying delays appear both in the linear and nonlinear parts of the Lure system, whereby the bound on the delay's magnitude and rate is known. A Lyapunov-Krasovskii type functional is used to obtain a bounding ellipsoid by solving linear matrix inequalities which are delay dependent. We illustrate the applicability of our results on an example derived from models of neuronal populations.

Keywords: Safety critical systems, Observers for nonlinear systems, Uncertain systems
Abstract: This paper proposes a data-driven set-based estimation algorithm for a class of nonlinear systems with polynomial nonlinearities. Using the system's input-output data, the proposed method computes a set that guarantees the inclusion of the system's state in real-time. Although the system is assumed to be a polynomial type, the exact polynomial functions, and their coefficients are assumed to be unknown. To this end, the estimator relies on offline and online phases. The offline phase utilizes past input-output data to estimate a set of possible coefficients of the polynomial system. Then, using this estimated set of coefficients and the side information about the system, the online phase provides a set estimate of the state. Finally, the proposed methodology is evaluated through its application on SIR (Susceptible, Infected, Recovered) epidemic model.

Keywords: Control over networks
Abstract: Cloud computing solutions enable Cyber-Physical Systems (CPSs) to utilize significant computational resources and implement sophisticated control algorithms even if limited computation capabilities are locally available for these systems. However, such a control architecture suffers from an important concern related to the privacy of sensor measurements and the computed control inputs within the cloud. This paper proposes a solution that allows implementing a set-theoretic model predictive controller on the cloud while preserving this privacy. This is achieved by exploiting the offline computations of the robust one-step controllable sets used by the controller and two affine transformations of the sensor measurements and control optimization problem. It is shown that the transformed and original control problems are equivalent (i.e., the optimal control input can be recovered from the transformed one) and that privacy is preserved if the control algorithm is executed on the cloud. Moreover, we show how the actuator can take advantage of the set-theoretic nature of the controller to verify, through simple set-membership tests, if the control input received from the cloud is admissible. The correctness of the proposed solution is verified by means of a simulation experiment involving a dual-tank water system.

Keywords: Discrete event systems, Supervisory control, Automata
Abstract: The nature of processing elements in embedded systems is witnessing a significant shift. Nowadays, embedded systems often composed of specialized multi-cores to satisfy the computation requirements of today's applications. On such heterogeneous multi-cores, execution of the same application may consume different processing times on distinct cores. Appropriate assignment of a given set of applications on the available heterogeneous cores can make a fundamental difference in terms of the performance achieved with respect to various constraints such as resource utilization, reliability, cost, space, etc. in resource-constrained safety-critical real-time systems. However, computation of optimal schedules for heterogeneous platforms is non-trivial and computationally expensive to be conducted on-line. This work presents an off-line optimal scheduler synthesis strategy for periodic non-preemptive task sets to be executed on heterogeneous multi-cores, using Supervisory Control of Timed Discrete Event Systems (SCTDES) as the design formalism. Existing SCTDES based scheduler synthesis mechanisms, prior to the work presented in this paper, do not provide the flexibility to model heterogeneity of the underlying platform. The efficiency of our scheme has been demonstrated using a practical case-study with Instrument Control application.

Keywords: Agents and autonomous systems, Cooperative control, Constrained control
Abstract: Control of multi-agent systems under temporal logical specifications has been popular due to its ability to tackle complex tasks that cannot be easily defined as classic control objectives. In this paper, a general class of leader-follower multi-agent systems subject to certain fragments of signal temporal logic (STL) specifications is considered. We first propose a funnel-based control strategy for the leader-follower multi-agent systems to enforce the satisfaction of the basic STL formulas by prescribing certain transient behavior on the funnels that constrain the closed-loop trajectories. A hybrid control strategy is then leveraged to satisfy the sequential STL formulas. Finally, a simulation example is given to illustrate the results.

Keywords: Cooperative autonomous systems, Predictive control for linear systems, Robotics
Abstract: This paper introduces a novel line of sight constraint for mixed-integer programming models based on intermediary points. Its effectiveness is evaluated in the context of a receding horizon trajectory planning problem for a multiagent system, in which connectivity of the communication network must be maintained. The proposed constraint is shown to provide better results in terms of overall conservatism when compared to an earlier approach. Simulations further illustrate our findings.

