Keywords: Energy systems, Power plants
Abstract: Power converters pose fascinating control challenges due to their switched linear system characteristic and the need to operate them in the microsecond time scale. Converters are omnipresent and vital to decarbonize our planet with renewable energy systems and the electrification of the transportation sector and industry. Model predictive control maximizes the performance of such converters by boosting their power, minimizing their cost and increasing their resilience to disturbances and faults. This plenary provides an introduction to the intriguing world of power conversion control, its research challenges and showcases some promising results that are already revolutionizing industry.

Keywords: Fault detection and identification, Safety critical systems, Optimization algorithms
Abstract: Model-based fault detection identifies anomalies by comparing a system’s output with the prediction from a model. Although such a technique can be very powerful, it may suffer from the computational complexity of its underlying models, especially for large systems. An alternative approach that circumvents this cost increase uses barrier functions, which abstract the system’s behaviour into a single value. In this paper, we propose a fault detection mechanism via output-based barrier functions, that does not require to estimate the full state, copes with noisy processes, and is tailored to safety-critical faults as given by a user-defined safe region. We leverage such a mechanism by introducing so-called p-fault tolerant sets, which guarantee that a faulty system requires at least p time steps before reaching any unsafe state. Our approach is validated through numerical experiments on two systems with linear and nonlinear dynamics, along with the classic three-tank model.

Keywords: Fault detection and identification, Fault diagnosis, Maritime
Abstract: This paper introduces a Fault Diagnosis (Detection, Isolation, and Estimation) method using Set-Membership Estimation (SME) designed for a class of nonlinear systems that are linear to the fault parameters. The methodology advances fault diagnosis by continuously evaluating an estimate of the fault parameter and a feasible parameter set where the true fault parameter belongs. Unlike previous SME approaches, in this work, we address nonlinear systems subjected to both input and output uncertainties by utilizing inclusion functions and interval arithmetic. Additionally, we present an approach to outer-approximate the polytopic description of the feasible parameter set by effectively balancing approximation accuracy with computational efficiency resulting in improved fault detectability. Lastly, we introduce adaptive regularization of the parameter estimates to enhance the estimation process when the input-output data are sparse or non-informative, enhancing fault identifiability. We demonstrate the effectiveness of this method in simulations involving an Autonomous Surface Vehicle in both a path-following and a realistic collision avoidance scenario, underscoring its potential to enhance safety and reliability in critical applications.

Keywords: Fault diagnosis, Linear systems, Uncertain systems
Abstract: This paper extends an asymptotic active fault diagnosis (AFD) framework using a bank of observers to linear time-invariant (LTI) systems under the effect of hybrid bounded and Gaussian uncertainties. Each observer is designed to match a healthy/faulty actuator mode. Bounded uncertainties are represented using zonotopes, while Gaussian uncertainties are modeled as ellipsoids based on the specified confidence level. The output confidence domains are expressed as the Minkowski sum of a zonotope and an ellipsoid, i.e., an ellipsotope. At each step, the proposed AFD method designs an input and a group of observer gains such that output confidence domains estimated by observers at the next step keep away from each other as far as possible. By designing inputs and injecting them into the system and optimizing observer gains step by step, AFD is eventually achieved. At the end of this paper, the effectiveness of the proposed method is illustrated by examples.

Keywords: Fault detection and identification, Electrical power systems
Abstract: The increasing stress on the electric power distribution grid exposes the grid to different types of disturbances and faults. Recent technology enables electric vehicles (EVs) to act as a distributed energy storage source, so that the stress on the power grid is reduced. However, unexpected power grid changes may still lead to on-board-chargers (OBCs) damages on EVs. Thus, promptly detecting changes on the three-phase voltage signal coming from the grid is of paramount importance to adapt the grid to the new condition without affecting the OBCs. In this work, we propose a power grid fault detection method able to detect changes on the grid. The method is based on the estimation of the three-phase voltage signal from the grid using an Unscented Kalman Filter (UKF). The method is evaluated on data from a simulated power grid and from a real one, considering different power grid changes

Keywords: Fault detection and identification, Fault diagnosis
Abstract: Deformable mirrors are critical components in adaptive optics systems, particularly for high-precision applications. However, due to their complex and highly dynamic behavior, it remains challenging to develop accurate mathematical models for deformable mirrors. Additionally, deformable mirrors are prone to faults, such as material fatigue and sensor failures, making it essential to implement a robust health monitoring system capable of detecting faults in real time. This paper proposes a model-free health monitoring algorithm for deformable mirrors. The algorithm leverages transmissibility operators—mathematical models that describe the relationship between a system's outputs. By using measurements from sensors attached to the deformable mirror, we construct transmissibility relationships under healthy operating conditions. These relationships are then used, along with real-time sensor data, to generate virtual sensor measurements. The discrepancy between actual and virtual measurements serves as an indicator of potential faults. To demonstrate the effectiveness of the proposed algorithm, we apply it to a deformable mirror model simulated using COMSOL Multiphysics, incorporating scenarios involving microcracks and sensor failures.

Keywords: Fault detection and identification, Modeling, Aerospace
Abstract: A methodology for fault detection and isolation (FDI) of electrical faults in a three-phase permanent magnet synchronous motor (PMSM) is presented. The considered faults are stator winding and current sensor faults. The former consist of wire degradation (characterized by a rise in the corresponding phase resistance), turn-to-turn short circuits and phase-to-phase short circuits. The sensor faults consist of gain changes and offsets in phase current measurements. A multi-model adaptive estimator of fault parameters is developed by resorting to specific signal models. These models characterize phase current measurements (i.e. the signals) in the healthy and in each faulty operating mode. Such signal models do not require knowledge of the three-phase motor parameters. The severity of the fault is evaluated through the estimators that take the form of three dual extended Kalman filters (DEKFs) and one Kalman filter (KF). The proposed approach is validated by using synthetic data of phase current measurements. These data were generated by a digital twin of an electromechanical actuator (EMA) driven by a PMSM. This motor drive is subject to closed-loop speed control and includes realistic sensor noise. To test the robustness of the approach, its performance is evaluated with different levels of sensor noise.

Keywords: Nonlinear system theory, Autonomous robots, Robotics
Abstract: The aim of this paper is to provide a Derivative-free Kalman Filter (DKF) as an alternative to the classical Extended and Unscented Kalman filters, thus removing the numerical instability risks they pose once implemented on platforms with limited numerical precision. The DKF is designed for an Ackermann steered six-wheeled rover which has nonlinear, coupled longitudinal and lateral dynamics. Starting from the first principles of the two axle car, we derive the rover's mathematical model, which we then express through a flat-output representation. Using the resulting Brunovsky form, we provide both a position estimate (via the DKF) and a control module (a reference tracking variation). The model and its ancillary blocks are tested in simulation to validate the flat representation's consistency and the performance of the navigation and control actions.

Keywords: Nonlinear system theory, Concensus control and estimation, Neural networks
Abstract: We study the convergence of the nonlinear Krasnoselskij iteration x(k+1)=(1-θ)x(k) + θT(x(k)) in real vector spaces of finite dimension equipped with a p-norm, which is relevant for stability analysis and distributed computation in several discrete-time dynamical systems. % Specifically, we provide sufficient conditions for the convergence of the Krasnoselskij iteration, derived via implications between the strict pseudocontractivity of the operator T and the nonexpansiveness of (1 − θ)Id + θT. % Interestingly, it turns out that strict pseudocontractivity of T is necessary for the Euclidean norm (p = 2) only; not necessary for non-Euclidean norms (p ≠ 2); sufficient for any finite norm p ∈ (1, ∞); not sufficient for the taxi-cab norm (p = 1) and the supremum norm (p = ∞). % We numerically verify the above results in the context of recurrent neural networks and multi-agent systems with nonlinear Laplacian dynamics.

Keywords: Nonlinear system theory, Optimization algorithms
Abstract: The scaled relative graph (SRG) is a powerful graphical tool for analyzing the properties of operators, by mapping their graph onto the complex plane. In this work, we study the SRG of two classes of nonmonotone operators, namely the general class of semimonotone operators and a class of angle-bounded operators. In particular, we provide an analytical description of the SRG of these classes and show that membership of an operator to these classes can be verified through geometric containment of its SRG. To illustrate the importance of these results, we provide several examples in the context of electrical circuits. Most notably, we show that the Ebers-Moll transistor belongs to the class of angle-bounded operators and use this result to compute the response of a common-emitter amplifier using Chambolle-Pock, despite the underlying nonsmoothness and multi-valuedness, leveraging recent convergence results for this algorithm in the nonmonotone setting.

Keywords: Nonlinear system theory, Output feedback
Abstract: This paper extends the method of signal injection to accommodate nonperiodic excitation. The proposed extension builds on two key elements from the periodic case: (1) the decomposition of the system into low- and high-frequency components using higher-order averaging, and (2) the extraction of the virtual output through a demodulation process.

Keywords: Nonlinear system theory, Stability of nonlinear systems, Computational methods
Abstract: An analog of the describing function method is developed using square waves rather than sinusoids. Static nonlinearities map square waves to square waves, and their behavior is characterized by their response to square waves of varying amplitude – their amplitude response. The output of an LTI system to a square wave input is approximated by a square wave, to give an analog of the describing function. The classical describing function method for predicting oscillations in feedback interconnections is generalized to this square wave setting, and gives accurate predictions when oscillations are approximately square.

Keywords: Nonlinear system theory, Stability of nonlinear systems
Abstract: Scaled Relative Graphs (SRGs) provide a novel graphical frequency-domain method for the analysis of nonlinear systems. However, we show that the current SRG analysis suffers from a pitfall that limit its applicability in analyzing practical nonlinear systems. We overcome this pitfall by modifying the SRG of a linear time invariant operator, combining the SRG with the Nyquist criterion, and apply our result to Lur'e systems. We thereby obtain a generalization of the celebrated circle criterion, which deals with a broader class of nonlinearities, and provides (incremental) L_2-gain performance bounds.

Keywords: Stability of nonlinear systems, Network analysis and control, Modeling
Abstract: We propose an epidemic model for the spread of vector-borne diseases. The model, which is built extending the classical susceptible-infected-susceptible model, accounts for two populations -humans and vectors and for cross-contagion between the two species, whereby humans become infected upon interaction with carrier vectors, and vectors become carriers after interaction with infected humans. We formulate the model as a system of ordinary differential equations and leverage monotone systems theory to rigorously characterize the epidemic dynamics. Specifically, we characterize the global asymptotic behavior of the disease, determining conditions for quick eradication of the disease (i.e., for which all trajectories converge to a disease-free equilibrium), or convergence to a (unique) endemic equilibrium. Then, we incorporate two control actions: namely, vector control and incentives to adopt protection measures. Using the derived mathematical tools, we assess the impact of these two control actions and determine the optimal control policy.

Keywords: Sliding mode control, Stability of nonlinear systems, Robust control
Abstract: For many control systems the presence of perturbations may decrease performance, so that their rejection has been an area of active research for the last decades. Roughly speaking, there are two kind of perturbations, those which are matched, i.e. the control input can act directly on them, and those which are mismatched, i.e. the control input cannot act directly on them. While there are many solutions for the first case, the second one remains a challenge. This note improves a former result by the authors to avoid singularities in a Sliding Mode Control (SMC) approach which provides ultimate boundedness of the desired output with an ultimate bound that can be made arbitrarily small. Simulation results are provided to validate the introduced theory.

Keywords: Stability of nonlinear systems, Concensus control and estimation, Lyapunov methods
Abstract: The paper is devoted to the problem of stability analysis for homogeneous systems with sector nonlinearities. Such systems are relevant to a various applications including finite-time controls, observers and consensus protocols. The proposed sufficient stability conditions are presented in the form of linear matrix inequalities. Based on the developed stability analysis method a new average-consensus protocol is proposed for a multi-agent system with double integrator dynamics. The theoretical results are supported with a numerical example.

Keywords: Stability of nonlinear systems, LMI's/BMI's/SOS's, Neural networks
Abstract: This study investigates the absolute stability criteria based on the framework of integral quadratic constraint (IQC) for feedback systems with slope-restricted nonlinearities. In existing works, well-known absolute stability certificates expressed in the IQC-based linear matrix inequalities (LMIs) were derived, in which the input-to-output characteristics of the slope-restricted nonlinearities were captured through static O'Shea-Zames-Falb multipliers. However, since these certificates are only sufficient conditions, we cannot draw any conclusions about the absolute stability in the case where the LMIs are infeasible. In this paper, by taking advantage of the duality theory of LMIs, we derive a condition for systems to be not absolutely stable when the above-mentioned LMIs are infeasible. In particular, we can identify a destabilizing nonlinearity within the assumed class of slope-restricted nonlinearities as well as a non-zero equilibrium point of the resulting closed-loop system, by which the system is proved to be not absolutely stable. We demonstrate the soundness of our results by numerical examples.

Keywords: Stability of nonlinear systems, Lyapunov methods, Nonlinear system theory
Abstract: The Koopman operator provides an infinite-dimensional linear description of nonlinear dynamical systems that can be leveraged in the context of stability analysis. In particular, Lyapunov functions can be obtained in a systematic way via the eigenfunctions of the Koopman operator. However, these eigenfunctions are computed from finite-dimensional approximations, resulting in approximated Lyapunov functions that must be validated. In this paper, we provide theoretical error bounds on the approximation of the eigenfunctions of the Koopman operator in the case of analytic vector field and finite-dimensional approximation in polynomial subspaces. We leverage these results to assess the validity of Koopman-based Lyapunov functions and obtain an optimization-free inner approximation of the region of attraction of an equilibrium.

Keywords: Stability of nonlinear systems, Observers for nonlinear systems, Aerospace
Abstract: When inertial information is unavailable to locate the inertial reference frame, i.e., the environment is unknown, it is usually difficult to conduct pose tracking tasks for rigid bodies. To address this problem, based on the adopted pose observer, we can determine an estimated reference frame from the estimates. Then we express the desired pose trajectory with respect to the estimated frame and develop a gradient-based pose tracking law on SE(3) to deal with relative pose trajectory tracking. The dynamics of the combined observer-controller are analyzed, rigorously showing almost global asymptotic stability of the combination. Simulations are also conducted to show the effectiveness of our method.

Keywords: Distributed parameter systems
Abstract: This paper presents a reduced order compensator for a class of SISO infinite dimensional systems. The compensator order is equal to the dimension of the input and is made possible via the use of a functional observer. A full state feedback controller is first designed and subsequently a functional observer is set-up to estimate such functional. When the estimated functional is used as the control signal, the resulting system is shown to be stable and match the performance of a full state feedback. As an added compensator improvement, the sensor is allowed to move, rendering the output operator time varying. Such a time variation results in a differential Sylvester equation. The results are also extended to the case of a time varying disturbance operator, often modelling a moving disturbance. Numerical results are included to demonstrate the performance improvement of the functional observer when a sensor is allowed to move.

Keywords: Distributed parameter systems, Linear systems, Robust control
Abstract: In this paper, we investigate the state feedback control for the H^infty disturbance-attenuation problem in the context of general parabolic systems. The control mechanism employed is Neumann boundary control, which permits disturbances to enter the system through any channel. We employ the Galerkin approximation method, which generates a sequence of finite dimensional systems serving as proxies for the original infinite-dimensional system. Our key finding is that the solutions to the corresponding finite-dimensional algebraic Riccati equations converge in norm to the solution of the infinite-dimensional operator algebraic Riccati equation. Furthermore, we establish that the state feedback controls derived from these finite-dimensional algebraic Riccati equations are gamma-admissible for the original system, thereby validating the effectiveness of our approximation approach.

Keywords: Fluid flow systems, Reduced order modeling, Distributed control
Abstract: A second-order in time discretization of the artificial compression proper-orthogonal decomposition reduced-order model (POD-ROM) for control of thermally convective flows is presented. It is a velocity-pressure reduced- order model that does not require satisfaction of div-stability (Ladyzhenskaya-Babuska-Brezzi) condition for mixed reduced-order subspaces. Second-order in time artificial compression POD-ROM approximations of solutions of the optimality system are defined and error estimates are derived. Numerical experiments are performed to validate the accuracy and efficiency of the scheme in solving a boundary control problem involving mixed convection flow.

Keywords: Observers for linear systems, Distributed parameter systems, Delay systems
Abstract: A new finite-dimensional adaptive observer is proposed for a class of uncertain wave systems with delayed measurements. The observer is based on the modal decomposition approach and uses a classical persistent excitation condition to ensure practical exponential convergence of both states and parameters estimation for arbitrarily long delays.

Keywords: Delay systems, Distributed parameter systems, Robust control
Abstract: We examine approximations of the H∞-optimal controller arising from the mixed sensitivity minimization for a first order unstable time delay system, with first order weights. The optimal controller is known to have a special structure where a PI controller and an FIR filter form the finite dimensional and the infinite dimensional parts, respectively. We approximate the infinite dimensional part by fixed order rational transfer functions using various methods from the literature. We compare the H∞ approximation errors obtained with the following rational approximation methods for the FIR part of the controller: systune and ssest commands of Matlab, Hankel norm approximation (HNA), moment matching (MM), balanced reduction (balred), iterative rational Krylov algorithm (IRKA) and Pade approximation.

Keywords: Distributed parameter systems, Chemical process control, Predictive control for linear systems
Abstract: The model predictive control of an axial dispersion tubular reactor equipped with a recycle stream is presented. The intrinsic time delay imposed by the recycle stream, an often overlooked aspect in chemical engineering process control studies, is modeled as a transport PDE, leading to a boundary-controlled system of coupled parabolic and hyperbolic PDEs under Danckwerts boundary conditions, suitable for this reactor type. Considering the digital nature of modern controllers, a discrete-time linear model predictive controller is designed to stabilize the system, coupled with a Luenberger state estimator to address the controller's limited access to the system's full state. The need for model reduction through spatial approximation is eliminated by following a late lumping approach, while utilizing Cayley-Tustin time discretization method to preserve the continuous-time system's characteristics. The controller's effectiveness is demonstrated through numerical simulations, showcasing its capability to stabilize an unstable system while adhering to input constraints, having access merely to output measurements.

