Keywords: Adaptive control, Disturbance rejection, Linear systems
Abstract: An adaptive algorithm, compensating for unknown harmonic disturbance acting for linear objects under conditions of the unavailable state vector with a defined delay in the control channel is proposed. One of the features of the proposed method in comparison with other methods is that the perturbation signal is considered in the form of products of sinusoids. A new approach is proposed for estimating the frequencies of harmonic signal. It is assumed that all parameters of the multiharmonic disturbance (amplitude, frequency, and phase) are unknown. The task is completed in several steps. First, an observer is constructed based on a frequency estimation scheme. Secondly, stabilization of the ouput of object to zero is carried out using feedback based on the predictor. Examples are given that confirm the relevance of the proposed approach. Our main contribution is to propose a new scheme for compensating external disturbances for a linear plant and a new approach for estimating the frequencies of a multisinusoidal signal.

Keywords: Adaptive control, Neural networks, Formation control
Abstract: This article proposes a distributed control law with a neural network to achieve leader-follower affine formation maneuver (AFM) of multi-agent systems (MASs) subjected to time-varying external disturbances. The leaders determine the desired collective formation maneuvering such as scaling, shearing, translation, and rotation. The distributed controller ensures that the followers are tracking the maneuvers of the leaders at all times despite the disturbances. The disturbances are approximated and compensated with the aid of neural networks. A Lyapunov candidate function is used to highlight that the closed-loop system is uniformly ultimately bounded. Finally, in order to validate the proposed control method, it is applied to a multi-agent system of quadrotors. The simulation results have shown that the proposed control protocol is able to maintain the collective maneuvering of the quadrotors in the presence of time-varying disturbances.

Keywords: Adaptive control, Nonlinear control, Robotics
Abstract: Remotely operated vehicles (ROVs) are often used for inspection in aquaculture net pens which serves the important purpose of localizing holes in the net and reporting potential irregularities and damages. Manual control of the vehicle inside a net pen, while simultaneously inspecting the net structure, is difficult and puts a lot of stress on the vehicle operators. Adaptation of new solutions that enables autonomous traversal of net pens where the vehicle maintains a fixed distance, heading, and velocity relative to the net is considered essential. One of the main challenges of such autonomous solutions is a robust and tight control of the vehicle's velocities. To target this challenge, this paper presents adaptive speed controllers for the surge and sway speeds of a remotely operated vehicle with unknown parameters and under the influence of unknown external disturbances. The stability properties of the controllers are proven through Lyapunov theory, and both simulations and field experiments demonstrate their ability to track the desired speeds through the use of a net following scheme.

Keywords: Adaptive control, Optimisation, Intelligent transportation systems
Abstract: Controlling platoons of heterogeneous vehicles is a relevant field of research. In terms of platoon architecture, many approaches to harness inter-vehicle communication to develop cooperative adaptive cruise control (CACC) systems have been investigated. For instance, optimal control has proven to yield high performance results. Yet, heterogeneity in vehicle dynamics over a platoon poses many challenges. For real world application, one major issue is the uncertainty of model parameters. Furthermore, differences in lagged response to a control input degrade performance criteria such as tracking and cohesion of a vehicle string. Adjusting the spacing with both the preceding and the succeeding vehicles using bidirectional communication can potentially improve the platoon cohesiveness and is crucial when considering the synchronized merging scenario. In this paper, an adaptive optimal controller is presented which relies on bidirectional platoon communication and deals with parameter uncertainties. Along with the adaptive optimal controller synthesis and string stability analysis, effectiveness of the controller is verified in a simulation example.