Keywords: Cooperative control, Agents and autonomous systems, Cooperative autonomous systems
Abstract: This paper addresses problems of bearing formation control for n-agent systems defined in d (dge 2)-dimensional space under switching sensing graph topologies. We extend concepts of textit{Bearing Persistently Exciting} (BPE) formation, textit{Relaxed Bearing Rigid} (RBR) formation, and textit{Infinitesimal Bearing Rigid} (IBR) formation by including dynamically changing graph topologies. Using bearing measurements, controllers are proposed for multi-agent systems under single-integrator dynamics. By assuming the desired formation is BPE, the proposed bearing-only controller ensures uniformly exponential convergence of the formation in terms of shape and scale under switching topologies. The special case of the desired formation involving constant bearings is also considered. Simulation results are provided to illustrate the performance of the proposed control laws.

Keywords: Cooperative control, Agents and autonomous systems, Cooperative autonomous systems
Abstract: This paper explores the properties of Bearing Persistently Exciting (BPE) formations defined in d (dge 2)-dimensional space, a class of time-varying bearing formations whose configuration can be uniquely determined up to a translational Euclidean vector using only bearing and velocity measurements. We provide results on necessary conditions, and necessary & sufficient conditions to guarantee that a formation is BPE. We also explore the concept of Relaxed Bearing Rigid (RBR) formation, a particular case of BPE formation, which relaxes the classical conditions on interaction topologies required by Bearing Rigidity to maintain a rigid shape. Distributed estimators are also proposed to estimate the configuration of the BPE formation using bearing and velocity measurements. Simulation results are provided to validate the proposed estimators.

Keywords: Cooperative control, Maritime, Lyapunov methods
Abstract: We propose a distributed control law that solves the tracking-in-formation problem for a group of underactuated autonomous marine vehicles interconnected over an undirected graph and subject to inter-agent collision-avoidance and con- nectivity constraints. The control approach is based on input-output feedback linearization using the so-called hand-position point as the output. Moreover, the control strategy is able to deal with limited knowledge on the target’s state and dynamics as well as with disturbances in the form of unknown irrotational ocean currents. We establish almost-everywhere uniform asymptotic stability of the output dynamics with guaranteed respect of the inter-agent constraints. A numerical simulation illustrates the effectiveness of the proposed approach.

Keywords: Agents networks, Distributed estimation over sensor nets, Cooperative autonomous systems
Abstract: In this paper we consider a state estimation problem featuring an autonomous linear system and a set of interconnected sensors, each able to measure, in general, different measures and able to acquire their measurements and interact with each other in an asynchronous fashion. In particular, as a first step in the solution of this problem, in this paper we focus on the case of two agents and we develop a provably convergent asynchronous algorithm. Specifically, assuming the sensors take measurements at distinct time instants, we develop a procedure to merge such an information and reconstruct the value assumed by the full system’s state at both time instants. The procedure is based on the idea of repeatedly projecting forward/backward in time the state estimated at one time instant and refine the estimation by comparing with the estimate obtained at the other time instant based on the other measure. As a result, the procedure is shown to converge to the actual systems’s state at both considered time instants. Notice that, although addressing a particular case, a step of the proposed algorithm has the potential to represent the basic interaction step among pairs of agents in a gossip-like interaction protocol, and in our experimental analysis we provide numerical evidence that supports this conclusion.

Keywords: Transportation systems, Traffic control, Cooperative autonomous systems
Abstract: Enhancing the adaptive cruise control (ACC) functionality with wireless communication links, in addition to sensors such as radars and/or lidars, cooperative adaptive cruise control (CACC) is facilitated. CACC enables driving at small intervehicle distances, forming a vehicle platoon system, while maintaining string stability (i.e., the disturbance is attenuated along the vehicle string). While string stability is a necessary requirement for the design of vehicle platoons, most of the works are limited to homogeneous vehicles or platoons with a single heterogeneity, which is not often the case in reality. In this work, we make an attempt to provide sufficient conditions for string stability for a very general case of heterogeneous vehicle platoons under the multiple-predecessor following (MPF) topology. More specifically, we focus on the following forms of heterogeneity: different parasitic lags, different time headways, different controller parameters and different communication time delays. First, we obtain sufficient conditions for string stability for the general case. Homogeneous vehicle platoons or platoons with a single heterogeneity constitute special cases. Since providing closed-form expressions for the general case is not feasible, we analyze some simpler, yet practical, cases. In particular, we allow the time headway and the controllers to be identical (this can be part of the design of the vehicle platoons) and we consider the case in which parasitic lags and communication time delays can be different. Case studies are considered in simulations in order to give more insights and limitations of our results.