Keywords: Aerospace, Neural networks, Identification
Abstract: This paper investigates the problem of impact angle identification for an incoming aerial vehicle, which is essential for constructing an effective Kalman Filter (KF) and accurately estimating the vehicles state. The considered scenario is that a vehicle attempts to hit a stationary target with the Trajectory Shaping Guidance (TSG) law. A desired impact angle regression identification model from the perspective of the target is proposed based on a Gated Recurrent Unit (GRU) neural network. The input of the model is a period of available information, including the relative distance and relative angle between the vehicle and target, while the output is the regression identification result. To increase the offline training efficiency and improve the online identification accuracy, the Multiple-Model Mechanism (MMM) is introduced into the network. Simulation results demonstrate the advantage of the proposed model over a conventional network, and verify its application potential.

Keywords: Aerospace, Robust control, Optimal control
Abstract: This article addresses the planetary landing problem by considering uncertainties and leveraging the presence of a detection area where precise measurements are available. The flight consists of two distinct phases: the first phase, subject to a high level of uncertainties, and the second phase, during which the vehicle is feed- back controlled to ensure precise landing. We propose a deterministic optimal control method to plan the trajectory of the initial phase, aiming to minimize fuel consumption for the entire trajectory while satisfying a probabilistic constraint that ensures the vehicle reaches the detection zone with a specified threshold.

Keywords: Aerospace, Transportation systems, Optimization
Abstract: This paper presents for the first time an approach to minimize direct operational costs (DOC) for all-electric aircraft during the climb phase, introducing a time-varying cost index (CI). The CI is modeled as a dynamic parameter commanded by Air Traffic Control (ATC), allowing the aircraft to maintain a constant airspeed throughout the climb, while respecting the air traffic regulations. This paper also explores the implications of a time-varying CI on the determination of optimal airspeed and climbing time for all-electric aircraft. Additionally, it provides the necessary equations to calculate both the optimal climb airspeed and climb duration. The proposed methodology has been validated through a simulated scenario that reflects actual operational procedures. As a result, optimal values for climb airspeed, climbing time, and energy consumption have been established, paving the way for future applications of this methodology to advanced air mobility all-electric vehicles.

Keywords: Aerospace, Autonomous systems, UAV's
Abstract: This paper proposes a practical homing guidance law for impact angle control of a re-entry vehicle. Conventional guidance laws for re-entry vehicles has suffered from high complexity in implementation and degradation of their performance when there exist large initial errors. To address this issue, diveline vector is introduced for desired impact angle and used to derive the guidance command. Afterwards, the derived guidance command is incorporated with typical proportional navigation (PN) guidance command for homing to balance between the homing and diveline alignment. The proposed guidance law makes the vehicle align with the diveline vector first, and then home on the target along the diveline vector, working as mid-course guidance law and terminal guidance law simultaneously. Computer simulation shows that the proposed method has superior performance even when there are large initial range errors and impact angle errors.

Keywords: Aerospace, H2/H-infinity methods, Linear systems
Abstract: This paper presents the design and verification of an attitude control algorithm for a CubeSat whose primary mission is stellar observation. The main sources of pointing errors are environmental input disturbances, which must be attenuated to achieve precise pointing. The frequency characteristics of these disturbances were analyzed, and a relationship was established to estimate the resulting pointing error. By employing H∞ control, the disturbance properties are directly incorporated into the control problem formulation, resulting in a simplified controller tuning process. The designed controller was verified through linear analysis. Software-in-the-loop simulations using a high-fidelity simulator confirmed the results of the linear analysis. Additionally, a Monte Carlo simulation campaign, accounting for parametric uncertainties, was conducted and further validated the results.

Keywords: Aerospace, Transportation systems, Optimization
Abstract: This paper finds, for the first time, the optimal airspeed that maximizes the endurance of hybrid-electric aircraft in cruise flight. The optimal control problem is formulated using energy-depletion dynamics and solved using Pontryagin's Maximum Principle (PMP). The solutions show that the optimal cruise airspeed is the single positive real root of a quintic polynomial. One important result is the computation of the proportion of the aircraft's thrust stemming from electrical energy such that all fuel and electricity are consumed for maximum endurance. The paper shows that this value depends on the ratio between the initial electrical and the initial fuel energy. Furthermore, the results indicate that the maximum endurance flight time is also influenced by the cruising altitude. Simulations illustrate the proposed methodology.

Keywords: Modeling, Reduced order modeling, Emerging control applications
Abstract: We study the foreground-background separation problem for a video stream in the moment matching framework. We first show how a video stream can be represented as the steady-state output of a dynamical system. We then leverage this technique to reconstruct image sequences in a video stream. In addition, we show that foreground-background separation is a special case of this reconstruction algorithm. Finally, we highlight the computational benefits of the proposed method when compared to other methods that associate a sequence of images with a dynamical system.

Keywords: Modeling, Aerospace, Complex systems
Abstract: An accurate actuator model is essential to evaluate the performance of a control system during time domain simulation. Linearized motor transfer function based on closed-loop actuation system can often be found in the literature but these simplified models do not address the non-linear behaviors that exist in the real physical system. Such simplified models are often sufficient for evaluating the closed-loop system with a small demand of precision. However, the dynamic model of the canard actuation system, as the main actuator of the ammunition, must be represented with very high precision to obtain reliable results via numerical simulations. This paper details the modeling problem of a high-fidelity closed-loop actuation system with an inner current-loop, based on three phase induction motor and motor driver, as well as an outer-loop for controlling the angular position of the motor. The simulation results are then compared with hardware-in-the-loop test realized with motor and its driver in the loop in a full-scale wind-tunnel to qualify the numerical model and quantify the uncertainties, which will later be used to assess realistic performances.

Keywords: Modeling, Differential algebraic systems, Chemical process control
Abstract: We develop a mathematical model for dynamic simulation of an alkaline electrolyzer plant. The plant includes the stack, a water recirculation system and hydrogen storage with compressor. We model each component of the system with mass and energy balances. Our modeling strategy consists of a rigorous and systematic formulation using differential algebraic equations (DAE), along with a thermodynamic library that evaluates thermophysical properties. We perform a simulation with step power input. Dynamic modeling enables simulation and model-based optimization and control for optimal hydrogen production under varying operating conditions.

Keywords: Modeling, Differential algebraic systems, Chemical process control
Abstract: We provide a cyclone model for dynamical simulations in the pyro-process of cement production. The model is given as an index-1 differential-algebraic equation (DAE) model based on first engineering principle. Using a systematic approach, the model integrates cyclone geometry, thermo-physical aspects, stoichiometry and kinetics, mass and energy balances, and algebraic equations for volume and internal energy. The paper provides simulation results that fit expected dynamics. The cyclone model is part of an overall model for dynamical simulations of the pyro-process in a cement plant. This model can be used in the design of control and optimization systems to improve energy efficiency and reduce CO2 emission.

Keywords: Modeling, Machine learning, Maritime
Abstract: Accurate power prediction is crucial for optimizing fuel consumption models in maritime vessels, particularly as the industry shifts towards more sustainable energy solutions. Traditional reliance on fossil fuels often leads to inefficiencies in large-scale ship engines, prompting a growing interest in Liquefied Natural Gas (LNG) as a cleaner alternative for propulsion systems. This study presents a comparative analysis of three approaches to power prediction: physics-based, data-driven, and a hybrid multi-method framework. The proposed integrated approach combines Computational Fluid Dynamics (CFD), grounded in first-principle equations, with machine learning techniques to better capture the complexity of real-world maritime operations. A deep learning gating mechanism is employed to dynamically select the most accurate output from the ensemble of models, enhancing both reliability and precision in power estimation. The hybrid framework offers a robust and adaptable solution that contributes to improved fuel efficiency and the reduction of greenhouse gas emissions in the maritime sector.

Keywords: Modeling, Neural networks, Autonomous systems
Abstract: We study the problem of synthetic generation of samples of environmental features for autonomous vehicle navigation. These features are described by a spatiotemporally varying scalar field that we refer to as a threat field. The threat field is known to have some underlying dynamics subject to process noise. Some "real-world" data of observations of various threat fields are also available. The assumption is that the volume of "real-world" data is relatively small. The objective is to synthesize samples that are statistically similar to the data. The proposed solution is a generative artificial intelligence model that we refer to as a split variational recurrent neural network (S-VRNN). The S-VRNN merges the capabilities of a variational autoencoder, which is a widely used generative model, and a recurrent neural network, which is used to learn temporal dependencies in data. The main innovation in this work is that we split the latent space of the S-VRNN into two subspaces. The latent variables in one subspace are learned using the real-world data, whereas those in the other subspace are learned using the data as well as the known underlying system dynamics. Through numerical experiments we demonstrate that the proposed S-VRNN can synthesize data that are statistically similar to the training data even in the case of very small volume of real-world training data.

Keywords: Autonomous robots, Robotics, Optimization
Abstract: As a parametric motion representation, Bezier curves have significant applications in polynomial trajectory optimization for safe and smooth motion planning of various robotic systems, including flying drones, autonomous vehicles, and robotic manipulators. An essential component of Bezier curve optimization is the optimization objective, as it significantly influences the resulting robot motion. Standard physical optimization objectives, such as minimizing total velocity, acceleration, jerk, and snap, are known to yield quadratic optimization of Bezier curve control points. In this paper, we present a unifying graph-theoretic perspective for defining and understanding Bezier curve optimization objectives using a consensus distance of Bezier control points derived based on their interaction graph Laplacian. In addition to demonstrating how standard physical optimization objectives define a consensus distance between Bezier control points, we also introduce geometric and statistical optimization objectives as alternative consensus distances, constructed using finite differencing and differential variance. To compare these optimization objectives, we apply Bezier curve optimization over convex polygonal safe corridors automatically constructed around a maximal-clearance minimal-length reference path. We provide an explicit analytical formulation for quadratic optimization of Bezier curves using Bezier matrix operations. We conclude that the norm and variance of the finite differences of Bezier control points lead to simpler and more intuitive interaction graphs and optimization objectives compared to Bezier derivative norms, despite having similar robot motion profiles.

Keywords: Autonomous robots, Coverage control, Cooperative control
Abstract: Efficient image sampling from diverse viewing angles is crucial for high-quality 3D map reconstruction, and coverage control provides a promising solution to this mission. Moreover, with the recent advancements in real-time 3D reconstruction algorithms, it is now possible to iteratively reconstruct 3D maps in real time, providing immediate map feedback to robot motion control. In this paper, we propose a novel coordinated image sampling algorithm that leverages realtime map feedback to enhance the quality of the reconstructed 3D model. We first formulate the problem as an angle-aware coverage control problem, where images are captured from multiple angles across the field of interest by drones. These images are processed in real time by so-called NeuralRecon to generate an evolving 3D mesh of the environment. Mesh changes across the field serve as feedback to update the importance index of the coverage control as the map evolves. We then design a QP-based controller to certify a sampling performance by constraining the decay rate of the objective function. Simulations in Unity and ROS2 demonstrate that our feedback-driven approach outperforms the conventional method without map feedback, resulting in a more complete and accurate 3D map. Project page: https://htnk-lab.github.io/map_feedback_coverage/.

Keywords: Autonomous robots, Robotics, Randomized algorithms
Abstract: Safe, smooth, and optimal motion planning for nonholonomically constrained mobile robots and autonomous vehicles is essential for achieving reliable, seamless, and efficient autonomy in the logistics, mobility, and service industries. In many such application settings, nonholonomic robots, like unicycles with restricted motion, require precise planning and control of both translational and orientational motion to approach specific locations in a designated orientation, such as for approaching charging, parking, and loading areas. In this paper, we introduce a new dual-headway unicycle pose control method by leveraging an adaptively placed headway point in front of the unicycle pose and a tailway point behind the goal pose. In summary, the unicycle robot continuously follows its headway point, which chases the tailway point of the goal pose and the asymptotic motion of the tailway point towards the goal position guides the unicycle robot to approach the goal location with the correct orientation. The geometric construction of dual-headway unicycle pose control enables an explicit convex feedback motion prediction bound on the closed-loop unicycle motion trajectory for safety verification. We present an application of dual-headway unicycle control and motion prediction for optimal sampling-based motion planning around obstacles. In numerical simulations, we show that optimal unicycle motion planning using dual-headway translation and orientation distances significantly outperforms Euclidean translation and cosine orientation distances in generating smooth motion with minimal travel and turning effort.

Keywords: Autonomous robots, Optimal control, Robotics
Abstract: This paper presents a novel problem of creating and regulating localization and navigation illusions considering two agents: a receiver and a producer. A receiver is moving on a plane localizing itself using the intensity of signals from three known towers observed at its position. Based on this position estimate, it follows a simple policy to reach its goal. The key idea is that a producer alters the signal intensities to alter the position estimate of the receiver while ensuring it reaches a different destination with the belief that it reached its goal. We provide a precise mathematical formulation of this problem and show that it allows standard techniques from control theory to be applied to generate localization and navigation illusions that result in a desired receiver behavior.

Keywords: Autonomous robots, Automotive, Agents and autonomous systems
Abstract: Localization is a crucial aspect of every autonomous driving vehicle, as determining its position in the navigation space enables the vehicle to safely plan its motion and interact with the environment. State-of-the-art approaches that rely on the Global Navigation Satellite System (GNSS) suffer from poor reliability in urban contexts. This issue can be overcome with Simultaneous Localization And Mapping (SLAM) methods. Generating and maintaining maps of large dimensions using these methods is a high-resource consuming task. This has motivated many to develop localization methods based on map databases provided by third-party sources. This paper presents a localization approach based on the OpenStreetMap (OSM) database. In particular, a local perception map generated from LiDAR and Camera observations is aligned to a graph representing an approximation of the road structure. Our method is validated through a comparison with a state-of-the-art approach.

Keywords: Autonomous robots, Optimization algorithms, Agents and autonomous systems
Abstract: Path planning is one of the most challenging aspects for the development of self-driving vehicles, due to the complexity and variety of contexts of urban and suburban environments. Prior knowledge of the navigation environment, in the form of highly accurate maps, is therefore often exploited in order to simplify this task. The need for detailed a-priori maps represents one of the major limiting factors for the widespread adoption of autonomous vehicles beyond small areas, for which managing and updating such maps is feasible in terms of resources required. For this reason, in the last years a growing interest has been put towards developing navigation approaches that are based on topological maps databases, for which many providers are already available. These types of maps are however less accurate, therefore requiring the implementation of a local planner that combines this information with online observations from a perception module, in order to generate trajectories that traverse the road safely. This paper presents a two-phase planning algorithm that exploits the OpenStreetMap database and an on-board perception system, and considers the vehicle's handling characteristics to generate feasible trajectories. The approach is validated with an experimental session on public roads open to traffic.

Keywords: Distributed estimation over sensor nets, Concensus control and estimation, Cooperative control
Abstract: Designing observers for linear systems with both known and unknown inputs is an important problem in several research contexts, for example, fault diagnosis and fault-tolerant control, and cyber-secure control systems, and presents significant challenges in distributed state estimation due to the limited sensing capabilities of individual nodes. Existing methods typically impose an individual input-to-output rank condition on each estimator node, which severely restricts applicability in practical applications. This paper presents a novel distributed unknown-input observer design scheme based on a geometric approach under much weaker assumptions than the ones available in the literature. By leveraging the properties of the (C, A)-invariant (conditioned invariant) subspace at each node, our methodology aims at reconstructing portions of the system state that remain unaffected by local unknown inputs, while integrating these estimates via a network-based information exchange. A case study on vehicle platoon control shows the effectiveness of the proposed approach.

Keywords: Communication networks, Filtering, Sensor and signal fusion
Abstract: In this paper, we study the design of an optimal transmission policy for remote state estimation over packet-dropping wireless channels with imperfect channel state information. A smart sensor uses a Kalman filter to estimate the system state and transmits its information to a remote estimator. Our objective is to minimize the state estimation error and energy consumption by deciding whether to transmit new information or retransmit previously failed packets. To balance the trade-off between information freshness and reliability, the sensor applies a hybrid automatic repeat request protocol. We formulate this problem as a finite horizon partially observable Markov decision process with an augmented state-space that incorporates both the age of information and the unknown channel state. By defining an information state, we derive the dynamic programming equations for evaluating the optimal policy. This transmission policy is computed numerically using the point-based value iteration algorithm.

Keywords: Distributed cooperative control over networks, Cooperative autonomous systems, Predictive control for nonlinear systems
Abstract: Communication delays in Connected and Automated Vehicle (CAV) networks significantly impact decentralized optimization-based coordination, increasing risks of collisions and degrading system performance. Existing methods are limited by real-time computational challenges, vulnerability to outdated data, scalability constraints, and difficulties in managing uncertainties. This paper presents a Stochastic Decentralized Model Predictive Control (SDMPC) framework to mitigate the adverse effects of communication delays by incorporating a novel stochastic approximation method for modeling uncertainties. Our approach provides a tight probabilistic bound on safety constraints, ensuring accurate trajectory predictions and improved coordination. Simulation results show that the proposed SDMPC framework reduces the average trajectory deviation and lowers collision risks compared to conventional methods under various communication latency conditions. These improvements make SDMPC an effective solution for large-scale CAV networks, enhancing both safety and efficiency.

Keywords: Decentralized control, Linear systems, Autonomous systems
Abstract: This study investigates the coordination of vehicle platoon systems across diverse Information Flow Topologies (IFTs), addressing associated challenges and opportunities. Key factors, including node dynamics, decentralized controllers, IFTs, and geometric formations, are considered. The analysis focuses on evaluating the performance of vehicle platoon systems under both unidirectional and bidirectional IFTs, assuming a spanning tree structure. A stability analysis is conducted to assess system behavior. Furthermore, the impact of introducing new directed links into existing IFTs on the eigenvalues of the Information Flow Matrix (IFM) is theoretically examined. The results demonstrate that the eigenvalues of the IFM with added edges are greater than or equal to those of the initial IFT. This increase in eigenvalues accelerates convergence, indicating that strategic modifications to IFTs can significantly enhance the performance of vehicle platoon systems.