Keywords: Adaptive control, Predictive control, Optimisation
Abstract: In the modern era of manufacturing automation, the integration of sensor technology into the system ensures that data acquisition and analysis from complex systems become more efficient than ever. With the support of such developments, artificial intelligence-powered control in industrial control domains gains popularity and the traditional human-based PLC control, where the machines can monitor themselves, learn from the experience, and make their own decisions. However, despite advances in sensor technologies, most of the sensors in industries have limitations in observing the current status of the system, which is mostly limited to Boolean output data instead of continuous output. Therefore, such limitation forms a low dimensional state representation of the system, which could be problematic to develop a self-control policy, e.g. using a model-free deep reinforcement learning. In this paper, we present an effective model predictive controller with adaptive PLC-based policy on low dimensional state representation specifically for industrial control domains. First, we learn the model of the production system using the deep learning method, in case the digital twin is not available. Second, we set up a native implementation of model predictive control. Third, we augment the model predictive control with adaptive PLC-based policy. The proposed method is implemented into a bulk good system showing its potential to self-optimize the system by satisfying the production objective without overflow and low power consumption.

Keywords: Adaptive control, Time-delay systems, Disturbance rejection
Abstract: The problem of unknown external disturbances compensation for a class of linear systems with an unmeasured and delayed state is considered. The proposed solution is based on the use of the internal model principle and the extended error adaptation algorithm. It is assumed that the disturbance is the output of an autonomous linear generator with unknown parameters. A special observer is constructed to estimate the disturbance. The proposed approach does not require identification of the disturbance parameters. It is shown that in the presence of any state delay, the control algorithm preserves the stability of the closed-loop system. The performance of the obtained result is confirmed using computer simulation.

Keywords: Autonomous systems, Predictive control, Optimisation
Abstract: While existing algorithms for open water navigation typically address path following and COLREGS compliant collision-avoidance, the unique challenges of inland waterways require a more tailored approach. We propose a two-level control strategy that employs Model Predictive Control (MPC) and Scenario-Based Model Predictive Control (SB-MPC) for path following and collision-avoidance. The algorithm proposes integrated strategies for handling riparian land, static obstacles, and dynamic obstacles. The method is tested in simulation.

Keywords: Process control, Predictive control, Networked systems
Abstract: In recent years, cyber-security of networked control systems has become crucial, as these systems are vulnerable to targeted cyber-attacks that compromise the stability, integrity and safety of these systems. In this work, secure and private communication links are established between sensor-controller and controller-actuator elements using semi-homomorphic encryption to ensure cyber-security in Model Predictive Control (MPC) of nonlinear systems. Specifically, Paillier Cryptosystem is implemented for encryption-decryption operations in the communication links. Cryptosystems, in general, work on a subset of integers. As a direct consequence of this nature of encryption algorithms, quantization errors arise in the closed-loop MPC of non-linear systems. Thus, the closed-loop encrypted MPC is designed with a certain degree of robustness to the quantization errors. Furthermore, the trade-off between the accuracy of the encrypted MPC and the computational cost is discussed. Finally, a two-state multi-input multi-output continuous stirred tank reactor (CSTR) example is presented to demonstrate the implementation of the proposed encrypted MPC design.

Keywords: Neural networks, Nonlinear control, Predictive control
Abstract: In recent years, modeling of nonlinear systems has increasingly involved machine learning (ML). Recurrent neural networks (RNNs), a type of supervised learning technique, have shown to be effective in modeling time series data. Particularly, it has been demonstrated in several works that physics-informed RNN models (where the network structure is informed by the pattern of interactions of physical process variables) are preferable to dense RNN models. Motivated by this, the present work focuses on the generalization error of partially-connected RNN models and its relationship to the corresponding error of fully-connected RNN models for the same training and testing data sets. The RNN models are subsequently used in model predictive control of nonlinear processes. Through the use of a chemical process example, the advantages of the use of partially connected RNN models in MPC are illustrated via open-loop and closed-loop simulations.