Keywords: Sliding mode control, Robust adaptive control, Stability of nonlinear systems
Abstract: In this paper, a newly developed strategy for adaptive super-twisting sliding mode control is proposed. The designed algorithm provides self-tuning gains in order to achieve better performance against unknown perturbations with various amplitudes. The proposed adaptive protocol has been modified and it is different from those studied in the literature, since it only has one parameter which significantly reduces the tuning effort. In this approach, the gain variation velocity is tuned in accordance with the received data from the estimated upper bound of imposed disturbances, while the remaining parameter is just associated with the targeted accuracy in the vicinity of the origin. On the basis of Lyapunov's theory, the convergence of states is proved in the sense of finite-time stability. Numerical simulations have been provided to demonstrate the effectiveness of the proposed method.

Keywords: Sliding mode control, Transportation systems, Traffic control
Abstract: This paper deals with an eco-driving control problem for autonomous electric vehicles in urban traffic networks with signalized intersections. Specifically, a novel control scheme is proposed to make the vehicle travel through a sequence of signalized intersections while always catching green lights with minimum energy consumption. The proposal includes a speed reference generator, a sliding mode local controller and an event-triggered decision maker whose intelligence is provided by a pre-specified condition. This mechanism enables to determine when it is necessary to re-plan a new optimal velocity profile by the speed planner, which solves a sub-optimal version of the non-convex eco-driving optimal control problem due to the traffic lights constraints. The sliding mode controller instead enables a finite time tracking of the optimized speed reference and plays the role of compensator of the uncertainties affecting the vehicle dynamics. Such a robustness property in turn allows to limit the triggering events when a new optimization is solved to update the speed reference profile. The performance of the whole control scheme are finally assessed in simulation.

Keywords: Robust control, Distributed parameter systems, Sliding mode control
Abstract: In this paper, a wave equation with control input in one of its boundaries is considered. Using a Lyapunov approach, the control is designed as a homogeneous algorithm to drive the states to a vicinity of a reference signal, achieving exponential Input to State Stability of the closed-loop system. This homogeneity-based tracking control strategy is evaluated in a simplified earthquake model, allowing a controlled dissipation of its energy. In addition, simulations are conducted to support the robustness and the smoothness of the control signal depending on the homogeneous control parameter selection.

Keywords: Observers for linear systems, Fault estimation, Uncertain systems
Abstract: We address state estimation in the presence of faults and unknown disturbances combining unknown input observers (UIOs) and sliding mode observers. We consider a well-established UIO design for linear time-invariant systems and augment it with a nonlinear sliding mode action. This latter term deals with matched disturbances affecting the actuation channels, such as actuator faults, while the UIO provides geometric decoupling from the remaining exogenous inputs. We thoroughly present the analysis of the proposed observer, together with existence conditions stemming from the joint design. We also investigate how our design geometrically relates with other known results in the field of unknown-input state estimation, and discuss its benefits and pitfalls. An advantage of our design is that it allows reconstruction of the fault in finite time, under just boundedness assumptions, while other disturbances are rejected by the UIO. Numerical simulations show the effectiveness of the proposed method.

Keywords: Sliding mode control, Adaptive control, Robust adaptive control
Abstract: The present paper investigates a new adaptive controller for the second nonlinear systems subject to disturbances. The proposed controller combines a nonsingular fast terminal sliding-mode control (NFTSMC) and an adaptive approach. The adaptive law is based on barrier function (BF). The proposed controller is applied to quadrotor whose disturbances are bounded unknown. The singularity problem is avoided because derivatives of terms with fractional powers do not appear in the control law. Without overestimating the control gain, the suggested adaptive mechanism based on BF ensures a good convergence of the output system and maintains it in a preset region of zero, regardless of the upper bound of the disturbance. A methodology design of the proposed control method for the uncertain quadrotor system is presented. Simulations validate the suggested BF-based adaptive NFTSMC (BF-ANFTSMC).

Keywords: Sliding mode control, Control over networks, Robust control
Abstract: Even though networked control systems offer many benefits, the presence of a communication network in the control loop leads to some unfavorable effects such as variable time delays. This paper proposes a new optimal predictive sliding mode control strategy based on switching reaching law for multivariable networked systems affected by disturbances. The effectiveness of the proposed technique is examined in simulations, where in addition, a comparison to existing switching and nonswitching reaching laws is made. Furthermore, the sensitivity of the control algorithms with respect to a first order actuator dynamics is considered. It is shown that the predictive switching reaching law reduces high frequency oscillations in the reaching phase characteristic for the switching reaching law. Moreover, the proposed approach can be applied even if the parasitic actuator dynamics are slow, in comparison to the switching and nonswitching reaching laws.