Keywords: Cooperative control, Distributed control, UAV's
Abstract: This paper investigates the cooperative guidance problem for unmanned aerial vehicles (UAVs) to simultaneously intercept a maneuvering target with constrained relative geometry. And a distributed optimal spatial-temporal cooperative pursuit strategy is proposed to address this issue. The impact angles and times of each UAV is first analytically predicted based on augmented ideal proportional navigation (AIPN) guidance. Then an analytical optimal distributed consensus protocol is utilized to coordinate the UAVs to achieve synchronized impact times and specified relative intercept geometry. The main benefit of the proposed approach lies in its optimal performance in a distributed system. Hence it consumes less control energy than the existing distributed cooperative guidance laws. Comparative simulations are performed to demonstrate the energy efficiency of the proposed approach.

Keywords: Distributed control, Large-scale systems, Optimal control
Abstract: In this work, we propose a distributed Linear-Quadratic-Gaussian (LQG) control scheme featuring event-triggered communication in order to reduce the communication load in the distributed setup. A consensus-based distributed state estimator is used, where only the nodes' local state estimates need to be transmitted to neighboring nodes through the communication network at event instants. Then, each actuator node computes its own control input based on its estimate. We show that an advantageous trade-off can be found between the communication load of the setup and the control performance, achieving a similar control cost as the optimal centralized LQG controller in a distributed fashion, while also significantly reducing the communication load. The proposal is validated through formal analysis and simulation experiments.

Keywords: Automotive, Emerging control applications, Optimization algorithms
Abstract: This work presents the first application of the recently developed Twin-in-the-Loop Control (TiL-C) approach to a disturbance rejection problem, focusing on the regulation of an active suspension system. By evaluating mismatches between the digital twin and the actual vehicle, the study identifies limitations in TiL-C and proposes new compensation strategies to address the simulation-to-reality gap in controller tuning during the end-of-line phase. The findings demonstrate TiL-C's potential to connect digital and physical systems, enabling broader applications of TiL-C across multiple fields.

Keywords: Automotive, Identification, Cooperative control
Abstract: Skill evaluation of driving maneuvers can lead to new personalization concepts in Advanced Driver Assistance Systems. An important intermediate step in this direction is understanding the multitude of the driver's intentions. We show at the example of parking maneuvers from a fixed parking scenario, how the parking skill can be assessed using the concept of pareto dominance. A skill measure is presented to rate parking maneuvers relative to the maneuvers in a given measurement set. The practical relevance of the skill measure is underlined by its correlation to the subjective evaluation of parking maneuvers.

Keywords: Automotive, Machine learning
Abstract: A key question that arises when dealing with sensors is how to process the available signals to obtain valuable information for monitoring and control, with limited amount of memory and computing power. In this paper, we build a vehicle speed black-box estimation method based on accelerometer data only. The key feature allowing this is the fact that the architecture of the overall algorithm is constructed by looking also at the physics of the problem. As a second contribution, we discuss how the obtained model can be further simplified via frequency-based considerations. Experimental tests carried out on a real vehicle show that the developed estimators are both effective and robust to varying environmental conditions. Specifically, the rms of the estimation error is never greater than 8 km=h.

Keywords: Automotive, Robust control
Abstract: This paper presents a robust H-infinity control design method that incorporates learning feature. The goal of the work is to improve the performance level of the robust controller, which is achieved through reinforcement learning (RL) process. The contribution of the presented method is that the design of the H-infinity controller and the RL-based controller is structured in a joint optimization. This leads to an iterative design for the controllers. The developed design method is applied to the problem of minimizing lap time in vehicle control design. The presented simulation-based analysis shows that the proposed method is able to provide improved performance level, i.e., reduced lap time, compared to the robust control or the independent design process. The outcome of the joint design is that the lap time achieved with the H-infinity controller approaches the lap time achievable with the RL process.

Keywords: Automotive, Predictive control for linear systems, Transportation systems
Abstract: This paper presents an eco-driving assistance system based on model predictive control (MPC) with variable sampling within the prediction horizon, such as quadratic point sampling (QPS) using a quadratic time-warping function and dynamic point sampling (DPS) where the Douglas-Peucker algorithm is used to dynamically select relevant sampling points, and provides comparison to uniform point sampling (UPS). The controllers aim to track the energy-optimal velocity profile derived from dynamic programming (DP), which incorporates long-term knowledge of speed limits and road slope, while maintaining a safe distance from a preceding vehicle in a car-following scenario, taking into account short-term safety-relevant knowledge. Variable sampling is used to consider the short- and long-term objectives simultaneously, promoting energy efficiency and safety and providing real-time capability. The controllers were evaluated in MATLAB simulations using the model of a Renault ZOE Q90, a compact battery electric vehicle (BEV), with scenarios both including and excluding knowledge of the preceding vehicle's velocity trajectory. The results show that variable sampling strategies (QPS and DPS) offer greater efficiency than uniform point sampling, especially in complex traffic scenarios, leading to improved energy savings while maintaining safety.

Keywords: Automotive, Constrained control
Abstract: This article addresses the challenges associated with the safety and performance of autonomous vehicles. Passenger safety and comfort are enhanced through the optimization of advanced driver assistance systems (ADAS), such as adaptive cruise control (ACC). A new longitudinal controller has been developed using Control Lyapunov Functions (CLF) and high-order Control Barrier Functions (CBF). This controller imposes strict safety constraints and integrates Lyapunov functions for an efficient adaptive cruise control system. The proposed CLF-CBF methodology significantly improves the adaptability and efficiency of vehicle control systems, particularly in speed regulation and stop-and-go scenarios. The approach has been validated through numerical simulations, demonstrating its effectiveness in achieving safety and control objectives.

Keywords: Optimization, Optimization algorithms, Robust control
Abstract: In this paper, we develop an online optimization algorithm with integral action for solving online optimization problems characterized by quadratic cost functions with a linearly varying optimal point. The proposed algorithm can be formulated as an uncertain linear feedback system incorporating a discrete time double integrator. By leveraging established results on the optimal gain margin, we demonstrate that the proposed algorithm achieves the optimal convergence rate compared to other methods including a discrete time double integrator applied to quadratic cost functions in finite-dimensional spaces.

Keywords: Optimization, Optimization algorithms, Signal processing
Abstract: Kalman filter, the gold standard of state estimation, has been used widely in many control theory, machine learning, and signal processing applications. In Kalman filter, the minimum variance state estimation is computed by combining the prior estimation and the output of a noise contaminated measurement. A more informative measurement can improve the precision and confidence of the posterior estimation. Finding the optimal experiment for Kalman filter is the topic of ongoing research. In this article, we introduce a new framework for computing the optimal measurement matrix via connecting the theory of Kalman filter to the theory of principal components. Through a couple of examples we show the performance of the algorithm in using available resources effectively by improving the speed and precision of a high-dimensional data acquisition system.

Keywords: Optimization algorithms, Behavioural systems, Decentralized control
Abstract: The paper introduces Split-as-a-Pro, a control framework that integrates behavioral systems theory, operator splitting methods, and alternating projection algorithms. The framework reduces dynamic optimization problems — arising in both control and estimation — to efficient projection computations. Split-as-a-Pro builds on a non-parametric formulation that exploits system structure to separate dynamic constraints imposed by individual subsystems from external ones — such as interconnection constraints and input/output constraints. This enables the use of arbitrary system representations, as long as the associated projection is efficiently computable, thereby enhancing scalability and compatibility with gray-box modeling. We demonstrate the effectiveness of Split-as-a-Pro by developing a distributed algorithm for solving finite-horizon linear quadratic control problems and illustrate its use in predictive control. Our numerical case studies show that algorithms obtained using Split-as-a-Pro significantly outperform their centralized counterparts in runtime and scalability across various standard graph topologies, while seamlessly leveraging both model-based and data-driven system representations.

Keywords: Optimization algorithms, Computer aided control design, Energy systems
Abstract: Recent years have seen a growing interest in airborne wind energy systems (AWESs), which pursue higher electricity generation by harvesting faster winds at higher altitudes. Particular attention has been given to a single aircraft that flies crosswind, reeling out a tether wound around a winch connected to a ground-based generator. The research has been strongly focussed on tackling the control problem, both optimal (trajectory generation/path planning) and feedback, with the development of dedicated optimization and tracking approaches respectively. However, the optimal design of these systems has been studied less although the design depends on the control and vice versa. This issue can be attributed to the strongly dynamic nature of the mission profile, which inhibits conventional static design routines. Therefore, this paper examines the concurrent optimal control and design of a AWES. To realize this multidisciplinary design problem a nested optimization problem is formulated and solved to discover integrated solutions that surpass those attainable through a decoupled approach. The nested formulation permits using dedicated trajectory optimization routines for AWES in the interior loop, such as the penality-based interior point homotopy (PIPH) approach, and data-efficient design optimization routines, such as fail-safe constrained Bayesian optimization, in the outer loop. The result is an optimal design and trajectory for an airborne wind energy system that outperforms a baseline design in power generation by 60%.

Keywords: Optimization algorithms, Observers for nonlinear systems, Optimization
Abstract: This paper explores the application of Newton-based methods for solving the nonlinear optimization problem typical of moving horizon estimation (MHE). Specifically, we examine the performance of Newton, quasi-Newton, and Gauss-Newton methods, with a focus on their computational efficiency and convergence properties. Through simulation results, we illustrate the difference in performance across different scenarios, providing insight into the strengths and limitations of each method in the context of fast MHE.

Keywords: Optimization algorithms, Optimization, Output regulation
Abstract: Proximal gradient methods are popular in sparse optimization as they are straightforward to implement. Nevertheless, they achieve biased solutions, requiring many iterations to converge. This work addresses these issues through a suitable feedback control of the algorithm's hyperparameter. Specifically, by designing an integral control that does not substantially impact the computational complexity, we can reach an unbiased solution in a reasonable number of iterations. In the paper, we develop and analyze the convergence of the proposed approach for strongly-convex problems. Moreover, numerical simulations validate and extend the theoretical results to the non-strongly convex framework.

Keywords: Intelligent systems, Optimization, Hybrid systems
Abstract: The definition and key concepts of hybrid learning are introduced. The utilization of both continuous dynamics, used to train a collection of learning models, and discrete dynamics, used to describe a structural or logical relationship between them, is motivated in the context of explainability and robustness. The role of hybrid learning in inference and control and its connections to neurosymbolic artificial intelligence is introduced. Finally, a preview of the main points discussed in the tutorial is given.

Keywords: Intelligent systems, Identification for hybrid systems, Optimization
Abstract: This tutorial explores the fundamentals of hybrid learning, highlighting recent advancements in homotopy optimization methods as a foundation for real-time data-driven inference and control. It introduces the Online Deterministic Annealing (ODA) method, a homotopy optimization technique that estimates the modes of a hybrid model through a sequence of bifurcation phenomena. This enables real-time training of hybrid learning models using adaptive two-timescale stochastic approximation algorithms. The tutorial underscores the significance of hybrid learning as a structural framework for analyzing robustness, explainability, and resource efficiency in data-driven inference and control. Practical applications in hybrid system identification, memory-efficient reinforcement learning, and learning-based control are also discussed.

Keywords: Switched systems, Game theoretical methods, Identification
Abstract: We highlight the emergence of hybrid phenomena in security applications involving cyber-physical systems. Leveraging results from differential games, we apply online deterministic annealing on a space of potential attacking behaviors that can inform the defender of the system about both the different types of attackers on the environment and their distribution. We further showcase how the complexity of switching on dynamical systems can be employed as a defensive tool on its own merit. Towards this, we develop a probabilistic switching feedback control mechanism that deceives an adversarial identifier. Our studies underscore the plethora of opportunities that hybrid system methods present for security analysts and defense experts in cyber-physical systems; as detection approaches as well as design techniques.

Keywords: Hybrid systems, Machine learning, Agents and autonomous systems
Abstract: We consider inference, control and decision making in single-agent or multi-agent dynamical systems, where the dynamical systems are hybrid automata, and the metrics and specifications involve hybrid variables, i.e. both Boolean and numerical variables. We first describe the great variety of applications that lead to such problems including autonomous cars and large scale smart robo-cab systems, sensor networks, to next generation hybrid terrestrial and non-terrestrial networks, smart energy systems, medical diagnosis problems, smart manufacturing systems, scheduling in modern semiconductor manufacturing foundries, network security and trust. The complexity of such autonomous systems necessitates the efficient integration of domain knowledge based methods with data driven methods, i.e. machine learning (ML) and artificial intelligence methods. We identify several key challenges in these problems: achieving high performance with limited data samples, unsupervised and supervised ML with hybrid variables, multi-criteria based AI, risk and performance duality and its relationship to efficient sampling and large deviations theory, novel model predictive problems with hybrid systems and on graphs, efficient integration of domain specific LLMs with Knowledge Graphs (KGs). We demonstrate how trade-off analysis via multi-objective optimization, with hybrid variables, can play an important role in solving such problems. We describe such methods and algorithms as examples of homotopy optimization with hybrid variables. We describe how the resulting algorithms relate closely to Neurosymbolic AI.

Keywords: Machine learning, Neural networks
Abstract: In contemporary engineering and scientific research, the interplay between automatic control and machine learning has become increasingly significant. This report explores two key aspects of this relationship: the application of machine learning techniques to enhance automatic control systems and the use of automatic control principles to improve machine learning algorithms. Firstly, we discuss how machine learning can be leveraged to optimize control strategies in complex systems, enabling adaptive and intelligent responses to dynamic environments. Techniques such as reinforcement learning and neural networks are examined for their ability to learn from data, resulting in more efficient control mechanisms that can handle uncertainty and nonlinearity. Secondly, we investigate how principles of automatic control can be applied to refine machine learning processes. Concepts such as feedback control can be utilized to stabilize learning algorithms, reduce overfitting, and ensure convergence in various machine learning applications. This dual perspective highlights the mutual benefits and synergies that arise from integrating these two fields. Through case studies and examples, we demonstrate the transformative potential of combining machine learning and automatic control, paving the way for advances in robotics, autonomous systems, and smart technologies. Ultimately, this report aims to provide insights into the future directions of research and the practical implications of merging these two domains.

Keywords: Hybrid systems, Mechatronics
Abstract: Control-oriented models of mechatronic systems often reveal, after suitable coordinate transformations, an underlying linear time-invariant (LTI) continuous-time structure. This LTI behavior is typically interrupted by resetting actions, which may be intrinsic to the system or available as design elements for the control engineer. These resets introduce nonlinear effects that create powerful degrees of freedom in the resulting hybrid motion.

This talk will explore how a variety of such systems -- despite their apparent differences -- share this common LTI flow combined with reset-induced discontinuities. We will illustrate how hybrid Lyapunov theory, combined with timers and logical states, provides a systematic approach to analyzing stability and convergence. The discussion will emphasize how recognizing and exploiting the LTI structure within the hybrid dynamics can open up effective leads for feedback design.

Keywords: Fault tolerant systems, Stability of nonlinear systems, Uncertain systems
Abstract: In this work, we enhance the robustness of the prescribed performance control (PPC) methodology in the presence of inelastic actuator faults. PPC was originally developed to impose pre-specified (user-defined) bounds on the output tracking error for both transient and steady-state behavior. The latter control objective is achieved when considering ideal actuator operation. In the presence of actuator faults, however, PPC proves incapable of guaranteeing closed-loop system stability as possible violation of the performance bounds leads directly to controller instability. To this end, we propose a robust modification on the convectional PPC design to effectively handle the error in case of performance bound violation with guaranteed stability, while achieving error recovery back to the prescribed region in finite time once the fault is resolved. Additionally, the proposed controller does not rely on fault-detection mechanisms keeping low the complexity of the control scheme. The theoretical results are verified via simulation studies.

Keywords: Fault tolerant systems, System reconfiguration, Feedback linearization
Abstract: Most of the existing approaches to fault tolerant control (FTC) of UAVs utilize the gyroscopic effect to stabilize a small and light weight quadrotor UAV with complete loss of motor(s), which are not suitable for large and heavy quadrotor UAVs as the quadrotors are required to spin at very high speed after one or more motor failures to maintain stability. This paper studies a reversible thrust-based active fault tolerant control approach for large and heavy quadrotor UAVs against any single-motor failure where the spin rate is low and the center of gravity (CG) may not be at the geometric center of quadrotor structure. This approach is based on nonlinear dynamic inversion and incremental control allocation using the negative thrust from reversed motor-propeller rotation. The flight experiment verifies the effectiveness of the proposed method. The demonstration video can be accessed via https://www.youtube.com/watch?v=Tcfe0o1SBr8.

Keywords: Fault detection and identification, IVHS, Fault diagnosis
Abstract: Modern vehicles have evolved from mechanical systems to complex and connected ones. While this development has improved their efficiency, it also brings new potential risks, particularly cyber-attacks. The vehicle's lateral dynamics are crucial for maintaining stability and control during turns and maneuvers, making them a key focus of research. However, only a few recent studies have specifically investigated the security of lateral dynamics. This paper explores the potential for zero dynamics attacks on the vehicle's lateral dynamics, where the attacker can remain undetected by leaving no trace on the system's outputs. Three scenarios are studied: when the output includes yaw rate, lateral acceleration, and their combination. These two critical measurements of a vehicle's lateral motion are accessible through the inertial measurement units (IMU) in every vehicle. For each scenario, the impact of zero dynamics attacks on system performance is analyzed and illustrated through simulations. Finally, the paper provides recommendations for securing vehicles' lateral dynamics against such attacks.

Keywords: Fault detection and identification, Machine learning, Manufacturing processes
Abstract: In modern industrial production lines, ensuring product quality, customer satisfaction and minimizing production costs is of primary importance. Many learning techniques to address the issue of fault detection require training on labelled databases with a large number of anomalous audio samples that, however, are difficult or impossible to obtain. Furthermore, understanding which audio features are really crucial for anomaly detection is non-trivial. The article presents a comparative analysis of three unsupervised machine learning techniques based on the analysis of audio files/features, suitable to the case where a significant number of anomalies is not available; and a strategy for isolating audio features that are really important for anomaly detection. Experimental results show that the technique based on the isolation of the correct audio features is better than brute-force techniques.