Keywords: Predictive control, Automotive control, Real-time control
Abstract: This paper presents an implementation of a nonlinear model predictive control algorithm for autonomous driving applications. The algorithm is based on a two-track nonlinear model of the vehicle that takes into account Kamm's friction circle and a modified slip definition. It enables all-wheel drive torque vectoring and active front steering. The control algorithm was developed and evaluated through experiments conducted on a scaled four-wheel-drive electric car tested on a treadmill that serves as a roadway simulator. The dSPACE MicroLabBox was used to implement the control algorithm. The algorithm controls both the steering and the torques applied to each wheel based on the desired yaw rate. Its performance was evaluated using double lane change and multiple lane change maneuvers.

Keywords: Predictive control, Intelligent control systems
Abstract: We propose a reinforcement learning framework where an agent uses an internal nominal model for stochastic model predictive control (MPC) while compensating for a disturbance. Our work builds on the existing risk-aware optimal control with stochastic differential equations (SDEs) that aims to deal with such disturbance. However, the risk sensitivity and the noise strength of the nominal SDE in the riskaware optimal control are often heuristically chosen. In the proposed framework, the risk-taking policy determines the behavior of the MPC to be risk-seeking (exploration) or riskaverse (exploitation). Specifically, we employ the risk-aware path integral control that can be implemented as a Monte-Carlo (MC) sampling with fast parallel simulations using a GPU. The MC sampling implementations of the MPC have been successful in robotic applications due to their real-time computation capability. The proposed framework that adapts the noise model and the risk sensitivity outperforms the standard model predictive path integral in simulation environments that have disturbances.

Keywords: Disturbance rejection, Automotive control, Predictive control
Abstract: This paper investigates the task of lateral disturbance compensation based on model predictive control (MPC) for autonomous vehicles. By considering external disturbances and parameter perturbations in the model term of the MPC, the steady-state offset can be compensated. However, in the presence of more dynamic disturbances, like side wind, the lateral path tracking performance deteriorates. To overcome this limitation, a cascaded approach is presented, which is a combination of an MPC-based and an underlying direct compensation. The performance of this approach is validated in simulations as well as in practice with real vehicle tests.

Keywords: Discrete-event systems, Industrial automation, manufacturing, Event based systems
Abstract: In this paper, the three units of a Modular Production System (MPS) benchmark with a parametric number of drilling tools are modelled using finite deterministic automata in the discrete event system framework. The desired behavior of the MPS benchmark process together with the desired sequence of manufacturing actions of the MPS are expressed in the form of specification rules. The rules have been expressed in the form of a set of regular languages. Each regular language has been realized by a supervisor automaton. Using these supervisors, a modular supervisory architecture has been proposed. The physical realizability and the nonblocking property of the controlled automaton have been proved.

Keywords: Industrial automation, manufacturing, Optimisation, Modelling and simulation
Abstract: In recent years, the global wood products market has become highly competitive. Due to this, sawmills seek to improve their efficiency throughout their production process. In this regard, improving sawing efficiency through improved cutting strategies is vital for preventing overproduction and waste issues. In this paper, we deal with the sawing optimization problem defined as the problem of cutting rectangular boards from cylindrical logs with circular cross sections. In particular, we consider a sawing pattern that is highly beneficial for wood manufacturing, namely cant sawing. We take into account feasibility, capacity, non-overlapping, and technical constraints of the sawing process. We first develop an exact model of this combinatorial optimization problem as a mixed-integer nonlinear programming (MINLP) problem. However, this exact model involves a high level of combinatorics and requires considerable computation time, becoming computationally intractable as the problem size increases. To deal with this challenge, we develop a constructive heuristic approach, namely strip-bottom-left-fill (SBLF) heuristic, that builds a feasible cutting according to a list of ordered rectangles and a set of placement policies. The simulation results confirm the superiority of our proposed approach over the MINLP model and a state-of-the-art heuristic approach in terms of computational effort as well as memory and search requirements while preserving cutting yield efficiency.