Keywords: Sliding mode control, Optimal control, Power plants
Abstract: This paper presents a robust control strategy for pressurized water type nuclear power plants by combining the optimal linear quadratic Gaussian control strategy with the fractional-order theory based integral sliding mode control strategy. The proposed control scheme follows the reference set-point effectively in spite of the presence of uncertainties in the system by spending minimal control efforts. The nonlinear nuclear power plant model adopted in this study is characterized by 38 state variables. The nonlinear model is first linearised around steady state operating point to obtain a linear model for which a proposed control strategy is designed. Stability of the closed-loop system is proved with the help of Lyapunov theory. Finally, efficacy of the proposed control scheme for different control loops of the nuclear power plant is demonstrated through simulations and compared with conventional techniques.

Keywords: Intelligent systems, Energy systems, Neural networks
Abstract: This paper aims to maximize the total power generation for wind farms subject to strong wake effects and stochastic inflow wind speeds. A data-driven control method that only requires the accessible measurements of every turbine in the farm is proposed via deep reinforcement learning (DRL). We employ a grouping strategy to mitigate the high computational complexity induced by DRL and enhance our method’s applicability to large-scale wind farms. Based on the levels of aerodynamic interactions among turbines, this grouping strategy divides the whole farm into small sub-groups. Therefore, one can execute DRL on these sub-groups instead of carrying on a complicated learning process for the entire farm. Simulations verify the advantages of the proposed DRL based wind farm control method over the commonly employed greedy strategy. Results also show that the proposed method can significantly reduce the overall computing cost compared with the direct execution of DRL on the whole wind farm.

Keywords: Predictive control for linear systems, Energy systems, Power plants
Abstract: As wind turbines are designed with longer blades and towers, it becomes increasingly important to factor structural modes into the design of the controller. In classical turbine controllers, where pitch-speed, torque-speed, drivetrain and tower dampers are designed separately, it has for years been commonplace to base that design on a linearisation of the existing high fidelity aeroelastic model. Furthermore, any measurement filters that are required at run-time are included in the control loop shaping process. In contrast, most previous work on model predictive control (MPC) for wind turbines uses simplified models and ignores the need or effect of measurement filters. In this work, we demonstrate a mostly automatic design process that takes a detailed linearised model from an aeroelastic simulation package and adds linear filters and feedback, to produce a model predictive controller with low run-time computational complexity. The tuning process is substantially simpler than classical control, making it an attractive tool in industrial applications.

Keywords: Identification for control, Nonlinear system identification, Predictive control for nonlinear systems
Abstract: This work demonstrates the application of Koopman-based system identification on wind farm control, where wake interactions are highly nonlinear in nature. The linear models identified using measurements and signals available in real-time, i.e effective wind speed at the turbine rotors and control signals, show more than 85% variance-accounted-for (VAF). Different Koopman lifting function combinations, motivated by the 2D Navier-Stokes equations, governing the underlying wake interaction, are compared. The obtained Koopman model is used in closed-loop in the WFSim environment. The design of the qLMPC wind farm controller is provided and it is shown that the underlying quadratic programming (QP) converges in milliseconds thus making this design applicable in real-time to small wind farms. Finally, the results for power reference tracking obtained with qLMPC are shown based on estimated wind. It is demonstrated that using Koopman extended dynamic mode decomposition (EDMD) for wind estimation can lead to high-quality farm level control in the absence of wind measurements.

Keywords: Robust control, H2/H-infinity methods, Energy systems
Abstract: To meet ever-growing global demand of renewable energy, wind turbines have seen an exponential growth in size over the past few decades to capture more energy from wind. This has led to a growing concern of increased loading of wind turbine components as a result of additional weight and flexibility, hence, affecting operational and power reliability. To tackle this challenge, this contribution proposes a robust observer-based control strategy for mitigating structural loads as well as regulating power production of utility-scale wind turbines. Variation of the rotor effective wind profile, which is responsible for fatigue loading of turbine components, acts as a disturbance to the wind energy conversion system. Additionally, linear models used to design most wind turbine control systems have inherent modeling errors and do not capture nonlinearities caused by changing operating conditions. Although robust controllers address this challenge, they do not incorporate wind disturbance models. A few robust disturbance accommodating controllers have been proposed before to mitigate modeling errors and nonlinearities due to wind disturbances. However, these have been tested on smaller wind turbines. Their performance have also not been benchmarked against the latest baseline controllers. Therefore, this contribution proposes to extend this control strategy to an onshore 5 MW National Renewable Energy Laboratory (NREL) reference wind turbine (RWT) and compares its performance against the recently developed proportional-integral (PI)-based baseline controller from NREL, the reference open-source controller (ROSCO). Based on simulations results for various wind profiles, the proposed control scheme improves tower load mitigation and generator speed regulation performance.