Keywords: Fault detection and identification, Statistical learning, Stochastic filtering
Abstract: In this work we introduce a novel adaptive anomaly detection framework specifically designed for monitoring sequential random finite set (RFS) observations. Our approach effectively distinguishes between in-control data (normal) and out-of-control data (anomalies) by detecting deviations from the expected statistical behavior of the process. The primary contributions of this study include the development of an innovative RFS-based framework that not only learns the normal behavior of the data-generating process online but also dynamically adapts to behavioral shifts to accurately identify abnormal point patterns. To achieve this, we introduce a new class of RFS-based posterior distributions, named Power Discounting Posteriors (PD), which facilitate adaptation to systematic changes in data while enabling anomaly detection of point pattern data through a novel predictive posterior density function. The effectiveness of the proposed approach is demonstrated by extensive qualitative and quantitative simulation experiments.

Keywords: Energy systems, Fault detection and identification, Fault diagnosis
Abstract: Advancements in the wind turbine technology have led to a rapid expansion in the wind energy sector placing it as one of the crucial components of the global renewable energy sectors. Wind energy farms include onshore and offshore farms. While onshore farms are cost-effective and accessible, offshore farms benefit from higher and more consistent wind speeds, although they operate in harsher conditions that increase the risk of faults. Health monitoring of wind turbines is thus essential to ensure safe operation of onshore and offshore wind turbines and to prevent any failures that can cause damage to the system and the environment. In this paper, we introduce a health monitoring algorithm for wind turbines that does not require knowledge of a model of the wind turbine or the input acting on it (i.e. wind speed and direction). The proposed algorithm only requires measurements from sensors, which are available during the wind turbine operation.

Keywords: Observers for linear systems, Distributed estimation over sensor nets, Concensus control and estimation
Abstract: In this work, we propose a novel distributed estimation scheme for discrete-time linear time-invariant (LTI) systems. This scheme achieves finite-time convergence without introducing high gain. Each local observer can reconstruct the state of the global system by combining its local measurement with the information shared among the communication network. The scheme relies on the application of cross-agent communication technology which, while introducing potential communication delays, eliminates the need for centralized redesign when the network topology changes. The latter significantly benefits the flexibility of the proposed scheme, especially in the context of a time-varying communication topology. The numerical example validates the effectiveness of the proposed design and its robustness against communication delays.

Keywords: Observers for linear systems, LMI's/BMI's/SOS's, Robust control
Abstract: Interest in assistive robots for rehabilitation purposes has raised due to the increase of people affected by motor impairment. An important challenge in this context is the interaction between user and robot to realize a cooperative task, where knowing the participation of the patient is crucial to the rehabilitation protocol. A rising topic is the knowledge of human contribution for personalized assistance, as the current range of solutions include cumbersome measurement devices, therapist-on-the-loop approach and sensor-less estimation based on human movement. This proposal allows estimating the human contribution in a rehabilitation task, dealing with unexpected abrupt forces, i.e., a disease such as Parkinson. In addition, a robust LMI-based computed torque controller is tested and validated with an OpenSim model.

Keywords: Observers for linear systems, Uncertain systems
Abstract: This paper introduces polytopic observer designs for discrete- and continuous-time linear systems with bounded uncertainties. In particular, by noting that polytopes are equivalent to constrained zonotopes with intervals in their generator spaces, we propose two choices of generators to fix the order of the polytopic observer. Moreover, this observation enables us to leverage existing interval observers to find interval estimates in the augmented generator space before projecting them onto the original state space as polytopes/constrained zonotopes. Further, we prove that the polytopic observers are at least as good as interval observers for the same uncertain linear system in terms of the volumes of their set estimates and the error system gains. As a side contribution, we also introduce a more computationally efficient approach to obtain interval observer gains. Finally, we demonstrate and discuss the effectiveness of the proposed approach on a broad range of examples.

Keywords: Observers for linear systems, Control over networks
Abstract: This paper proposes a robust observer strategy for networked control systems affected by time-varying delays and data loss. Using a switched Lyapunov function based approach, the observer not only reduces the impact of delays by one sampling period but also compensates for data loss from the sensor to the controller by providing state estimates even when data is unavailable, enabling continuous control signal computation. Furthermore, the design preserves the separation principle, allowing independent design of observer and controller while maintaining stability. Simulation results demonstrate the effectiveness and robustness of the proposed approach in handling both delays and data loss.

Keywords: Observers for linear systems, Uncertain systems, Optimization
Abstract: This paper proposes a new cost criterion to enhance the precision of a zonotopic state estimator for discrete-time descriptor linear systems. Originally, the algorithm solves a minimum-trace problem involving zonotopes, whose evolution is given by an interval observer structure containing extra design matrices, called degrees of freedom. Although the minimization of trace yields explicit solutions, it does not necessarily imply minimization of volume, and thereby, the precision of the output zonotope cannot be improved effectively. The volume measure for zonotopes is computationally expensive and, when used as cost criterion, implies nonlinear optimization problems. To address these issues, we propose here a minimum-radius criterion where the smallest box enclosing the output zonotope is minimized. The resulting optimization problem is nonlinear, but its convexity is exploited to yield an equivalent linear program, whose complexity order is derived. The effectiveness of our approach is illustrated over two numerical examples, with one of them involving an interval method here modified to descriptor systems.

Keywords: Distributed estimation over sensor nets, Observers for linear systems, Linear systems
Abstract: In this paper we consider the problem of distributed estimation of the state of a discrete-time, linear and time-invariant (LTI) state space model affected by disturbances. We assume that there is a connected network of sensors having access to some output measurements as well as to part of the control inputs applied to the system. Such sensors exchange information with the goal of achieving consensus and providing an asymptotically correct estimate of the original system state. Necessary and sufficient conditions for the existence of distributed unknown input observers with augmented states that achieve both goals are derived. The problem solution exploits the theory of decentralized output feedback control, thus making it possible to inherit the algorithms available for the solution of that problem.

Keywords: Stochastic control, Optimal control, Uncertain systems
Abstract: This paper proposes a local solution method tailored to stochastic optimal control problems with Gaussian process (GP) representation of the dynamics leaning on the stochastic dynamic programming (DP) approach. We explore two methods—Fourier-Hermite DP (FHDP) and its recent extension, Fourier-Hermite Probabilistic DP (FHPDP)—for incorporating GP-based model learning. Compared to other model learning techniques, GP-based model learning explicitly quantifies model uncertainty and mitigates the effects of structural model errors. These Fourier-Hermite methods provide derivative-free versions of the differential dynamic programming (DDP) method through iterative backward-forward sweeps using sigma-point integration schemes for probabilistic value function approximation. Unlike the deterministic nature of the state-of-the-art GP-based DDP methods, the probabilistic foundation of the Fourier-Hermite methods makes them well-suited for integrating GPs. Therefore, we leverage GP-based forward uncertainty propagation within the Fourier-Hermite methods to propose sample-efficient data-driven methods, called GP-FHDP and GP-FHPDP, that can be applied to both stochastic and risk-sensitive optimal control problems. Furthermore, our methods can actively adjust exploration based on the uncertainty level, leading to accelerated convergence. The capabilities of the proposed algorithms are demonstrated on a simulated nonlinear vehicle system.

Keywords: Stochastic control, Predictive control for linear systems, Markov processes
Abstract: This paper presents a stochastic model predictive control (SMPC) algorithm for linear systems subject to additive Gaussian mixture disturbances, with the goal of satisfying chance constraints. We focus on a special case where each Gaussian mixture component has a similar variance. To solve the SMPC problem, we formulate a branch model predictive control (BMPC) problem on simplified dynamics and leverage stochastic simulation relations (SSR). Our contribution is an extension of the SMPC literature to accommodate Gaussian mixture disturbances while retaining recursive feasibility and closed-loop guarantees. We illustrate the retention of guarantees with a case study of vehicle control on an ill-maintained road.

Keywords: Stochastic control, Predictive control for nonlinear systems, Uncertain systems
Abstract: Switched latent force models (LFMs) are combinations of a first-principles physical model and a Gaussian process prior, where the driving force of the LFM may switch at certain time points. This allows to use expert knowledge to create an analytical state space model that describes large parts of the system behavior, while deviating parts are modeled using data-based methods. This paper proposes the combination of stochastic model predictive control and switched LFMs by reformulating the Gaussian process priors as linear state space models with additive white Gaussian noise. For this purpose, a stochastic optimization problem is formulated that can be solved by a deterministic approximation of the uncertainty propagation and the chance constraints. The switching points of the LFM introduce further uncertainty to the system that must be considered for the prediction of the state trajectories. Therefore, Gaussian mixture models are used to describe the probability density functions of the predicted states. The computation cost of the approach can be reduced by using a separate disturbance predictor, which allows to formulate the optimization problem of the model predictive controller independently of the internal disturbance states. The performance of the proposed method is illustrated for the control of a building energy system.

Keywords: Nonlinear system theory, Stochastic systems, Stability of nonlinear systems
Abstract: In this article, a novel reduced inversion zeroing neural network (RIZNN) is first proposed to solve the discrete periodic Riccati matrix equation (DPRE). To enhance the convergence and robustness performance of the RIZNN model, a nonlinear activation function, called Sine-Exponential activation function (SEAF), is constructed by combining a hyperbolic sine function with an exponential function. The convergence properties of the RIZNN model activated by SEAF are also proven under different noise conditions. Finally, simulation results are employed to illustrate the superiority of the RIZNN model activated by SEAF.

Keywords: Stochastic systems, Predictive control for linear systems, Adaptive control
Abstract: Recently data-driven predictive control (DDPC) methods have attracted considerable interest, primarily due to the benefit that optimal control inputs can be obtained directly from raw data without identifying an explicit model. However, time-varying characteristics of complex systems can substantially degrade the control performance of existing DDPC methods. In this work, we propose a new adaptive DDPC scheme in which innovations play a critical role. By leveraging the (approximate) white-noise and Gaussian properties of innovations, we develop a data-driven scheme that can effectively detect changes in system dynamics. This naturally motivates a statistically grounded adaptive DDPC framework in which the online adaptation is triggered by the detection of changes in system dynamics. Numerical examples demonstrate the efficacy of the proposed detection scheme and the enhanced control performance of the adaptive DDPC method in contrast to existing heuristic methods.

Keywords: Optimization algorithms, Stochastic control, Adaptive control
Abstract: In this paper a novel stochastic optimization and extremum seeking algorithm is presented, one which is based on time-delayed random perturbations and step size adaptation. For the case of a one-dimensional quadratic unconstrained optimization problem, global exponential convergence in expectation and global exponential practical convergence of the variance of the trajectories are proven. The theoretical results are complemented by numerical simulations for one- and multi-dimensional quadratic and non-quadratic objective functions.

Keywords: Distributed parameter systems
Abstract: This work is concerned with finding optimal actuator locations for infinite-dimensional systems with fractional operators. The system under consideration contains space-fractional derivatives and considers the mathbb{H}_2-control performance metric as an optimization function for location optimization. Numerical studies for a diffusion problem are presented to illustrate the effects of the orders of fractional operators on the optimal actuator location.

Keywords: Electrical power systems, Model/Controller reduction, Distributed estimation over sensor nets
Abstract: We introduce a novel power transmission line model featuring both continuous and discontinuous joints, where sensors are strategically placed at the discontinuous joints to capture real-world voltage discontinuities. This unique sensor topology sets it apart from conventional models. The system is governed by wave equations derived from the telegrapher’s equations, representing voltage propagation in an LC transmission line with distinct wave speeds and segment lengths. Voltage continuity is enforced at even-numbered joints, while the discontinuous joints at odd-numbered segments allow for precise sensor placement, creating a hybrid framework. In contrast to traditional models, where sensors at continuous joints may compromise exact observability, we employ the multiplier method to achieve exact observability with minimal observation time, utilizing energy identities and hidden regularity estimates. Furthermore, we develop a Finite Difference approximation that preserves observability uniform as the discretization parameter approaches zero. The proof hinges on discrete multipliers, closely mimicking the continuous PDE system. This work offers practical insights into controlling voltage propagation, providing more robust and efficient strategies for monitoring and stabilizing power transmission lines.

Keywords: Distributed parameter systems
Abstract: In this paper, an adaptive algorithm is proposed for identifying the unknown parameters in a stable linear single-input single-output (SISO) infinite-dimensional system. First, the transfer function of the infinite-dimensional system is assumed to be expressible as a ratio of two infinite series in the Laplace variable s. Additionally, the algebraic expressions relating the coefficients of these infinite series and the unknown parameters of the infinite-dimensional model are assumed to be known. Finally, certain identifiability conditions are also assumed to hold. Then, using an n-th order truncation of the transfer function, an adaptive update law driven by real-time input-output data is proposed for estimating the unknown parameters of the infinite-dimensional system. The estimates for the unknown parameters generated by the update law are shown to converge close to the true values at large times provided n is sufficiently large and the initial guess for the unknown parameters is within an identifiable region. The class of systems to which the proposed approach is applicable includes many partial differential equations (PDEs) with constant/spatially-varying parameters and distributed/boundary input and output. The efficacy of the approach is illustrated using two examples: a 1D heat equation with an unknown constant diffusion coefficient and a 1D wave equation with an unknown linearly (in space) varying parameter.

Keywords: Iterative learning control, Distributed parameter systems, Uncertain systems
Abstract: Iterative learning control for lumped-parameter systems is very well developed. Still, comparatively less attention has been directed to applying this method to those whose dynamics evolve in space and time. This paper gives new results in assessing the performance of various iterative learning control designs, focusing on a heat transfer application described by partial differential equations using a network of sensors and actuators.

Keywords: Optimization algorithms, Distributed parameter systems, Process control
Abstract: This paper presents an efficient approach for solving Optimal Control Problems of an industrial heating process for furniture production, called edge banding. The proposed algorithm solves the Quadratic Program that arises within a Sequential Quadratic Programming framework, where a Gauss-Newton approximation of the Hessian matrix is applied. The core of this approach is a tailored condensing method designed to exploit the unique structure of the system model, which results from discretizing the governing Partial Differential Equation using Lagrangian coordinates. We present an efficient implementation of the algorithm and provide numerical results, showing a substantial computational speedup compared to an ad-hoc implementation using a state-of-the-art QP solver.

Keywords: Stability of nonlinear systems, Chemical process control, Lyapunov methods
Abstract: We investigate a nonlinear parabolic partial differential equation whose boundary conditions contain a single control input. This model describes a chemical reaction of the type 'A -> product', occurring in a dispersed flow tubular reactor. The existence and uniqueness of solutions to the nonlinear Cauchy problem under consideration are established by applying the theory of strongly continuous semigroups of operators. We also prove the stability of the equilibrium of the closed-loop system with a proposed feedback law. Additionally, using Lyapunov's direct method, we evaluate the exponential decay rate of the solutions.

Keywords: Computational methods, Algebraic/geometric methods, Nonlinear system theory
Abstract: This article presents a novel and intrinsic approach to deriving a geometric integrator for a nonholonomic mechanical system based on the notions of retraction maps and their lifts to tangent and cotangent bundles. We use adapted Hamel coordinates to transform the original system with constraints to a reduced set of states on the dual of the non-integrable distribution.

Keywords: Computational methods, Differential algebraic systems, Stability of nonlinear systems
Abstract: Singularly perturbed systems (SPSs) are prevalent in engineering applications, where numerically solving their initial value problems (IVPs) is challenging due to stiffness arising from multiple time scales. Classical explicit methods require impractically small time steps for stability, while implicit methods developed for SPSs are computationally intensive and less efficient for strongly nonlinear systems. This paper introduces a Stabilized Multirate Explicit Scheme (SMES) that stabilizes classical explicit methods without the need for small time steps or implicit formulations. By employing a multirate approach with variable time steps, SMES allows the fast dynamics to rapidly converge to their equilibrium manifold while slow dynamics evolve with larger steps. Analysis shows that SMES achieves numerical stability with significantly reduced computational effort and controlled error. Its effectiveness is illustrated with a numerical example.

Keywords: Computational methods, Feedback linearization, Algebraic/geometric methods
Abstract: Mechanical systems are most often described by a set of continuous-time, nonlinear, second-order differential equations (SODEs) of a particular structure governed by the covariant derivative. The digital implementation of controllers for such systems requires a discrete model of the system and hence requires numerical discretization schemes. Feedback linearizability of such sampled systems, however, depends on the discretization scheme employed. In this article, we utilize retraction maps and their lifts to construct feedback linearizable discretizations for SODEs which can be applied to many mechanical systems.

Keywords: Computational methods, Identification, Randomized algorithms
Abstract: In this paper, we propose step-size adaptation methods for the Hamiltonian Monte Carlo (HMC) and Metropolis-adjusted Langevin algorithms (MALA). The adaptation procedures consist of an acceptance rate estimator which is implemented as a Bayesian filter on the observed acceptance indicator sequence. This sequence is modeled as a Bernoulli sequence with a time-varying probability, and its distribution is represented by a beta distribution. Therefore, the resulting filter is called the Beta-Bernoulli filter. The acceptance rate is then controlled to the desired target acceptance rate using a linear feedback controller. The resulting adaptation mechanism is experimentally evaluated in practical MCMC sampling tasks.

Keywords: Computational methods, Linear systems, Predictive control for linear systems
Abstract: This study focuses on the numerical discretization methods for the continuous-time discounted linear-quadratic optimal control problem (LQ-OCP) with time delays. By assuming piecewise constant inputs, we formulate the discrete system matrices of the discounted LQ-OCPs into systems of differential equations. Subsequently, we derive the discrete-time equivalent of the discounted LQ-OCP by solving these systems. This paper presents three numerical methods for solving the proposed differential equations systems: the fixed time-step ordinary differential equation (ODE) method, the step-doubling method, and the matrix exponential method. Our numerical experiment demonstrates that all three methods accurately solve the differential equation systems. Interestingly, the step-doubling method emerges as the fastest among them while maintaining the same level of accuracy as the fixed time-step ODE method.