Keywords: Networked systems, Optimisation, Multi-agent systems, Data reconciliation
Abstract: Data reconciliation is an essential tool in data processing in various industries. It helps to improve accuracy of decision-making algorithms by reducing the influence of random errors in measurements. In this paper, we consider large-scale data reconciliation problems in which multiple areas communicate over a network to obtain an optimal solution of the centralized problem. Our proposed approach accounts for the boundaries between different areas avoiding a mismatch and sub-optimality as well as reduces computational and communication complexities. The proposed distributed data reconciliation method is compared to a centralized reference in different scenarios.

Keywords: Predictive control
Abstract: The Parallel Flow Regenerative (PFR) lime kiln process is challenging with respect to the energy efficiency, product quality and production stops, due to the inability of the human operators to accurately predict the evolution of the process. Monitoring and controlling of such processes encounter several issues, related to the high mass and heat inertia of the process, data quality, production stops, operator’s experience, as well as unknown exogenous factors (e.g., quality of the fuel, and raw material properties). Hence, an automated control/optimization mechanism for properly configuring the process is not straightforward. In this paper, we present a selection of mechanisms for data preprocessing together with domain specific feature analysis that allow for capturing the short-term changes of the critical parameters of the process. Through these mechanisms, automated predictive modeling can be performed that can be used by the kiln operator or a predictive-based controller to modify fuel feed strategies to meet energy efficiency and product quality requirements. We validate the proposed data-based preprocessing and modeling approaches through experiments in real-world data sources

Keywords: Swarms, Linear systems, Algebraic and geometric methods
Abstract: The location selection of sensors in large-scale swarm systems is a prerequisite for further design of mechanisms to monitor the system states. This paper considers the required number and location of the sensors in a large-scale swarm system so that the observability of the overall system is satisfied. Firstly, by extending observability theory for swarm systems, some necessary and/or sufficient observability conditions related to the node-dynamics, network topology, coupling mode and measured outputs are obtained. Secondly, based on the above observability conditions, an algorithm for deciding how many and where to place the sensors is designed, which can be implemented in a polynomial complexity time. Finally, an unmanned aerial vehicle (UAV) swarm system is employed to verify the effectiveness of the theoretical results.

Keywords: Wireless sensor networks, Networked systems
Abstract: In the present and rapidly evolving industrial scenario, wireless networked controls are gaining importance. This brings about new problems concerning the use of the radio channel, as well as the energy efficiency of the involved devices (often running on battery). We argue that such problems, among which a fundamental one is clock synchronisation, should be addressed by dedicated control structures embedded in the used hardware/software architecture, and that the construction of the said controls should follow strictly system-theoretical principles to the maximum extent. In this paper, building on previous experience, we present such a synchronisation solution together with a formal model for its operation, also accounting for non-idealities in the reference time base. Some experimental results are reported to support the statements made.

Keywords: Aerospace control, Modelling and simulation, Complex systems
Abstract: The aim of this study is to examine the effects of dynamic morphing of the trailing-edge. This particular type of aerodynamic lift mechanism brings new opportunities to control lift distribution and thus control an aircraft. Morphing wing concepts are developed for their high efficiency and other benefits. However, this variable geometry can potentially perform even better by using non-stationary aerodynamic effects. In this article, we demonstrate the first results of this novel approach which will be used to investigate further control strategy development.

Keywords: Feedback stabilization, Nonlinear systems, Nonlinear control
Abstract: We consider the stabilization problem for driftless control-affine systems under the bracket-generating condition. In our previous works, a class of time-varying feedback laws has been constructed to stabilize the equilibrium of a nonholonomic system under rather general controllability condition. The latter stabilization scheme is based on the sampling concept, which is not equivalent to the definition of classical solutions for the corresponding non-autonomous closed-loop system. In the present contribution, we refine the previous results by presenting sufficient conditions for the convergence of classical solutions of the closed-loop system to the equilibrium. These conditions are illustrated with numerical simulations of the Brockett integrator.