Keywords: Neural networks, Machine learning, Maritime
Abstract: Accurate wind farm power prediction is of vital importance for the performance improvement of wind farms and their grid integration. In this paper, a novel method based on the state-of-the-art deep learning model (i.e. the Transformer network) is developed to tackle this issue. Specifically, the prediction task is modeled as a segmentation problem and the powerful Vision Transformer (ViT) is employed to predict each individual turbine’s power generation in a wind farm with wake interaction effects. The proposed method, called Wind Transformer (WiT), is evaluated by carrying out a set of numerical experiments. The results show that the proposed method achieves accurate and efficient wind farm power prediction and it outperforms other deep learning baseline models significantly. Particularly, the maximum mean absolute percentage error by the proposed method is only 1.030%, while they are 4.350% for LSTM and 3.510% for CNN models.

Keywords: Electrical machine control, Observers for nonlinear systems, Power plants
Abstract: This paper proposes a voltage control law based on the parameter estimation of a resistive load via a non-linear observer. The main benefit of the proposed observer is to show a dynamic of convergence independent of the state value. Simulations and experimental verifications are performed in order to verify the effectiveness of the controller and the observer.

Keywords: Electrical machine control, Adaptive control, Power electronics
Abstract: Predictive Stator Current Control (PSCC) is a flexible technique that has been the subject of research in connection with multi-phase drives. The flexibility that the Cost Function (CF) provides allows to consider different criteria usually found in variable speed drives. However, the tuning of the cost function is cumbersome. Intensive trial and error methods are usual in this context. In this paper an adaptive procedure is proposed based on the concept of Model Reference Adaptive Control (MRAC). The method provides on-line tuning of the CF parameters of PSCC: the Weighting Factors (WF). The proposal is motivated by the idea of using optimal WF for each operating point. A case study is developed for a six-phase Induction Machine (IM). The adaptive method is modified to include additional terms (cross-terms and momentum) to cope with the particular behavior observed in the gradient of the figures of merit with respect to the WF. The proposal is assessed with simulations and real experimentation on a laboratory setup.

Keywords: Discrete event systems
Abstract: This paper presents a mathematical tool for stochastic filter design based on reach sets for general Uncertain Max-Plus Linear (uMPL) systems. The reach sets are defined as the computation of the set of all states that can be reached from a known previous state vector (forward) and from an available source of measurement (backward). The existing approaches in [13, 10] have exponential complexity, which is an important drawback in high-dimensional systems. In this work, we propose a max-plus polyhedra-based procedure with complexity that is in practice polynomially-bounded.

Keywords: Computational methods, Complex systems, Nonlinear system theory
Abstract: Community detection is a challenging and relevant problem in various disciplines of science and engineering like power systems, gene-regulatory networks, social networks, financial networks, astronomy etc. Furthermore, in many of these applications the underlying system is dynamical in nature and because of the complexity of the systems involved, deriving a mathematical model which can be used for clustering and community detection, is often impossible. Moreover, while clustering dynamical systems, it is imperative that the dynamical nature of the underlying system is taken into account. In this paper, we propose a novel approach for clustering dynamical systems purely from time-series data which inherently takes into account the dynamical evolution of the underlying system. In particular, we define a emph{distance/similarity} measure between the states of the system which is a function of the influence that the states have on each other, and use the proposed measure for clustering of the dynamical system. For data-driven computation we leverage the Koopman operator framework which takes into account the nonlinearities (if present) of the underlying system, thus making the proposed framework applicable to a wide range of application areas. We illustrate the efficacy of the proposed approach by clustering three different dynamical systems, namely, a linear system, which acts like a proof of concept, the highly non-linear IEEE 39 bus transmission network and dynamic variables obtained from atmospheric data over the Amazon rain forest.

Keywords: Differential algebraic systems, Computational methods, Constrained control
Abstract: Introducing flexibility in the time-discretisation mesh can improve convergence and computational time when solving differential equations numerically, particularly when the solutions are discontinuous, as commonly found in control problems with constraints. State-of-the-art methods use fixed mesh schemes, which cannot achieve superlinear convergence in the presence of non-smooth solutions. In this paper, we propose using a flexible mesh in an integrated residual method. The locations of the mesh nodes are introduced as decision variables, and constraints are added to set upper and lower bounds on the size of the mesh intervals. We compare our approach to a uniform fixed mesh on a real-world satellite reorientation example. This example demonstrates that the flexible mesh enables the solver to automatically locate the discontinuities in the solution, has superlinear convergence and faster solve times, while achieving the same accuracy as a fixed mesh.