Keywords: Computational methods, Variational methods, Optimization algorithms
Abstract: In this paper, we present monviso (monotone variational inequalities solver), a novel open-source Python package for solving monotone variational inequalities. We detail the package's structure and baseline functionality, discussing a simple example that illustrates the essential methods and parameters. Moreover, we characterize how the proximal operator, which is the foundation of many iterative schemes, is handled through cvxpy, an open-source Python library for convex optimization. We list the available algorithms and describe the basic implementation of any general iterative method to enable users to build additional and (possibly new) algorithms. Finally, we illustrate several examples of possible use cases for monviso, showcasing the different applications the package can support across various fields, including control, optimization, dynamic game theory, and machine learning.

Keywords: Process control, Chemical process control, Sliding mode control
Abstract: Dead-time is a common phenomenon in industrial processes, which can degrade the overall performance of a control system, causing poor setpoint tracking, slow disturbance rejection, and reduced control accuracy. This paper discusses a hybrid control strategy based on sliding-mode control and the Smith Predictor approach using a sliding surface with iterated integral terms. For evaluation, two examples are utilized: the first deals with a higher-order linear system characterized by substantial delay, emphasizing tracking, disturbance rejection, and referencing variable tracking. The second example pertains to experimental research conducted with the TCLab device.

Keywords: Process control, Computational methods, Manufacturing processes
Abstract: This study introduces a novel control approach for an aluminum oxide Atomic Layer Etching (ALE) process, combining an ex-situ Run-to-Run (R2R) and real-time Endpoint (EP) controller. The method dynamically adjusts process times for both ALE half-cycles to address disturbances and optimize outcomes. The EP controller employs a machine learning-based soft sensor, using a transformer architecture to analyze variable-length pressure profiles from five wafer inspection points and make binary classification decisions. To ensure robust performance, a multiscale modeling method integrates Computational Fluid Dynamics (CFD) with kinetic Monte Carlo (kMC) simulations, offering accuracy and low computational cost. Various R2R-EP combinations, including traditional Exponential Weighted Moving Average and a novel Standard Case Corrector, were tested, showing significant improvements in reducing misprocessing rates compared to traditional controllers.

Keywords: Process control, Machine learning, Fault detection and identification
Abstract: The increasing demand for cost-effective moni- toring solutions has led to widespread adoption of virtual sensors, which estimate critical system states and parameters using available measurements rather than dedicated physical sensors. However, these data-based solutions face a fundamental challenge: their accuracy often deteriorates significantly when deployed in field conditions that differ from laboratory testing environments. Such distribution shifts are often inevitable due to varying environmental conditions, equipment aging, and different usage patterns across installations. Although virtual sensors are typically trained and validated under controlled laboratory conditions, maintaining their performance for both state estimation and parameter estimation under real-world distribution shifts remains an open problem. This work addresses this challenge by providing two complementary mitigation strategies: an adaptation approach based on importance sam- pling and a robust design method using uniform training. We demonstrate our approach in an industrial freezer monitoring system, where the goal is to estimate the insulation quality despite the varying operating conditions between laboratory testing and field deployment. The results provide theoretically grounded, yet practical tools for deploying reliable virtual sensing solutions.

Keywords: Process control, Complex systems, Mechatronics
Abstract: The new aspects provided by the current paper consist of merging the features of the two data-driven algorithms, Active Disturbance Rejection Control (ADRC) and Fictitious Reference Iterative Tuning (FRIT), that are reunited as ADRC-FRIT algorithm to improve the control performances of the three degrees of freedom Tower Crane Systems (TCSs). The ADRC-FRIT algorithm determines the optimal controller parameters by solving a gradient problem using the classical Gauss-Newton algorithm. Both ADRC and FRIT algorithms belong to the category of data-driven algorithms, i.e., the input/output data are used to compute the controller parameters. A comparison study of the new ADRC-FRIT algorithm with the ADRC algorithm is performed, where both algorithms are proven through experimental evaluation using the TCS laboratory equipment.

Keywords: Control of metal processing, Manufacturing processes, Reduced order modeling
Abstract: Laser-based directed energy deposition is an additive manufacturing technique used for 3d printing of metals. In processes of this type, the laser power is often kept constant or controlled with very simple controllers. An appropriate feedforward laser power signal has the potential to increase the geometric quality of the printed product. We derive a computationally efficient reduced order model (ROM) from a finite-element (FE) model and simulation that describes the temperature field induced by the moving laser heat source. We show the ROM can be used to find an optimal laser power signal by solving an optimal control problem that would be cumbersome with the FE model. The optimal control solution achieves reference tracking for the melt pool width and outperforms a PI width controller even if this PI controller is tuned systematically.

Keywords: Quantum control, Predictive control for nonlinear systems, Stochastic systems
Abstract: This paper explores the application of Model Predictive Control (MPC) strategies to quantum systems undergoing continuous-time measurement. The MPC control law is formulated based on the estimated state obtained from quantum filtering. As quantum filtering is affected by uncertainties, applying MPC for real-time monitoring enhances the stability of the filtering process. Our analysis demonstrates that the MPC framework ensures exponential stability in mean square. Additionally, we establish that a nonlinear MPC model guarantees system stability under specified conditions. Furthermore, the inherent robustness of MPC allows the system to remain exponentially bounded in mean square, even in the presence of environmental disturbances. Numerical simulations are provided to validate these control properties of quantum MPC.

Keywords: Autonomous systems, Autonomous robots, Optimization
Abstract: This paper presents a novel point cloud registration approach tailored to meet the requirements of autonomous racing, where rapid and precise LiDAR-based localization is essential in high-speed environments. The proposed method, track constrained GICP (tc-GICP), is a variant of the well-established Generalized Iterative Closest Point (GICP) algorithm, where we increase the computational efficiency by lowering the degrees of freedom of the optimization problem from six to three. This modification, that accelerates convergence significantly without affecting localization accuracy, is accomplished by restricting vehicle pose estimations to a predefined track plane. The performance of tc-GICP is evaluated using real-world data from competitions attended by the PoliMOVE racing team, including the Goodwood Festival of Speed and the 2024 Indianapolis Autonomous Challenge. The results demonstrate the capability of our algorithm to preserve high localization accuracy while achieving a notable speed improvement.

Keywords: Autonomous systems, Machine learning, Intelligent systems
Abstract: In the field of autonomous surface vehicles (ASVs), devising decision-making and obstacle avoidance solutions that address maritime COLREGs (Collision Regulations), primarily defined for human operators, has long been a pressing challenge. Recent advancements in explainable AI and machine learning have shown promise in enabling human-like decision-making. Notably, there have been significant developments in the application of Large Language Models (LLMs) to the decision-making of complex systems, such as self-driving cars. The textual and somewhat ambiguous nature of COLREGs (from an algorithmic perspective), however, poses challenges that align well with the capabilities of LLMs, suggesting that LLMs may become increasingly suitable for this application soon. This paper presents and demonstrates the first application of LLM-based decision-making and control for ASVs. The proposed method establishes a high-level decision-maker that uses online collision risk indices and key measurements to make decisions for safe manoeuvres. A tailored design and runtime structure is developed to support training and real-time action generation on a realistic ASV model. Local planning and control algorithms are integrated to execute the commands for waypoint following and collision avoidance at a lower level. To the authors' knowledge, this study represents the first attempt to apply explainable AI to the dynamic control problem of maritime systems recognising the COLREGs rules, opening new avenues for research in this challenging area. Results obtained across multiple test scenarios demonstrate the system's ability to maintain online COLREGs compliance, accurate waypoint tracking, and feasible control, while providing human-interpretable reasoning for each decision.

Keywords: Autonomous systems, Robotics
Abstract: The development of autonomous vessels is a topic of great interest in the maritime industry. Generating trajectories or vessels over long time horizons while respecting multiple sets of rules, is a complicated task. In this contribution, a procedure to generate trajectories that comply with the rules of preventing collisions at sea (COLREGs) is presented. The approach is based on the combination of three methods, namely the rapidly exploring random tree (RRT) algorithm, velocity obstacles (VOs), and motion primitives (MPs). In this approach, the dynamic situation is analyzed and the corresponding COLREG rule is enforced by selecting only valid, i.e., COLREG conform, MPs from a database. This database is created by solving a boundary value problem (BVP) for a set of initial and final conditions. The approach is capable of respecting static and dynamic obstacles and can be used for a variety of vessel classes and time horizons.

Keywords: Autonomous systems, Autonomous robots, Robotics
Abstract: In diverse applications such as large industrial settings, automated guided vehicles (AGVs) deal with discontinuous localization feedback from optical sensors or QR code readers. These discontinuities may induce aggressive behavior in reactive controllers, which is dangerous in practical operations. Additionally, many AGVs are chosen for their low power consumption, to be used for longer operation times, which commonly reduces the computational resources of the system. This work proposes a real-time path-following algorithm based on Bézier curves to address these challenges. The proposed strategy selects third-order Bézier curve control points in a logical and structured manner to create smooth recovery trajectories as discontinuities appear. Numerical simulations and practical experiments were carried out to validate the performance of the proposed approach. A comparison with model predictive control (MPC) showcased that the Bézier-based approach achieves a similar performance in terms of control smoothness, while the computational complexity is reduced.

Keywords: Autonomous robots, Automotive, Machine learning
Abstract: This work proposes an end-to-end learning-based controller for the task of lane-following by an autonomous vehicle. The perception module is developed as a unified camera model that is able to perform multiple tasks namely- object detection, lane segmentation, free-space segmentation, lane-detection and depth-estimation from a single RGB camera image. Since, there is no single data-source that contains labels for training all the above mentioned tasks at once, we propose a novel training methodology that enables us to train a single network on many different task-specific datasets in an asynchronous manner. Based on empirical studies we demonstrate that the multi-dataset training outperforms standard method in combined task with a marginal trade-off in the object detection task. Subsequently, we combine the perception module with a path-following learning based controller and deploy it on real-vehicle and test the end-to-end controller on real roads.

Keywords: Autonomous systems, Sensor and signal fusion, Maritime
Abstract: Global navigation satellite systems (GNSS) is the gold standard for aiding of inertial navigation systems (INS), but is highly susceptible to naturally occurring, unintentional, and intentional interference due to its low signal power. GNSS is also not suitable for use indoors or in other areas where the signal is blocked. In recent years, phased-array radio systems (PARS) have shown to have potential for navigation. After Bluetooth introduced direction finding in its specification, the technology has emerged as a low-cost PARS-based alternative. In this paper, an estimation scheme fusing PARS and inertial sensor data based on factor graph optimisation (FGO) using incremental smoothing and mapping with fixed-lag smoothing is applied and compared against a standard error-state Kalman filter (ESKF) solution as a benchmark. The results are obtained from 100 Monte Carlo runs of a simulated USV trajectory, where performance is compared when aided by both fault-free and erroneous PARS measurements, with and without robust estimation schemes. We find that our estimator slightly outperforms ESKF in North-East position and roll/pitch estimates, but struggles more with Down-position and yaw angle, possibly due to additional cross-covariance available to the ESKF.

Keywords: Machine learning, Neural networks, Intelligent systems
Abstract: This paper presents a novel method for generating realistic counterfactual explanations (CFEs) in machine learning (ML)-based control for mobile robots using 2D LiDAR. ML models, especially artificial neural networks (ANNs), can provide advanced decision-making and control capabilities by learning from data. However, they often function as "black boxes," making it challenging to interpret them. This is especially a problem in safety-critical control applications. To generate realistic CFEs, we parameterize the LiDAR space with simple shapes such as circles and rectangles, whose parameters are chosen by a genetic algorithm, and the configurations are transformed into LiDAR data by raycasting. Our model-agnostic approach generates CFEs in the form of synthetic LiDAR data that resembles a base LiDAR state but is modified to produce a pre-defined ML model control output based on a query from the user. We demonstrate our method on a mobile robot, the TurtleBot3, controlled using deep reinforcement learning (DRL) in real-world and simulated scenarios. Our method generates logical and realistic CFEs, which helps to interpret the DRL agent's decision making. This paper contributes towards advancing explainable AI in mobile robotics, and our method could be a tool for understanding, debugging, and improving ML-based autonomous control.

Keywords: Machine learning, Optimal control, Automotive
Abstract: This paper introduces a hierarchical motion planning strategy for autonomous tractor-trailer systems, designed for efficient long-horizon, collision-free maneuvering in complex environments. By combining high-level reference line graph search with low-level primal-dual reinforcement learning (RL)-based trajectory optimization, our approach addresses the computational challenges inherent to the motion planning of tractor-trailer dynamics. The high-level graph search decides waypoints guided by Reeds-Shepp cost, and the low-level RL connects the waypoints with dynamically feasible and collision-free trajectories. To enhance safety and accuracy, we incorporate reachability constraints and batch trajectory sampling in the algorithm design. Empirical results show that our method significantly reduces computation time, outperforming traditional state-lattice-based planning approaches and enabling real-time applicability.

Keywords: Machine learning, Agents and autonomous systems, Cooperative autonomous systems
Abstract: We present a decentralized reinforcement learning (RL) approach to address the multi-agent shepherding control problem, departing from the conventional assumption of cohesive target groups. Our two-layer control architecture consists of a low-level controller that guides each herder to contain a specific target within a goal region, while a high-level layer dynamically selects from multiple targets the one an herder should aim at corralling and containing. Cooperation emerges naturally, as herders autonomously choose distinct targets to expedite task completion. We further extend this approach to large-scale systems, where each herder applies a shared policy, trained with few agents, while managing a fixed subset of agents.

Keywords: Intelligent systems, Machine learning, Mechatronics
Abstract: Bearing-only tracking, localisation, and circumnavigation is a problem in which a single or a group of agents attempts to track a target while circumnavigating it at a fixed distance using only bearing measurements. While previous studies have addressed scenarios involving stationary targets or those moving with an unknown constant velocity, the challenge of accurately tracking a target moving with a time-varying velocity remains open. This paper presents an approach utilising a Long Short-Term Memory (LSTM) based estimator for predicting the target's position and velocity. We also introduce a corresponding control strategy. When evaluated against previously proposed estimation and circumnavigation approaches, our approach demonstrates significantly lower control and estimation errors across various time-varying velocity scenarios. Additionally, we illustrate the effectiveness of the proposed method in tracking targets with a double integrator nonholonomic system dynamics that mimic real-world systems.

Keywords: Filtering, Machine learning, Adaptive systems
Abstract: We present a novel approach to processing periodic signals with non-stationary fundamental frequency. These quasi-periodic signals feature a perpetually recurring underlying signal pattern and arise in various fields of science and engineering. The proposed method integrates a recursive extraction of the signal pattern with dynamic tracking of the instantaneous phase to effectively suppress measurement noise without prior knowledge of the signal characteristics or the frequency variation. The performance is showcased both in simulation and using experimental measurements of the cardiac cycle obtained by a nitrogen-vacancy diamond quantum sensor. Overall, high-fidelity signal reconstruction and convincing pattern learning is achieved, even in the presence of complex non-linear disturbances and non-Gaussian noise. Conclusively, the proposed technique constitutes a flexible and efficient solution, addressing limitations of existing methods and offering real-world applicability.

Keywords: Control of metal processing, Machine learning, Manufacturing processes
Abstract: Reinforcement Learning (RL) continues to attract considerable attention in academia and industry. Its data-driven nature, combined with its varied and flexible formulation, makes it applicable in a variety of complex control tasks, in which control theory techniques can be challenging to implement. The Powder Bed Fusion (PBF) process, comprises an example of such a complex control task. However, there are still critical challenges in RL that need to be addressed in order to fully enable its use in PBF implementations. For instance, while constraint satisfaction comprises a necessity in PBF process control, there are still gaps in demonstration and analysis of RL algorithmic behaviour and control performance under constraint satisfaction. Existing constraint techniques in the literature, such as radial squashing, can provide zero constraint violation guarantees in process control. However, a constraint framework that also accounts for satisfactory control performance must be established. In this work, an attempt to address the above challenges is presented, providing a thorough analysis on constrained RL for PBF process control, and assessing the impact of the radial squashing technique. The results of this analysis show that tuning the intensity of radial squashing can be vital for maintaining satisfactory control performance under constraints.

Keywords: Transportation systems, Automotive, Cooperative control
Abstract: In this paper, we introduce the extended delayed spacing policy for vehicle platoons, motivated by its capability to deal with input delays. The extended delayed spacing policy can be asymptotically tracked with a state feedback controller that does not depend on future or past states. As the input can be regarded as the result of input-output linearization, it remains to show that the controller also guarantees stable internal dynamics. However, internal stability is shown to depend on the choice of parameters in the spacing policy. Finally, it is shown that string stability is a consequence of the choice of spacing policy and depends on the choice of parameters. A sufficient condition on the parameters of the spacing policy guaranteeing string stability is presented. The results are illustrated through simulations.

Keywords: Transportation systems, Neural networks
Abstract: Efficient traffic monitoring is crucial for managing urban transportation networks, especially under congested and dynamically changing traffic conditions. Drones offer a scalable and cost-effective alternative to fixed sensor networks; however, deploying fleets of low-cost drones for traffic monitoring poses challenges in adaptability, scalability, and real-time operation. To address these issues, we propose a learning-based framework for decentralized traffic monitoring with drone swarms, targeting the uneven and unpredictable distribution of monitoring needs across urban areas. Our approach introduces a semi-decentralized reinforcement learning model, which trains a single Q-function using the collective experience of the swarm. This model supports full scalability, flexible deployment, and, when hardware allows, the online adaptation of each drone’s action-selection mechanism. We first train and evaluate the model in a synthetic traffic environment, followed by a case study using real traffic data from Shenzhen, China, to validate its performance and demonstrate its potential for real-world applications in complex urban monitoring tasks.