Keywords: Optimisation, Nonlinear systems
Abstract: We study an optimal control problem for a nonlinear hyperbolic equation with boundary input which describes isothermal chemical reactions in a plug flow reactor. The considered optimization problem is analyzed in the class of periodic controls under an isoperimetric constraint. A dimensionless formulation of this isoperimetric problem is derived, and optimality conditions with piecewise constant controls are formulated. The behavior of the cost functional for such bang-bang controls is illustrated by numerical simulations.

Keywords: Linear systems, Time-delay systems
Abstract: We study the asymptotics of C_0-semigroups on the domain of the generator. In particular we analyze the behavior of |T(t)(A-lambda I)^{-1}| as time goes to infinity and develop some existing stability results (semi-uniform stability) to the case when the intersection of the spectrum of the generator with the imaginary axis is non-empty. We also give an example of a class of delay differential equations for which our theorem is applicable.

Keywords: Linear systems
Abstract: The problem of exact observability of a model of elastic-coupled strings is considered. The lack of regular exact observability is noted and required additional smoothness of observed signal is proposed.

Keywords: Nonlinear systems, Optimisation
Abstract: We consider nonlinear multi-input non-autonomous control systems and analyze their invariants analogous to those introduced in Sklyar K. On mappability of control systems to linear systems with analytic matrices. Systems Control Lett. 134 (2019) 104572. We show that, compared to single-input systems, new invariants should be introduced. We give a complete set of invariants for one subclass of multi-input non-autonomous systems and propose a method of solving the time-optimal problem for such systems.

Keywords: Biomedical engineering, Optimisation, Adaptive control
Abstract: With the help of in-vitro simulators, it is possible to simulate human physiological conditions to test medical equipment, accelerating innovation cycles and exploring the search for new and efficient solutions. In this paper, we consider the Pulse Duplicator in use at the University of Padova Healing Research Laboratory in Italy, for testing the effectiveness of prosthetic heart valves under realistic cardiac settings. By using a dither-free extremum seeking controller, that uses 1st order least squares fits for gradient estimation, we automatically adjust a fundamental system parameter in real-time, i.e. a system valve closing degree, that ensures a physiological pressure drop to simulate the peripheral resistance to flow in the human systemic circulation.

Keywords: Discrete-event systems, Computational methods, Biologically inspired systems
Abstract: In this paper, we quantify the impact of non-pharmaceutical interventions (NPIs) on the spread of COVID - both individually and in various combinations. We utilize the previously developed PanSim agent-based model to accurately capture various aspects of the epidemic and the interventions and show how the transmission rate (β) commonly used in compartmental ODE models can be matched to the agent-based model and used to compare interventions. Through a specific example of targeting a desired level of peak hospitalization, we give several equivalent intervention packages that can be imposed at various times during a single wave to achieve this goal. By mapping out the effect of different combinations of interventions on the transmission rate, we pave the way for coupling the PanSim model with advanced feedback control.

Keywords: Neural networks, Biomedical engineering, Modelling and simulation
Abstract: Hypoglycemia prevention is one of the main challenges of an efficient Type 1 diabetes control. Alarm Systems (ASs) that alert the patients about upcoming hypoglycemia are useful instruments to act in advance and avoid it. Model-based ASs use patient models to predict future Blood Glucose (BG) levels and to activate alarms, so model prediction capabilities highly influence the AS performance. In recent studies, neural network techniques for BG forecasting obtained promising results, both as population and personalized models. In particular, Personalized Long Short-Term Memory models (P-LSTMs) for BG predictions obtained good results in literature on the 100 in silico patients of the UVA/Padova simulator. In this work, personalized ASs for hypoglycemia prevention based on P-LSTMs are proposed. The ASs are able to predict correctly the 79% of the hypoglycemia events with a precision of 87%. These performances could be improved for some critical patients that present poor prediction performances. An enhanced version of the P-LSTM is currently under study.