Keywords: Agents and autonomous systems, Distributed control, Computational methods
Abstract: This work analyses the formation control problem of a Multi Agent System (MAS) composed by unicycle-like robots with synchronous intermittent communication policies. In particular, two communication architectures have been considered. The first one is a formation control protocol for a directed networks of agents using synchronous intermittent information feedback, in which the consensus can be reached if the communication duration time across each interval is larger than a threshold value. The second setup is instead a formation control problem for time-varying formation of agents under sampling with multiple leaders. The system with sampled data is then converted into a time delay system, and the problem is tackled by Lyapunov analysis based on LMIs.

Keywords: Applications in neuroscience, Computational methods, Filtering
Abstract: In mathematical neuroscience, a special interest is paid to a working memory mechanism in the neural tissue modeled by the Dynamic Neural Field (DNF) in the presence of model uncertainties. The working memory facility implies that the neurons' activity remains self-sustained after the external stimulus removal due to the recurrent interactions in the networks and allows the system to cope with missing sensors' information. In our previous works, we have developed two reconstruction methods of the neural membrane potential from incomplete data available from the sensors. The methods are derived within the Extended Kalman filtering approach by using the Euler-Maruyama method and the Ito-Taylor expansion of order 1.5. It was shown that the Ito-Taylor EKF-based restoration process is more accurate than the Euler-Maruyama-based alternative. It improves the membrane potential reconstruction quality in case of incomplete sensors information. The aim of this paper is to investigate their pattern recognition facilities, i.e. the quality of the pattern formation reconstruction in case of model uncertainties and incomplete information. The numerical experiments are provided for an example of the stochastic DNF with multiple active zones arisen in a neural tissue.

Keywords: Statistical learning, Sensor and mesh networks, Optimization
Abstract: In this paper, we consider a subset selection problem in a spatial field where we seek to find a set of k locations whose observations provide the best estimate of the field value at a finite set of prediction locations. The measurements can be taken at any location in the continuous field, and the covariance between the field values at different points is given by the widely used squared exponential covariance function. One approach for observation selection is to perform a grid discretization of the space and obtain an approximate solution using the greedy algorithm. The solution quality improves with a finer grid resolution but at the cost of increased computation. We propose a method to reduce the computational complexity, or conversely to increase solution quality, of the greedy algorithm by considering a search space consisting only of prediction locations and centroids of cliques formed by the prediction locations. We demonstrate the effectiveness of our proposed approach in simulation, both in terms of solution quality and runtime.

Keywords: Linear parameter-varying systems, Model/Controller reduction
Abstract: This paper shows how the computational effort required for the computation of parameter-dependent generalized Gramians (PDGGs) for linear parameter-varying (LPV) systems with more than two scheduling parameters can be significantly reduced by using a grid-free approach. This approach is applicable to LPV systems that can be expressed as linear fractional representations. The constraints for generalized Gramians are reformulated according to the full block S-procedure (FBSP) lemma, which eliminates the need for gridding by imposing a DG-scale structure on the multipliers. A case study on numerical examples illustrates the benefits of PDGGs in the balanced truncation method and shows a considerable improvement in computation time when there are more than two scheduling parameters. Moreover, the results suggest that the conservatism due to using DG-scalings does not affect the quality of the reduced models.

Keywords: Switched systems, Decentralized control
Abstract: We propose a switched control law for a DC-DC Buck converter that enables a grid-aware operation of a distributed generator (DG) in a low-voltage islanded DC microgrid (MG). The qualifier grid-aware means that the DG adjusts its operation mode in dependency of the MG status. The resulting closed-loop system is a switched system, in which the subsystems possess different equilibrium points. By means of a time- and state-dependent switching logic together with dwell-time stability analysis methods, we derive sufficient stability criteria that ensure the existence of a unique and globally exponentially stable equilibrium point of the resulting closed- loop switched system. The performance of the proposed control is illustrated via a numerical example.

Keywords: Switched systems, Differential algebraic systems, Power electronics
Abstract: Switched descriptor systems with pulse width modulation are characterized by modes whose dynamics are described by differential algebraic equations; this type of models can be viewed as switched impulsive systems, i.e. switched systems with ordinary differential equations as modes dynamics and state jumps at the switching time instants. The presence of possible jumps in the state makes the application of the classical averaging technique nontrivial. In this paper we propose an averaged model for switched impulsive systems. The state trajectory of the proposed averaged model is shown to approximate the one of the original system with an error of order of the switching period. The model reduces to the classical averaged model when there are no jumps in the state. The practical interest of the theoretical averaging result is demonstrated through numerical simulations of a switched capacitor electrical circuit.