Keywords: Transportation systems, Automotive, Robust control
Abstract: This paper proposes a novel approach to a Cooperative Adaptive Cruise Control (CACC) system using structured H∞ synthesis. The approach allows vehicle platoons to consist of heterogeneous dynamics models. This method allows a platoon to be considered a multi-agent system (MAS) and the control problem is solved by including the different models of the vehicles at once, while performance specifications are satisfied in the frequency domain. The suggested control strategy ensures the consensus of vehicle velocities and spacing distances at the steady state of the platoons. The simulation results, presented below, validate the capabilities and the potential of the proposed control framework in a platoon with different vehicle masses in a) predecessor-following (PF), and b) two-predecessor-following (TPF) communication topologies.

Keywords: Transportation systems, Game theoretical methods, Computational methods
Abstract: We investigate the problem of platoon matching through the lens of revealead preference. Specifically, we consider a noncooperative game among a number of vehicles that decide to form or not to form a platoon on a single road. Given a collection of observed vehicles’ decisions, we provide tractable procedures to decide whether such decisions are rational under some game and some model of rationality, i.e., whether the decisions are rationalizable. We further show that for each model of rationality in our consideration, there exists a collection of observed vehicles’ decisions that are not rationalizable, showing that our model of decision making in vehicle platooning is not too general as to admit any vehicle’s decision as rational under some game and some model of rationality.

Keywords: Filtering, Transportation systems, Stochastic filtering
Abstract: This paper presents an approach to tackle the problem of tram localization through utilizing a custom processing of Global Navigation Satellite System (GNSS) observables and the track map. The method is motivated by suboptimal performance in dense urban environments where the direct line of sight to GNSS satellites is often obscured which leads to multipath propagation of GNSS signals. The presented concept is based upon the iterated extended Kalman filter (IEKF) and has linear complexity (with respect to the number of GNSS measurements) as opposed to some other techniques mitigating the multipath signal propagation. The technique is demonstrated both on a simulated example and real data. The root-mean-squared errors from the simulated ground truth positions show that the presented solution is able to improve performance compared to a baseline localization approach. Similar result is achieved for the experiment with real data, while treating orthogonal projections onto the tram track as the true position, which is unavailable in the realistic scenario. This proof-of-concept shows results which may be further improved with implementation of a bank-of-models method or chi-squared-based rejection of outlying GNSS pseudorange measurements.

Keywords: Concensus control and estimation, V&V of control algorithms, Agents networks
Abstract: This paper is concerned with a recently proposed scalable protocol for vehicle platooning, known as the serial consensus protocol. This achieves coordination of second-order integrator systems through a series connection of first-order conventional consensus protocols, and has been theoretically shown to have advantageous stability and performance properties. We implement this protocol on a vehicle platoon of five robots in a lab setting, where it is subject to measurement noise. We present experimental verification that the system is scalably stable, which is in accordance with theoretical findings. Moreover, we show theoretically that the parameters for which scalable stability occurs can be relaxed when information of the communication topology is available, which is also verified experimentally. In parallel, a conventional consensus protocol is implemented, which is known to not be scalably stable, which we also demonstrate in experiments. Lastly, we implement experiments to demonstrate the string stable behavior of the the serial consensus protocol in directed vehicle platoons, as predicted by theory, which again does not hold for conventional consensus.

Keywords: Optimization algorithms, Optimization, Process control
Abstract: Online Feedback Optimization uses optimization algorithms as dynamic systems to design optimal control inputs. The results obtained from Online Feedback Optimization depend on the setup of the chosen optimization algorithm. In this work we analyse the sensitivity of Online Feedback Optimization to the parameters of projected gradient descent as the algorithm of choice. We derive closed-form expressions for sensitivities of the objective function with respect to the parameters of the projected gradient and to time-varying model mismatch. The formulas are then used for analysis of model mismatch in a gas lift optimization problem. The results of the case study indicate that the sensitivity of Online Feedback Optimization to the model mismatch depends on how long the controller has been running, with decreasing sensitivity to mismatch in individual timesteps for long operation times.

Keywords: Optimization, Optimization algorithms, H2/H-infinity methods
Abstract: In this paper, we propose control-theoretic methods as tools for the design of online optimization algorithms that are able to address dynamic, noisy, and partially uncertain time-varying quadratic objective functions. Our approach introduces two algorithms specifically tailored for scenarios where the cost function follows a stochastic linear model. The first algorithm is based on a Kalman filter-inspired approach, leveraging state estimation techniques to account for the presence of noise in the evolution of the objective function. The second algorithm applies H-infinity-robust control strategies to enhance performance under uncertainty, particularly in cases in which model parameters are characterized by a high variability.

Through numerical experiments, we demonstrate that our algorithms offer significant performance advantages over the traditional gradient-based method and also over the optimization strategy proposed in [1] based on deterministic models.

Keywords: Optimization algorithms, Linear systems, Electrical power systems
Abstract: In this paper, we propose a gradient projection algorithm aimed at improving the transient performance of feedback-based optimization (FO) for linear dynamical systems. Our approach leverages a specifically designed gain matrix, replacing the usual scalar step size to enhance trajectory efficiency and reduce oscillations. By solving a semi-definite programming, we select the gain matrix to trade off between convergence rate and oscillation minimization. Compared to the standard FO algorithms, our method demonstrates improved transient performance in numerical simulations and in turn faster convergence.

Keywords: Optimization algorithms, Stability of nonlinear systems, Sliding mode control
Abstract: We propose a novel primal-dual gradient dynamics (PDGD) algorithm to dynamically solve an optimization problem with linear equality constraints in finite time. To ensure finite-time convergence, we endow the PDGD with suitable homogeneity properties. More precisely, departing from the standard PDGD and based on the associated Lagrangian of the optimization problem, the algorithm is derived by suitably combining a change of coordinates of the standard PDGD with the multivariable super-twisting algorithm. In our new coordinates, the proposed PDGD's global convergence to the optimal solution of the optimization problem is then proven via a smooth, strong Lyapunov function. Additionally, we provide a numerical example to compare the performance of our algorithm with existing approaches from the literature.

Keywords: Computer aided control design, Optimization algorithms, Mechatronics
Abstract: Commissioning industrial processes is often tedious and difficult because certain settings, like controller parameters, need repeated adjustment to reach a specific goal. Passing knowledge on system tuning from experienced operators to junior colleagues is also challenging, as this knowledge is often tacit and hard to formalize.

We approach the tuning problem with Bayesian optimization (BO). We discuss different ways of combining the best of both worlds, i.e., the insights of the expert with the goal-driven exploration of BO. With the goal of incorporating recorded expert tuning sessions into BO, we delve into two main directions.

First, we explore the effect of context changes on a Human First-Computer Last approach, where the expert conducts an initial rough tuning, and BO performs a fine-tuning afterward. Second, we outline initial steps toward a subgoal-based BO method, based on the observation that humans often divide complex tasks into multiple subgoals achieved in sequence.

We evaluate these methods on two simulation setups: a cascade controller of a pusher-slider robotic system and a projectile trajectory planning problem.

Keywords: Optimization algorithms, Optimization
Abstract: The Augmented Lagrangian Alternating Direction Inexact Newton (ALADIN) method is a cutting-edge distributed optimization algorithm known for its superior numerical performance. It relies on each agent transmitting information to a central coordinator for data exchange. However, in practical network optimization and federated learning, unreliable information transmission often leads to packet loss, posing challenges for the convergence analysis of ALADIN. To address this issue, this paper proposes Flexible ALADIN, a random polling variant of ALADIN, and presents a rigorous convergence analysis, including global convergence for convex problems and local convergence for non-convex problems.

Keywords: Optimization, Optimization algorithms, Emerging control applications
Abstract: After introducing the session by reviewing recent results on control-theoretic approaches to analysis and design of optimization algorithms, this tutorial talk introduces a novel framework for equality-constrained optimization based on control theory. The main idea is to design a feedback control system in which the Lagrange multipliers serve as the control inputs while the output represents the constraints. This system is forced to converge to a stationary point of the constrained optimization problem through suitable regulation of the Lagrange multipliers. More specifically, we present two control laws: proportional-integral control and feedback linearization that lead to a variety of different methods. We rigorously develop the related algorithms, analyze their convergence theoretically, and present several numerical experiments that demonstrate their effectiveness compared to the state-of-the-art approaches.

Keywords: Optimization algorithms, Optimization, Emerging control applications
Abstract: Sparse optimization is the science of learning parsimonious models through the solution of suitable minimization problems. Sparsity is a desirable condition for several purposes. For example, it reduces the numerical complexity and memory footprint of a model for edge deployment and it enhances the physical interpretability. For this motivation, sparsity is studied in several contexts, ranging from system identification to neural networks.

Proximal, gradient-based iterative algorithms are a popular approach to sparse optimization problems, because they are straightforward to implement. However, they require the setting of hyperparameters that affect the performance. Typically, this is a sensitive task that entails a tradeoff between estimation accuracy and convergence speed. To overcome this drawback, some recent works propose a time-varying tuning of the hyperparameters according to specific feedback control laws; see, e.g., cite{fox23} and references therein. More precisely, iterative sparse optimization algorithms are discrete-time dynamical systems, in which the hyperparameters can serve as control inputs to regulate the behavior and enhance the performance.

The goal of this tutorial talk is to propose a complete overview of control-theoretic approaches to sparse optimization algorithms. Specifically, we compare the different results obtained in the recent literature and we discuss possible future directions in this research area.

Keywords: Optimization, Optimization algorithms
Abstract: In many applications - ranging from control, to image processing, to machine learning - we are interested in solving online optimization problems, that is, optimization problems whose data change over time. Designing online optimization algorithms presents a challenging set of constraints, foremost of which is the need for real-time operations. In this talk, I will discuss the fruitful application of model-based techniques to design efficient online algorithms. After a background on online optimization, I will discuss how we can leverage techniques from (robust) control theory to design online algorithms that outperform the state of the art. I will show theoretical and numerical results for both unconstrained and constrained optimization problems, and discuss some interesting future directions.

Keywords: Optimization algorithms, Nonlinear system theory, Safety critical systems
Abstract: In this tutorial we consider the problem of designing optimization algorithms for settings where the solution is used to regulate a physical plant. This scenario is ubiquitous in engineering applications e.g. model predictive controllers for regulating industrial processes, safety filters in robotics, and online feedback optimization for power systems. In these applications, the optimization problem often incorporates constraints which, when violated, would threaten the safe operation of the system. Thus, we require that the algorithm is anytime, meaning that it is guaranteed to return a feasible point even when terminated before it converges to a solution. To address this problem, rather than adopting the dominant view of an optimization algorithm as simply a computational method, we instead approach its design from a systems theoretic perspective. This is because the optimization algorithm, along with the physical process it regulates, can be viewed as a pair of interconnected dynamical systems, which is naturally analyzed using control theoretic tools. Our solution begins by considering a "static" nonlinear optimization problem and introducing an anytime solver in the form of a continuous time dynamical system — derived using techniques from safety critical control — called the safe gradient flow. Next, we "close the loop" and analyze the interconnection of the safe gradient flow with a dynamical system modeling the plant. Finally, to address the practical implementation of these methods on digital computers, we tackle the challenge of discretizing the safe gradient flow.

Keywords: Hybrid systems, Reduced order modeling, Modeling
Abstract: In this paper, a port-Hamiltonian approach to modelling of mechanical systems subject to switching contact constraints is presented. It is well known that a class of constrained systems can be represented as port-Hamiltonian systems without Lagrange multipliers by considering a reduced momentum space. Here, we revisit the modelling of these systems for the purpose of demonstrating that the discrete dynamics associated with switching constraints can be obtained directly through the process of mapping between reduced-order continuous systems. The modelling framework is illustrated by applying it to the compass-like biped robot, and the obtained port-Hamiltonian model is used to derive a passivity-based controller.

Keywords: Switched systems, Hybrid systems, Uncertain systems
Abstract: In the paper, we utilize the prescribed performance control (PPC) design framework to achieve tracking with preselected and user-defined, transient and steady-state, performance characteristics, for a class of uncertain, state impulsive switched nonlinear systems. The state impulses and the switched system dynamics are synchronous. The impulsive time sequence can be aperiodic with unknown impulse time instants. Besides the boundedness of all signals in the closed loop, the proposed controller guarantees that between any two successive impulse time instances, the output tracking error converges to a predefined and arbitrarily small neighborhood of zero within a user-specified fixed time. Simulation studies verify and clarify the theoretical findings.

Keywords: Discrete event systems, Automata, Fault diagnosis
Abstract: We consider a robust decentralized diagnosis problem for uncertain discrete event systems. We assume that, instead of the single exact model, a set of multiple possible models, each of which has its own event set and own normal behavior, is given for the task of diagnosis. The robust decentralized diagnosis problem requires us to synthesize a decentralized diagnoser such that, for all possible models, it detects the occurrence of any failure string within a finite number of steps. To characterize the existence of such a robust decentralized diagnoser, we introduce a notion of robust codiagnosability of the set of possible models. Then, we develop a method for verifying robust codiagnosability effectively.

Keywords: Hybrid systems, Switched systems, Observers for nonlinear systems
Abstract: This paper presents sufficient reconstructability and observability conditions for piecewise linear systems affected by known and unknown inputs. Based on these conditions, an observer design using a High Order Sliding Mode (HOSM) differentiator is proposed to estimate both the continuous and discrete states. Additionally, numerical examples and simulations are given to explain and illustrate the results.

Keywords: Statistical learning, Discrete event systems, Machine learning
Abstract: Accurately forecasting the direction of financial returns poses a formidable challenge, given the inherent unpredictability of financial time series. The task becomes even more arduous when applied to cryptocurrency returns, given the chaotic and intricately complex nature of crypto markets. In this study, we present a novel prediction algorithm using limit order book (LOB) data rooted in the Hawkes model, a category of point processes. Coupled with a continuous output error (COE) model, our approach offers a precise forecast of return signs by leveraging predictions of future financial interactions. Capitalizing on the non-uniformly sampled structure of the original time series, our strategy surpasses benchmark models in both prediction accuracy and cumulative profit when implemented in a trading environment. The efficacy of our approach is validated through Monte Carlo simulations. The research draws on LOB measurements from a centralized cryptocurrency exchange where the stablecoin Tether is exchanged against the U.S. dollar.

Keywords: Hybrid systems, Switched systems
Abstract: This paper reports a theory of Koopman operators for a class of hybrid dynamical systems with globally asymptotically stable periodic orbits, called hybrid limit-cycling systems. We leverage smooth structures intrinsic to the hybrid dynamical systems, thereby extending the existing theory of Koopman operators for smooth dynamical systems. Rigorous construction of an observable space is carried out to preserve the inherited smooth structures of the hybrid dynamical systems. Complete spectral characterization of the Koopman operators acting on the constructed space is then derived where the existence and uniqueness of their eigenfunctions are ensured. Our results facilitate global analysis of hybrid dynamical systems using the Koopman operator.

Keywords: Observers for nonlinear systems, Sliding mode control, Uncertain systems
Abstract: In this paper, we propose a global third-order homogeneous in the bi-limit sliding mode observer (BL-H SMO) that can estimate the states of an uncertain nonlinear system with unknown inputs in finite time. The system must be strongly observable w.r.t. unknown inputs (UIs) and uniformly observable w.r.t. known inputs. To handle the UIs and non-Lipschitz nonlinearities, the proposed observer combines sliding-mode and high-gain observers, extending the generalized super-twisting observer. The convergence of the BL-H SMO to the system states is proven through a Lyapunov function. Finally, to showcase the effectiveness of the proposed method, the paper includes an academic example and a practical example, namely a three-phase converter with LCL-filter.

Keywords: Observers for nonlinear systems, Stability of nonlinear systems, Lyapunov methods
Abstract: This paper addresses the problem of designing non-linear observers for bearing-only target motion analysis (BOTMA). For the case where a source and target agent are navigating in a plane, we design non-linear state observers to estimate the target’s trajectory using only noisy bearing measurements collected by the source. Considering linear relative dynamics for the mobile agents and non-linear relative bearing measurements, the observer gains are designed by solving a set of time-varying linear matrix inequality (LMI) conditions. Compared to existing filters in the stochastic framework used to solve the BOTMA problem, the proposed non-linear observer provides guarantees based on the Lyapunov stability criterion. The effectiveness of the proposed observer for target state estimation is demonstrated through numerical simulations and comparison with extended Kalman filter.

Keywords: Observers for nonlinear systems, Stability of nonlinear systems, Nonlinear system theory
Abstract: This paper addresses the problem of Simultaneous Localization and Mapping (SLAM) for rigid body systems in three-dimensional space. We introduce a new matrix Lie group (SE_{3+n}(3)), whose elements are composed of the pose, gravity, linear velocity and landmark positions, and propose an almost globally asymptotically stable nonlinear geometric observer that integrates Inertial Measurement Unit (IMU) data with landmark measurements. The proposed observer estimates the pose and map up to a constant position and a constant rotation about the gravity direction. Numerical simulations are provided to validate the performance and effectiveness of the proposed observer, demonstrating its potential for robust SLAM applications.

Keywords: Observers for nonlinear systems, Switched systems
Abstract: State of charge is a crucial metric for monitoring and controlling metal hydride storage tanks. However, designing model-based estimators for this system is complex due to the nonlinear, time-varying, and switched dynamics of the system. Additionally, this switched system presents operational modes where the system becomes unobservable. Considering these obstacles, this paper proposes a novel nonlinear switched observer for metal hydride storage tanks. Leveraging differential detectability and recent advances in contraction theory for switched systems, we establish sufficient conditions for the observer's convergence, assuming synchronized switching between the observer and the plant. Furthermore, to ensure this synchronization, we propose an unknown system dynamics estimator to identify the current operating mode. The proposed approach is validated through a numerical simulation.