Keywords: Switched systems, Differential algebraic systems, Stability of nonlinear systems
Abstract: In this paper, we propose some sufficient conditions for checking the asymptotic stability of switched nonlinear differential-algebraic equations (DAEs) under arbitrary switching signal. We assume that each model of a given switched DAE is externally equivalent to a nonlinear Weierstrass form. With the help of this form, we can define nonlinear consistency projectors and jump-flow solutions for switched nonlinear DAEs. Then we use a different approach from the paper [12] to study the stability of switched DAEs via a novel notion called the jump-flow explicitation, which attaches a nonlinear control system to a given nonlinear DAE and can be used to simplify the common Lyapunov function conditions for both the flow and the jump dynamics of switched nonlinear DAEs. At last, a numerical example is given to illustrate how to check the stability of a switched nonlinear DAE by constructing a common Lyapunov function.

Keywords: Switched systems, Linear systems, Fault tolerant systems
Abstract: In this paper, we focus on the smoothness of transient responses after a system switch to propose a new augmented switching L2 gain including the evaluation of their differential. Moreover, we apply it to the initial state design of a newly-activated controller at a controller switch to establish its potential practicality as a design index.

Keywords: Optimal control, Energy systems, Switched systems
Abstract: This work investigates the solution of Mixed-Integer Optimal Control Problems (MIOCPs) for residential heating networks. The network consists of several buildings that are interconnected through a district heating network and a central heating plant. All buildings have access to decentralized heat generation, in the form of solar thermal collectors on the rooftops of the buildings. Buildings with surplus heat are intended to transfer heat to buildings with heating demands in order to prevent the activation of the central heating plant. Additional storages provide further flexibility for storage and utilization of heat. Binary variables represent the exchange relations between buildings and the central plant. For this system, we solve an MIOCP in two different ways. On the one hand, we keep the system related nonlinearities and apply the Combinatorial Integral Approximation (CIA) method to the arising Mixed-Integer Nonlinear Program (MINLP). On the other hand, we apply a linear reformulation yielding a Mixed-Integer Linear Program (MILP) which we solve using a standard MILP solver. We show that the MINLP approach has a computational advantage over the MILP approach, while yielding only slightly worse results in the single-digit percentage range for selected key figures.

Keywords: Computational methods, Hybrid systems, Complex systems
Abstract: We provide discrete abstractions of impulsive systems on Banach spaces. Thereby we explicitly allow infinite- dimensional state and input spaces, which makes it possible to cover a crucially important class of dynamical systems modeled by partial differential equations with jumps, referred to as impulsive evolution systems. Using the notion of the so-called alternating simulation function, we prove, under an incremental stability assumption, that there exists an approximate alter- nating simulation relation between the impulsive system and its discrete abstraction. We also provide conditions for the existence of an approximate alternating bisimulation relation. A notable feature of our work is that we propose a time-varying alternating simulation function that allows the construction of discrete abstractions for a broad class of impulsive systems in which both the flow and jumps are possibly unstable. Our method also covers the classes of time-varying impulsive systems and impulsive systems with an output map.

Keywords: Electrical power systems, Stability of hybrid systems, Discrete event systems
Abstract: As more distributed energy resources (DERs) are connected to the power grid, it becomes increasingly important to ensure safe and effective coordination between legacy voltage regulation devices and inverter-based DERs. In this work, we show how a distribution circuit model, composed of two LTC's and two inverter devices, can create voltage oscillations even with reasonable choices of control parameters. By modeling the four-device circuit as a switched affine hybrid system, we analyze the system’s oscillatory behavior, both during normal operation and after a cyber-physical attack. Through the analysis we determine the specific region of the voltage state space where oscillations are possible and derive conditions on the control parameters to guarantee against the oscillations. Finally, we project the derived parameter conditions onto 2D spaces, and describe the application of our problem formulation to grids with many devices.

Keywords: Network analysis and control, Stochastic systems, Agents networks
Abstract: We study a concentration phenomenon in a gossip model that evolves over a stochastic block model (SBM) with two communities. We study the conditional mean of the stationary distribution of the gossip model over the SBM, and show that it is close to the mean of the stationary distribution of the gossip model over an averaged graph, with high probability. As a consequence, regular (non-stubborn) agents in the same community of the gossip model over the SBM have stationary states with similar expectations. The results show that it is possible to use the gossip model over the averaged graph to approximate and analyze the gossip model over the SBM, and establish a correspondence between agent states and community structure of a network. We present numerical simulations to illustrate the results.