Keywords: Observers for nonlinear systems, Uncertain systems
Abstract: This paper extends existing mathcal{H}_{infty} and L_1 interval observer designs for discrete-time (DT) and continuous-time (CT) nonlinear systems with bounded Jacobians and uncertainties by significantly reducing their computational complexity from mixed-integer programs to (generally) linear or semidefinite programs, without adding any conservatism. Our methods similarly employ a multiple-gain observer structure and are based on mixed-monotone decomposition and embedding systems to design correct-by-construction interval framers that inherently bound the true system state and whose framer error system is input-to-state stable. without needing additional positivity constraints. Moreover, we propose a novel CT mixed-monotone decomposition function that is less conservative than previous nonlinear bounding approaches, resulting in smaller framer errors. Finally, we demonstrate the effectiveness of the proposed interval observer on several DT and CT examples.

Keywords: Observers for nonlinear systems, Adaptive systems, Identification
Abstract: A new adaptive observer is proposed for a certain class of nonlinear systems with bounded unknown input and parametric uncertainty. Unlike most existing solutions, the proposed approach ensures simultaneous asymptotic convergence of the unknown parameters, state and perturbation estimates to an arbitrarily small neighborhood of the equilibrium point. The solution is based on the novel augmentation of a high-gain observer with the dynamic regressor extension and mixing (DREM) procedure enhanced with a perturbation annihilation algorithm. The aforementioned properties of the proposed solution are verified via numerical experiments.

Keywords: Stochastic systems, Stochastic control, Stochastic filtering
Abstract: In this paper we extend the classical dynamic programming approach of stochastic optimal control to decentralized stochastic discrete-time nonlinear dynamical optimal control problems. The control strategies have access to different information structures, and aim to optimize a common pay-off. We derive a set of generalized Hamilton-Jacobi-Bellman (HJB) equations, where the information states are pathwise conditional distributions that satisfy the Markov property, and play the role of the sufficient statistics for the different strategies. The main theorem sates that each control strategy, which is obtained from the solutions of the HJB equations is optimal, and a functional of its own pathwise information state. The paper settles the long standing open question of extending the classical dynamic programming approach to decentralized stochastic optimal control problems.

Keywords: Stochastic filtering, Computer aided learning, Nonlinear system identification
Abstract: This paper presents a discrete-time framework for parameter identification and state reconstruction of stochastic nonlinear systems with additive noise. Specifically, we build on the Physics-Informed Neural Network (PINN) framework and integrate a discrete-time Extended Kalman Filter (EKF) to manage stochastic noise. The EKF serves as the information model for the neural network, complemented by the likelihood function of the output innovation sequence to enhance accuracy. Numerical simulations on the Chialvo oscillator demonstrate the effectiveness of the proposed approach in handling noisy and nonlinear system dynamics.

Keywords: Stochastic filtering, Uncertain systems, Stochastic systems
Abstract: This article addresses sequential Bayesian filtering for nonlinear and stochastic dynamical systems. We extend a Galerkin-approach that was previously used for the prediction of non-Gaussian probability density functions, to incorporate linear and non-linear measurement updates. The proposed method results in a linear pipeline of prediction and update steps, which are computed as sparse matrix operations on the finite-dimensional coefficient vector. The performance of our approach is demonstrated in numerical experiments for nonlinear dynamical 2D- and 4D-systems, using results of a standard particle filter as reference, both in terms of accuracy and computational expenses.

Keywords: Observers for nonlinear systems, Stochastic filtering, Robotics
Abstract: This paper introduces an Unscented Kalman Filter (UKF) on the Special Euclidean group SE(2) based on optimal quantization, integrating it within the Invariant Linear Quadratic Gaussian (ILQG) regulator to solve trajectory tracking featuring high-angle variations. Traditional trajectory tracking methods like the Linear Quadratic Gaussian (LQG) regulator often face limitations due to linearization, especially under highly nonlinear model, high-angle maneuvers and/or stochastic uncertainties. The proposed approach overcomes these limitations by employing a sigma-point filter, the Unscented Kalman Filter on Lie groups (UKF-LG) enhanced with optimal quantization. This method improves the accuracy of the covariance computation during the filter’s prediction step, enabling better estimation performances to provide robust trajectory tracking in scenarios with significant angular deviations. Simulations demonstrate the effectiveness of the proposed Optimal Quantization-based Left-UKF-LG (OQ-Left-UKF-LG) over the IEKF and standard UKF-LG in complex tracking scenarios. The approach offers promising potential for applications requiring precise tracking in dynamic environments, such as autonomous mobile robots or Unmanned Aerial Vehicles (UAVs).

Keywords: Distributed estimation over sensor nets, Machine learning, Agents and autonomous systems
Abstract: We consider the problem of learning time-varying functions in a distributed fashion, where agents collect local information to collaboratively achieve a shared estimate. This task is particularly relevant in control applications, whenever real-time and robust estimation of dynamic cost/reward functions in safety critical settings has to be performed. In this paper, we adopt a finite-dimensional approximation of a Gaussian Process, corresponding to a Bayesian linear regression in an appropriate feature space, and propose a new algorithm, DistKP, to track the time-varying coefficients via a distributed Kalman filter. The proposed method works for arbitrary kernels and under weaker assumptions on the time-evolution of the function to learn compared to the literature. We validate our results using a simulation example in which a fleet of Unmanned Aerial Vehicles (UAVs) learns a dynamically changing wind field.

Keywords: Stochastic control, Stochastic filtering, Stochastic systems
Abstract: We formulate and analyze the operational definition called, control-coding capacity, C_{FB} in bits/second, of discrete-time nonlinear partially observable stochastic systems. C_{FB} is the maximum rate of encoding signals or messages into randomized controller-encoder strategies, and reproducing the messages at the output of the system using a decoder, with arbitrary small asymptotic error probability. In the first part of the paper, we characterize C_{FB} by an information theoretic optimization problem over randomized strategies, using posterior distributions of nonlinear filtering theory and information states. In the second part of the paper, we derive the formula of C_{FB}, for linear-quadratic Gaussian partially observable stochastic systems (LQG-POSSs). We express C_{FB} in terms of solutions of two Riccati equations of Kalman-filtering and a control Riccati equation. To derive C_{FB} we prove, (1) optimal randomized strategies consist of signalling, control, and estimation, and (2) a decentralized semi-separation principle holds between the control problem and the signalling and estimation problems.

Keywords: Identification for control, Stochastic systems, Switched systems
Abstract: This paper presents a novel approach to imitation learning from observations, where an autoregressive mixture of experts model is deployed to fit the underlying policy. The parameters of the model are learned via a two-stage framework. By leveraging the existing dynamics knowledge, the first stage estimates the control input sequences and hence reduces the problem complexity. At the second stage, the policy is learned by solving a regularized maximum-likelihood estimation problem using the estimated control input sequences. We further extend the learning procedure by incorporating a Lyapunov stability constraint to ensure asymptotic stability of the identified model, for accurate multi-step predictions. The effectiveness of the proposed framework is validated using two autonomous driving datasets collected from human demonstrations, showcasing its practical capability in modelling complex nonlinear dynamics.

Keywords: Predictive control for linear systems, Uncertain systems, Robust control
Abstract: In this paper, we study homothetic tube model predictive control (MPC) of discrete-time linear systems subject to bounded additive disturbance and mixed constraints on the state and input. Different from most existing work on robust MPC, we assume that the true disturbance set is unknown but a conservative surrogate is available a priori. Leveraging the real-time data, we develop an online learning algorithm to approximate the true disturbance set. This approximation and the corresponding constraints in the MPC optimisation are updated online using computationally convenient linear programs. We provide statistical gaps between the true and learned disturbance sets, based on which, probabilistic recursive feasibility of homothetic tube MPC problems is discussed. Numerical simulations are provided to demonstrate the efficacy of our proposed algorithm and compare with state-of-the-art MPC algorithms.

Keywords: Optimization, Optimization algorithms, Uncertain systems
Abstract: In this paper, we develop a two-stage data-driven approach to address the adjustable robust optimization problem, where the uncertainty set is adjustable to manage infeasibility caused by significant or poorly quantified uncertainties. In the first stage, we synthesize an uncertainty set to ensure the feasibility of the problem as much as possible using the collected uncertainty samples. In the second stage, we find the optimal solution while ensuring that the constraints are satisfied under the new uncertainty set. This approach enlarges the feasible state set, at the expense of the risk of possible constraint violation. We analyze two scenarios: one where the uncertainty is non-stochastic, and another where the uncertainty is stochastic but with unknown probability distribution, leading to a distributionally robust optimization problem. In the first case, we scale the uncertainty set and find the best subset that fits the uncertainty samples. In the second case, we employ the Wasserstein metric to quantify uncertainty based on training data, and for polytope uncertainty sets, we further provide a finite program reformulation of the problem. The effectiveness of the proposed methods is demonstrated through an optimal power flow problem.

Keywords: Safety critical systems, Optimization algorithms, Statistical learning
Abstract: Bayesian Optimization (BO) is a data-driven strategy for minimizing/maximizing black-box functions based on probabilistic surrogate models. In the presence of safety constraints, the performance of BO crucially relies on tight probabilistic error bounds related to the uncertainty surrounding the surrogate model. For the case of Gaussian Process surrogates and Gaussian measurement noise, we present a novel error bound based on the recently proposed Wiener kernel regression. We prove that under rather mild assumptions, the proposed error bound is tighter than bounds previously documented in the literature which leads to enlarged safety regions. We draw upon a numerical example to demonstrate the efficacy of the proposed error bound in safe BO.

Keywords: Robust control, Optimization, Machine learning
Abstract: Disturbance rejection in high-precision control applications can be significantly improved upon via online convex optimization (OCO). This includes classical techniques such as recursive least squares (RLS) and more recent, regret-based formulations. However, these methods can cause instabilities in the presence of model uncertainty. This paper introduces a safety filter for systems with OCO in the form of adaptive finite impulse response (FIR) filtering to ensure robust disturbance rejection. The safety filter enforces a robust stability constraint on the FIR coefficients while minimally altering the OCO command in the infty-norm cost. Additionally, we show that the induced ell_infty-norm allows for easy online implementation of the safety filter by directly limiting the OCO command. The constraint can be tuned to trade off robustness and performance. We provide a simple example to demonstrate the safety filter.

Keywords: Robotics, Autonomous robots, Uncertain systems
Abstract: With Edge Computing and 5G, industrial mobile robots will be able to offload computationally expensive algo- rithms such as sensor-based localization to the edge. However, multiple robots streaming large volumes of data over the network simultaneously will create network congestion, leading to high latencies and data loss, which can severely impact the robot operation. In this paper, we address this problem from a safety perspective by looking at how much communication can be reduced before risking safety violations due to increased localization uncertainty. We propose a co-design approach that adjusts communication and control jointly according to a requirement on localization uncertainty and show that by satisfying this requirement, safety can also be achieved. The method leverages a data-driven model of how the uncertainty depends on both communication and control. The performance of the optimization problem is evaluated experimentally on an ABB Mobile YuMi® Research Platform robot in both simulations and on hardware, and the results indicate that communication can be reduced without compromising safety.

Keywords: Electrical machine control, Robust control, LMI's/BMI's/SOS's
Abstract: This paper addresses the design and experimental validation of a state-feedback robust control law for electric motors with norm-bounded parametric uncertainties. The main objective is to propose a streamlined synthesis method that allows practitioners to take advantage of advanced control methods while only selecting a few, physically meaningful, tuning parameters. The synthesis is obtained by means of convex optimisation, specifically robust H2 and robust pole placement within linear matrix inequality (LMI) regions. The proposed formulation can be verified to be less conservative than pre-existing results from the literature. The effectiveness of the approach is also validated experimentally on a permanent magnet synchronous motor (PMSM) using a low-cost industrial microcontroller. A robust controller is designed to ensure that two distinct motors meet defined performance criteria. Experimental results comparing the performance of this robust controller with a non-robust design are presented.

Keywords: Electrical power systems, Nonlinear system theory, Fault detection and identification
Abstract: In this paper, we propose an active fault-tolerant control law, based on a fault estimation method and on differential flatness, for a three-phase inverter, connected to the grid by LCL filters. The system is vulnerable to multiple faults, therefore an active fault-tolerant control is required to preserve the electrical power conversion between renewable resources and the grid. First, the fault estimation is achieved using our recent algorithm [Laaziz et. al. 2024](based on a left inversion technique and on the super-twisting differentiator) and then, an active fault-tolerant control law based on a differential flatness approach is applied. In the paper, we provide a flatness analysis of the inverter and its LCL filters in healthy and faulty conditions. In particular, we show that the flat output is the same for both healthy and faulty systems, which is crucial for the active fault-tolerant control law design. This common flat output is computed thanks to a new model of the inverter and its output LCL filters. Several simulation results highlight the well-founded of the proposed method in healthy and faulty (symmetric and asymmetrical) conditions.

Keywords: Electrical power systems, Decentralized control, Constrained control
Abstract: This study proposes a unified decentralised control scheme for dc microgrids comprised of boost converters and constant power loads. A control theoretic approach is presented, where analytic tools are used to derive the system model and prove the existence of a positive invariant set under the system dynamics. This property guarantees that the effect of constant power load demand fluctuations does not destabilize the system and limits high transients that can otherwise damage the electronic components. A careful design of the controller structure leads to rigorously proving closed-loop system stability, as well as allows the derivation of necessary conditions on the controller parameters that guarantee the desired behaviour. Hence, the results are not affected by the network topology nor require knowledge of the network parameters, enhancing the overall implementability of the proposed design. The proposed control scheme is tested in a simulated scenario, verifying the theoretic results and demonstrating network resilience to sudden load perturbations.

Keywords: Electrical power systems, Distributed estimation over sensor nets, Reduced order modeling
Abstract: This paper addresses the monitoring of inter-area oscillations, which pose significant risks to power system’s stability. We present distributed data-driven algorithms to estimate inter-area oscillation modes and their corresponding left and right eigenvectors. Specifically, the power system is divided into coherent areas with local estimators that process real-time PMU data. The averaged state information are exchanged over a strongly connected communication network to distributively estimate the eigenvalues of the reduced order model. By using the estimated eigenvalues and leveraging the structure of the solution to the reduced-order dynamical model, the eigenvectors are then computed via solving a least square problem in a distributed manner. Simulations conducted on a 50-bus power system model comprising four areas demonstrate that the algorithm offers a scalable and effective solution for eigen-analysis in power systems.

Keywords: Electrical power systems, Nonlinear system theory, Decentralized control
Abstract: In this paper, a novel virtual inertia control algorithm is presented to enhance the frequency dynamic response of inverter-interfaced Distributed Energy Resources (DERs). By utilising the simple integral control structure, the proposed controller dynamically alters the value of the virtual inertia according to the value of the Rate of Change of Frequency (RoCoF), in order to decrease frequency oscillations and achieve fast convergence to the equilibrium point. Using vector field theory, it is shown that the controller structure imposes a lower limit on its output (virtual inertia) while the boundedness of the virtual inertia and the inverter frequency deviation around their nominal values are analytically proven using the ultimate boundedness theory. Furthermore, convergence to the desired equilibrium for the closed loop system consisting of an inverter-interfaced DER is proven using eigenvalue analysis. Finally, the efficacy of the the proposed approach in improving the frequency response of DERs during events of external disturbances (e.g. sudden variations in load demand) is validated for a 4-node and 4-DER microgrid created on the Typhoon HIL real-time simulator.

Keywords: Safety critical systems, Delay systems, Robust adaptive control
Abstract: Input delays affect systems such as wirelessly connected autonomous vehicles, and may lead to safety violations. One promising way to ensure safety in the presence of delay is to employ control barrier functions (CBFs), and extensions thereof that account for uncertainty: delay adaptive CBFs (DaCBFs). This paper proposes an online adaptive safety control framework for reducing the conservatism of DaCBFs. The main idea is to reduce the maximum delay estimation error bound so that the state prediction error bound is monotonically non-increasing. To this end, we first leverage both delay estimation and the estimation error bound of a disturbance observer to derive an upper bound on the current state prediction error from the previous state. Second, we design two nonlinear programs to update the maximum delay estimation error bound satisfying the obtained state prediction error bound previously and afterwards update the maximum error bound of the future state prediction used in DaCBFs. The proposed method ensures the maximum state prediction error bound with the delay estimation is monotonically non-increasing, yielding less conservatism in DaCBFs. We verify the proposed method in an automated connected truck application, showing that the proposed method reduces the conservatism of DaCBFs.

Keywords: Safety critical systems, Optimal control, Nonlinear system theory
Abstract: Articulated Frame Steering (AFS) vehicles are widely used in heavy-duty industries, where they often operate near operators and laborers. Therefore, designing safe controllers for AFS vehicles is essential. In this paper, we develop a Quadratic Program (QP)-based safety filter that ensures feasibility for AFS vehicles with affine actuator dynamics. To achieve this, we first derive the general equations of motion for AFS vehicles, incorporating affine actuator dynamics. We then introduce a novel High-Order Control Barrier Function (HOCBF) candidate with equal relative degrees for both system controls. Finally, we design a Parametric Adaptive HOCBF (PACBF) and an always-feasible, QP-based safety filter. Numerical simulations of AFS vehicle kinematics demonstrate the effectiveness of our approach.

Keywords: Safety critical systems, Constrained control
Abstract: Control invariant sets play an important role in safety-critical control and find broad application in numerous fields such as obstacle avoidance for mobile robots. However, finding valid control invariant sets of dynamical systems under input limitations is notoriously difficult. We present an approach to safely expand an initial set while always guaranteeing that the set is control invariant. Specifically, we define an expansion law for the boundary of a set and check for control invariance using Linear Programs (LPs). To verify control invariance on a continuous domain, we leverage recently proposed Lipschitz constants of LPs to transform the problem of continuous verification into a finite number of LPs. Using concepts from differentiable optimization, we derive the safe expansion law of the control invariant set and show how it can be interpreted as a second invariance problem in the space of possible boundaries. Finally, we show how the obtained set can be used to obtain a minimally invasive safety filter in a Control Barrier Function (CBF) framework. Our work is supported by theoretical results as well as numerical examples.

Keywords: Safety critical systems, Optimization, Nonlinear system theory
Abstract: This paper introduces an innovative safety-enhanced control framework that incorporates safety and stability constraints into nominal controllers through the quadratic programming approach. These constraints are defined using properly defined control Lyapunov functions and control barrier functions. In contrast to the traditional method, this framework enhances the nominal controller with stability and safety elements, eliminating the need for a controller redesign. Consequently, this increases the robustness and stability of the system while maintaining its performance. The approach is validated through a case study involving an overhead crane, which illustrates the real-world efficacy of this method.

Keywords: V&V of control algorithms, Safety critical systems, Computational methods
Abstract: Behavior Trees (BTs) have emerged as a powerful framework for structuring complex and long-horizon robotic behaviors and tasks. To ensure their correctness and safety, rigorous formal verification techniques are essential. Temporal Behavior Trees (TBTs) offer a promising approach by leveraging temporal logic to specify and verify desired properties of BT executions. However, traditional TBTs are primarily suited for offline analysis. We propose a framework for online monitoring of TBTs (OMTBT). Our method combines Signal Temporal Logic (STL) and propositional ternary logic to define quantitative semantics for partial BT executions. This enables real-time assessment of a BT's adherence to temporal specifications, allowing for timely feedback and adaptation. By providing a conservative approximation of the final TBT robustness, our online monitoring technique is well-suited for closed-loop control systems. We demonstrate its effectiveness through simulation on a robotic manipulation task, where it is used to define logic-based reward functions to guide reinforcement learning policy optimization, leading to improved performance and safety.

Keywords: Safety critical systems, Sliding mode control
Abstract: In this paper, we address robust stabilization and safety for a class of nonlinear systems. To achieve this, we employ sliding mode control and control barrier function approaches to ensure both objectives. Additionally, our proposed method guarantees high-order safety by integrating the control barrier function with a nonsingular fast terminal sliding variable and incorporating a robust reaching law in the safety design. To reduce the chattering problem, we used the super twisting algorithm for our nominal control algorithm. The robustness of the proposed strategy in the presence of multiple safety constraints is validated through numerical simulation examples.

Keywords: Autonomous systems, Observers for nonlinear systems, Sensor and signal fusion
Abstract: This work presents a fusion framework for pose estimation that integrates incremental pose measurements with IMU data within a multi-state (augmented state) extended Kalman filter. While relative transformations between consecutive camera (or LiDAR) frames can be estimated via numerous algorithms, fusing these incremental transformations with IMU measurements for enhanced pose accuracy demands a rigorous mathematical approach. We address this problem by employing the logarithmic map of SE(3) to represent incremental pose as a measurement model within the Kalman filter framework, with error states defined through the exponential map of SE(3). We validate the proposed method on the publicly available KITTI dataset using Monte Carlo runs, highlighting the effectiveness of our approach in improving trajectory accuracy.

Keywords: Autonomous systems, Switched systems, Linear systems
Abstract: This paper provides a benchmark example where a standard control system is modified with state-of-the-art safety guarantees implemented via a High-Order Control Barrier Function as safety filter. We will demonstrate that this modification ensures safety but induces unstable behaviour. In particular, we show that this instability is induced by a persistent switching between the original closed-loop dynamics and the CBF dynamics. Stability conditions will be needed to ensure that safety does not damage the stability of the closed-loop system.

Keywords: Large-scale systems, Autonomous systems, Distributed parameter systems
Abstract: In this paper, we address the large-scale shepherding control problem using a continuification-based strategy. We consider a scenario in which a large group of follower agents (targets) must be confined within a designated goal region through indirect interactions with a controllable set of leader agents (herders). Our approach transforms the microscopic agent-based dynamics into a macroscopic continuum model via partial differential equations (PDEs). This formulation enables efficient, scalable control design for the herders' behavior, with guarantees of global convergence. Numerical and experimental validations in a mixed-reality swarm robotics framework demonstrate the method’s effectiveness.

Keywords: Autonomous systems, Predictive control for linear systems, UAV's
Abstract: Automated inspection with Unmanned Aerial Systems (UASs) is a transformative capability set to revolutionize various application domains. However, this task is inherently complex, as it demands the seamless integration of perception, planning, and control which existing approaches often treat separately. Moreover, it requires accurate long-horizon planning to predict action sequences, in contrast to many current techniques, which tend to be myopic. To overcome these limitations, we propose a 3D inspection approach that unifies perception, planning, and control within a single data-driven predictive control framework. Unlike traditional methods that rely on known UAS dynamic models, our approach requires only input-output data, making it easily applicable to off-the-shelf black-box UASs. Our method incorporates back-face elimination, a visibility determination technique from 3D computer graphics, directly into the control loop, thereby enabling the online generation of accurate, long-horizon 3D inspection trajectories.

Keywords: UAV's, Agents and autonomous systems, Autonomous systems
Abstract: Nowadays, unmanned aerial vehicles (UAVs) are increasingly utilized in search and rescue missions, a trend driven by technological advancements, including enhancements in automation, avionics, and the reduced cost of electronics. In this work, we introduce a collaborative model predictive control (MPC) framework aimed at addressing the joint problem of guidance and state estimation for tracking multiple castaway targets with a fleet of autonomous UAV agents. We assume that each UAV agent is equipped with a camera sensor, which has a limited sensing range and is utilized for receiving noisy observations from multiple moving castaways adrift in maritime conditions. We derive a nonlinear mixed integer programming (NMIP) -based controller that facilitates the guidance of the UAVs by generating non-myopic trajectories within a receding planning horizon. These trajectories are designed to minimize the tracking error across multiple targets by directing the UAV fleet to locations expected to yield targets measurements, thereby minimizing the uncertainty of the estimated target states. Extensive simulation experiments validate the effectiveness of our proposed method in tracking multiple castaways in maritime environments.

Keywords: Supervisory control, Machine learning, Automata
Abstract: Obtaining safety guarantees for reinforcement learning is a major challenge to achieve applicability for real-world tasks. Safety shields extend standard reinforcement learning and achieve hard safety guarantees. However, existing safety shields commonly use random sampling of safe actions or a fixed fallback controller, therefore disregarding future performance implications of different safe actions. In this work, we propose a predictive safety shield for model-based reinforcement learning agents in discrete space. Our safety shield updates the Q-function locally based on safe predictions, which originate from a safe simulation of the environment model. This shielding approach improves performance while maintaining hard safety guarantees. Our experiments on gridworld environments demonstrate that even short prediction horizons can be sufficient to identify the optimal path. We observe that our approach is robust to distribution shifts, e.g., between simulation and reality, without requiring additional training.

Keywords: Machine learning, Optimization, Hybrid systems
Abstract: We study the problem of approximating a model predictive controller (MPC) with learning models to facilitate real-time operation. In particular, we investigate how the use of a hybrid learning model can tighten the statistical learning bounds used for stability guarantees given by existing robust data-driven MPC approaches. We propose a hybrid learning framework with a finite set of state-dependent modes, each consisting of a supervised regression model. The mode-switching signal corresponds to a state space partition produced by solving a homotopy optimization problem that implicitly minimizes the Lipschitz constant of the regression model in each mode. The cardinality of the partition is decided by a bifurcation phenomenon, inducing a performance-complexity trade-off that is discussed. The proposed MPC approximation framework is validated on a nonlinear benchmark problem.

Keywords: Machine learning, Neural networks, Predictive control for nonlinear systems
Abstract: The use of Recurrent Neural Networks (RNNs) is interesting to obtain prediction models for control. Among them, Echo State Networks (ESNs) display advantages such as simplified training and implementation. However, its high dimensionality might become an obstacle for optimization and interpretability. For that reason, this paper presents a comparative study of Model Predictive Control (MPC) implementations using Echo State Networks and their reduced models. The original ESN model and a reduced model obtained through Proper Orthogonal Decomposition (POD) are evaluated in terms of model precision, control performance and computational cost, in the task of controlling a quadruple-tank process. The results indicate that, despite a slight decrease in precision due to model reduction, the reduced model provides a significant reduction in execution time with almost the same control performance, making it a viable alternative for real-time applications.

Keywords: Machine learning, Neural networks, Optimization
Abstract: The federated learning (FL) paradigm effectively distributes the training burden among several units, each possessing local datasets. This paper presents a novel privacy-preserving FL algorithm, FedSecure. Our approach is distinct from the state of the art as it redefines the conventional distributed optimization problem inherent in FL. Unlike traditional frameworks that assume a common weight vector as the global decision variable, our method introduces an equivalent constrained problem. Each agent maintains its own weight vector as the local decision variable under the constraint that these local weight vectors must be equal. This enables the dual decomposition method to solve the distributed optimization problem. A key advantage of FedSecure is its ability to eliminate the necessity of sharing both the initial weights and the updated network weight values of each agent with the server. This feature ensures that the information related to agents’ training samples remains impervious to potential state-of-the-art cyber espionage attempts, underscoring the robust security measures of our algorithm. We validate FedSecure on the MNIST and CIFAR-10 datasets and compare it to a differential privacy algorithm, in which artificial noise is added to parameters at the clients’ side before aggregating, namely, noising before model aggregation FL (NbAFL).

Keywords: Machine learning, Fault detection and identification, Communication networks
Abstract: Distributed learning (DL) uses multiple nodes to accelerate training, enabling efficient optimization of large-scale models. Stochastic Gradient Descent (SGD), a key optimization algorithm, plays a central role in this process. However, communication bottlenecks often limit scalability and efficiency, leading to increasing adoption of compressed SGD techniques to alleviate these challenges. Despite addressing communication overheads, compressed SGD introduces trustworthiness concerns, as gradient exchanges among nodes are vulnerable to attacks like gradient inversion (GradInv) and membership inference attacks (MIA). The trustworthiness of compressed SGD remains unexplored, leaving important questions about its reliability unanswered.

In this paper, we provide a trustworthiness evaluation of compressed versus uncompressed SGD. Specifically, we conducted empirical studies using GradInv attacks, revealing that compressed SGD demonstrates significantly higher resistance to privacy leakage compared to uncompressed SGD. In addition, our findings suggest that MIA may not be a reliable metric for assessing privacy risks in distributed learning.

Keywords: Optimization, Communication networks
Abstract: We consider distributed optimization as motivated by machine learning in a multi-agent system: each agent holds local data and the goal is to minimize an aggregate loss function over a common model, via an interplay of local training and distributed communication. In the most interesting case of training a neural network, the loss functions are non-convex and the high dimension of the model poses challenges in terms of communication and computation. We propose a primal-dual method that leverages second order information in the local training sub-problems in order to accelerate the algorithm. To ease the computational burden, we invoke a quasi-Newton local solver with linear cost in the model dimension. Besides, our method is communication efficient in the sense of requiring to broadcast the local model only once per round. We rigorously establish the convergence of the algorithm and demonstrate its merits by numerical experiments.

Keywords: Autonomous systems, Automotive, Neural networks
Abstract: In this paper, a Physics-Informed Neural Network (PINN) is presented for control purposes. The dynamics of the system are integrated into the training process of the neural network to improve training accuracy. In the control loop, an optimization-based algorithm is also used, which ensures the safe operation of the system. One of the main requirements against the optimization-based algorithm is the low computational cost.  Moreover, in the paper, a comparative analysis is presented of the conventionally trained and the PINN-based results for control purposes. The results are illustrated through a vehicle-oriented problem, considering parameter uncertainty and nonlinear effects.

Keywords: Automotive, Predictive control for nonlinear systems, Machine learning
Abstract: Trajectory planning for automated vehicles commonly employs optimization over a moving horizon — Model Predictive Control — where the cost function critically influences the resulting driving style. However, finding a suitable cost function that results in a driving style preferred by passengers remains an ongoing challenge. We employ preferential Bayesian optimization to learn the cost function by iteratively querying a passenger's preference. Due to increasing dimensionality of the parameter space, preference learning approaches might struggle to find a suitable optimum with a limited number of experiments and expose the passenger to discomfort when exploring the parameter space. We address these challenges by incorporating prior knowledge into the preferential Bayesian optimization framework. Our method constructs a virtual decision maker from real-world human driving data to guide parameter sampling. In a simulation experiment, we achieve faster convergence of the prior-knowledge-informed learning procedure compared to existing preferential Bayesian optimization approaches and reduce the number of inadequate driving styles sampled.

Keywords: Automotive, Optimal control, Safety critical systems
Abstract: This paper presents an uncertainty-responsive model predictive control (MPC) framework designed to ensure the safe operation of autonomous vehicles (AVs) in those environments like, e.g., urban traffic, where the predicted behavior of the surrounding road users (RUs) could be highly uncertain, thus leading to potential collisions between AV and RUs or an unacceptably over-conservative behavior of the AV.

In our framework, the collision-avoidance constraints are adjusted on-line to adapt the AV's behavior to the varying uncertainty of the RUs' prediction model, such that safety is preserved.

Simulations of urban and highway driving scenarios, constructed upon real-world data, show that the proposed approach avoid collisions in presence of unpredicted behaviors of the surrounding RUs.

Keywords: Automotive, Optimal control, Autonomous systems
Abstract: This paper introduces a novel minimum-order, multi-objective centralized MPC control architecture for autonomous vehicles. The contributions are fourfold. First, we propose a hierarchical architecture that integrates path-tracking, speed control, and stability control based on a minimum-order predictive model. Second, stability control is dynamically activated and relaxed using an adaptive weight, based on a stability index. Third, speed control employs an adaptive weight to improve robustness, reduce overshoot and oscillations, and enhance energy efficiency. Fourth, path-tracking control is enhanced with an adaptive weighting scheme that considers lateral error, road adherence, and curvature to ensure smooth convergence and prevent oscillations, while also improving accuracy under uncertain and low-adherence conditions. The architecture is validated in a joint simulation between Simulink/Matlab and SCANeR Studio vehicle dynamics simulator. Our findings demonstrate the effectiveness of the architecture in enhancing stability and comfort at low runtime, while maintaining path-tracking precision and speed control robustness, at high speeds, high curvature, and low adhesion.

Keywords: Automotive, Autonomous systems, Game theoretical methods
Abstract: This work proposes a safe control strategy for an autonomous vehicle to overtake a human-driven vehicle (HDV) using a predictive safety filter (PSF) mechanism that hierarchically combines an end-to-end Reinforcement Learning (RL) agent with a predictive controller. To create a more realistic RL environment, a Stackelberg game based on a first-principles model is employed to capture the HDV’s real-time response during overtaking rather than relying on a predefined empirical or purely statistical driver model. In the lower layer, a distributionally robust chance-constrained predictive controller is implemented to manage uncertainties in HDV behavior, ensuring robust safety guarantees. The effectiveness of the proposed synthetic controller is verified in a gym environment with comparisons against traditional schemes.

Keywords: Lyapunov methods, Stability of nonlinear systems
Abstract: In this paper we introduce a nonlinear bidirectional cruise controller that uses spacing and speed measurements from the preceding and following vehicles to decide on the appropriate control action (acceleration). We rigorously prove that the set of equilibrium points is globally asymptotically stable and provide KL estimates that guarantee uniform convergence to the said set. Finally, we show that the solutions of the closed-loop system are Lagrange stable and provide certain conditions under which we also have exponential convergence to the set of equilibrium points.

Keywords: Predictive control for linear systems, Identification, Behavioural systems
Abstract: By considering the class of constrained LTI systems with unknown time delays, we propose a set-theoretic direct data-driven predictive controller to provide closed-loop guarantees. In particular, first, starting from input/output data series, we propose a sample-based method to build N-step input output backward reachable sets. Then, we leverage the constructed family of backward reachable sets to derive a data-driven control law. The proposed method guarantees finite-time convergence and recursive feasibility, independent of objective function tuning. It requires neither explicit state estimation nor an explicit prediction model, relying solely on input-output measurements; therefore, unmodeled dynamics can be avoided. Finally, a numerical example highlights the effectiveness of the proposed method in stabilizing the system, whereas direct data-driven predictive control without terminal ingredients fails under the same conditions.

Keywords: Predictive control for linear systems, Linear systems
Abstract: The integration of control applications into cloud and edge expands the capabilities of modern control systems, but also introduces variability in shared resource availability and competition with other applications, posing new challenges for control design. This paper presents a multi-mode Model Predictive Control (MPC) framework tailored for resource-aware systems. By treating the controller period as a scalable parameter, our approach dynamically adjusts control accuracy and computational complexity in response to changing resource and state-space conditions. Unlike existing event- and self-triggered strategies, our multi-mode design allows users to actively manage trade-offs between computational load and control quality. We provide feasibility and stability guarantees for the proposed control framework and demonstrate its effectiveness in a simulated cart-pole system, showcasing significant improvements in computational resource efficiency without compromising control performance.

Keywords: Predictive control for linear systems, Automotive, Uncertain systems
Abstract: Self-driving vehicles rely on sensory input to monitor their surroundings and continuously adapt to the most likely future road course. Predictive trajectory planning is based on snapshots of the (uncertain) road course as a key input. Under noisy perception data, estimates of the road course can vary significantly, leading to indecisive and erratic steering behavior. To overcome this issue, this paper introduces a predictive trajectory planning algorithm with a novel objective function: instead of targeting a single reference trajectory based on the most likely road course, tracking a series of target reference sets, called a target funnel, is considered. The proposed planning algorithm integrates probabilistic information about the road course, and thus implicitly considers regular updates to road perception. Our solution is assessed in a case study using real driving data collected from a prototype vehicle. The results demonstrate that the algorithm maintains tracking accuracy and substantially reduces undesirable steering commands in the presence of noisy road perception, achieving a 56% reduction in input costs compared to a certainty equivalent formulation.

Keywords: Predictive control for linear systems, Distributed control, Process control
Abstract: This paper proposes a novel synthesis procedure to find adaptive terminal constraints for distributed economic model predictive control (DEMPC) for a class of linear systems subject to periodic disturbances. Ellipsoidal sets centered at optimal periodic state solutions, with time-varying radii, are used to construct the terminal constraints. A set of linear matrix inequality conditions are provided to update the radius of the terminal set at each time step to guarantee recursive feasibility. Finally, the effectiveness of the proposed method is demonstrated through simulation of a water distribution network.