Keywords: Autonomous robots, Stochastic control, Randomized algorithms
Abstract: In this paper we provide a learning based algorithm for solving a state constrained stochastic control problem and apply it to controlling the motion of a 2D Cartpole system (2DCPS) i.e. a planar inverted pendulum robot. The goal of the proposed algorithm is to learn to generate a optimal Markov control policies, under the presence of environmental uncertainties, for the task of path following and pole balancing and avoiding obstacles along the prescribed path. To model the environment uncertainties, we use the standard machinery of stochastic differential equations (SDEs). The resulting state constrained stochastic control problem (SCP) is solved statistically via a maximum likelihood estimator (MLE). Relying on universal approximation power of neural-networks (NNs), we build our MLE using gated-recurrent-units (GRUs). Apart from being able to include state constraints, the novel feature of the estimator is the incorporation of ergodic policy generation and step length generation networks. Consequently, the estimator is practically able to handle (i) general SCP with generic (non-quadratic) cost functions, (ii) deal with heterogeneous and asynchronous sequential data. We apply our algorithm on a model dynamical system i.e. the 2DCPS and evaluate several training configurations. Based on the results we are led to conclude that the direct policy sampling method to control the forward process performs better than using the forward-backward SDEs (FBSDEs) framework for learning. Finally, by comparison, we show that the results of the optimally configured estimator is much better than results of a classical optimization method.

Keywords: Emerging control applications, Stochastic control, Neural networks
Abstract: Controlling the environmental parameters, including light in greenhouses, increases the crop yield; however, the electricity cost of supplemental lighting can be high. Therefore, the importance of applying cost-effective lighting methods arises. In this paper, an optimal supplemental lighting control approach is developed considering a variational inference Bayesian Neural Network (BNN) model for sunlight prediction. The predictive model is validated through testing the model on the historical solar data of a site located at North Carolina (R^2=0.9971, RMSE=1.8%). The proposed lighting approach is shown to minimize electricity cost by considering the BNN-based sunlight prediction, plant light needs, and variable electricity pricing when solving the underlying optimization problem. For evaluation, the new strategy is compared to: 1) a Markov-based prediction method, which solves the same optimization problem, assuming a Markov model for sunlight prediction; 2) a heuristic method which aims to supply a fixed amount of light. Simulation studies are conducted to examine the electricity cost improvements of the BNN-based approach. The results show that the BNN-based approach reduces cost by (on average) 2.27% and 43.91% compared to the Markov prediction-based method and the heuristic method, respectively, throughout a year.

Keywords: Stochastic control, Optimal control, Optimization
Abstract: We consider the problem of optimal investment in a market with borrowing and stochastic interest rates. We assume a quadratic-affine model for the bond rate which includes the Hull-White and Cox-Ingersoll-Ross models as special cases. Due to borrowing, this is an optimal stochastic control problem with nonlinear system dynamics. An explicit closed-form solution is found for the case of a power and logarithmic utility from terminal wealth as a linear state-feedback control.

Keywords: Stochastic control, Stochastic systems, Optimal control of communication networks
Abstract: We characterize the signalling and control rate of nonlinear partially observable stochastic control or decision models (DMs), called operational control-coding (CC) capacity. This is defined as the maximum signalling rate in bits/second, of encoding information signals into randomized control strate- gies, and reproducing them asymptotically at the output of the DM, with arbitrary small error probability. We show that the CC capacity is characterized by an extremum problem of an information theoretic pay-off, with information state the posteriori distribution of nonlinear filtering, subject to an average cost constraint. The dual of the CC capacity is characterized by the minimization of the average cost subject to a rate constraint. As an application example, we consider the partially observable, linear-quadratic Gaussian (LQG)-DM. We show that optimal randomized control strategies consist of an estimation and control part which controls the unobserved state of the DM, and an information transmission/signalling part which signals information through the DM.

Keywords: H2/H-infinity methods, Stochastic control, Robust control
Abstract: The problem of robust static Linear Quadratic Regulation (LQR) problem is solved for uncertain, continuous time, retarded systems with multiplicative noise. A design solution is obtained, based on a descriptor approach, for the uncertain polytopic case where the parameters of the system matrices reside in a given polytope. The latter solution method enables the derivation of the required constant output gain by solving a set of linear matrix inequalities that correspond to the vertices of the uncertainty polytope. The theory developed is extended also to the gain-scheduling case.

Keywords: Genetic regulatory systems, Modeling, Stochastic systems
Abstract: We present a general framework for studying intrinsic fluctuations in gene expression that arise from RNA post-transcriptional splicing. In particular, we decompose the coefficient of variation for the full system into terms that capture the effects from the various stochastic processes that make up the model. Our analysis reveals that the pre-mRNA conversion to mature mRNA stochastically deletes the steady state correlation between the involved chemical species. Moreover, the noise decomposition suggests that under constitutive gene expression the pre-mRNA conversion is capable to restore the Poissonian limit of the downstream chemical species. Finally, interesting biological insights were found suggesting the role of splicing factors in altering the stochastic noise of the chemical species.