Keywords: Large-scale systems
Abstract: The rapid proliferation of data in the new information era has increased the complexity of data-driven problems across various fields of science and engineering. This development has led to a paradigm shift in AI, moving towards unsupervised and self-supervised representation learning, as well as multimodal learning. Significant advancements have emerged not only in mainstream Natural Language Processing and Computer Vision but also in Earth observation applications. These advancements exploit the synergies between self-supervised learning and the expanded availability of High-Performance Computing (HPC) systems, resulting in the emergence of AI Foundation Models (FMs). Originating from the concept of building upon an existing 'foundation', these models are developed by training on large and diverse data sets. This training enables them to capture a broad spectrum of informative features, making them extremely versatile and applicable across multiple domains. This keynote will provide an overview of the current efforts toward FMs for Earth observation at the 'Remote Sensing Simulation and Data Lab' of the Icelandic HPC community, University of Iceland, which collaborates closely with the Jülich Supercomputing Centre at the Forschungszentrum Jülich in Germany. The presentation will highlight the necessary tools and efforts required for the development of FMs, showcasing interdisciplinary research that intersects AI, supercomputing, and remote sensing applications. This research aims to enhance our understanding of complex Earth processes and advance the development of Digital Twins of Earth.

Keywords: Nonlinear system theory
Abstract: Nonlinear systems can be lifted to higher dimensional spaces in which the dynamics evolve linearly, and thus, enabling their analysis and control. However, most of the common objectives focus on the linearity of the lifted dynamics and the subsequent reconstruction while ignoring whether these features of interest can actually be recovered from the measured outputs. To address this issue we need to develop an observability-aware Koopman lifting. Accordingly, in this paper, we introduce such a lifting that adds a short-horizon observability check alongside standard invariance and reconstruction terms. The design keeps roles separate: a linear output map for identifiability and a nonlinear reconstruction map for higher-fidelity reconstructions. Therefore, instead of verifying observability afterwards, we incorporate it directly into the training objective to actively shape the learned model. Applied on benchmark dynamical systems with nonlinear outputs, namely the Van der Pol and Duffing oscillators, we have demonstrated that the method yields a better posed lifted-to-output mapping, stronger identifiability, lower long-horizon error, and a more reliable spectral profile than invariance-only and fixed-dictionary baselines.

Keywords: Nonlinear system theory
Abstract: The Koopman operator framework enables global analysis of nonlinear systems through its inherent linearity. This study aims to clarify spectral properties of the Koopman operators for nonlinear systems with control inputs. To this end, we treat the inputs as parameters throughout this paper. We then introduce the Koopman operator for a parameterized dynamical system with a globally exponentially stable equilibrium point and analyze how eigenfunctions of the operator depend on the parameter. As a main result, we obtain a global linearization, which enables one to transform the nonlinear system into a finite-dimensional linear system, and we show that it depends continuously on the parameter. Subsequently, for a control-affine system, we investigate a condition under which the transformation providing a global bilinearization does not depend on the parameter. This provides the condition under which the global bilinearization for the control-affine system is independent of the parameter.

Keywords: Predictive control for nonlinear systems, UAV's, Safety critical systems
Abstract: This paper presents a Koopman-based model predictive control (MPC) framework for safe UAV navigation in dynamic environments using real-time LiDAR data. By leveraging the Koopman operator to linearly approximate the dynamics of surrounding objects, we enable efficient and accurate prediction of the position of moving obstacles. Embedding this into an MPC formulation ensures robust, collision-free trajectory planning suitable for real-time execution. The method is validated through simulation and ROS2-Gazebo implementation, demonstrating reliable performance under sensor noise, actuation delays, and environmental uncertainty.

Keywords: Predictive control for linear systems, Chemical process control, Identification for control
Abstract: This paper presents a deep Koopman-based Economic Model Predictive Control (EMPC) for efficient operation of a laboratory-scale pasteurization unit (PU). The method uses Koopman operator theory to transform the complex, nonlinear system dynamics into a linear representation, enabling the application of convex optimization while representing the complex PU accurately. The deep Koopman model utilizes neural networks to learn the linear dynamics from experimental data, achieving a 45% improvement in open-loop prediction accuracy over conventional subspace identification. Both analyzed models were employed in the EMPC formulation that includes interpretable economic costs, such as energy consumption, material losses due to inadequate pasteurization, and actuator wear. The deep Koopman EMPC and subspace EMPC are numerically validated on a nonlinear model of multivariable PU under external disturbances. These results demonstrate that the deep Koopmand EMPC achieves a 32% reduction in total economic cost compared to the subspace baseline.

Keywords: Lyapunov methods, Optimization, Stability of nonlinear systems
Abstract: Koopman operator-based methods enable data-driven bilinear representations of unknown nonlinear control systems. Accurate representations often demand significantly higher dimensions than the original system, making control design challenging. Control Lyapunov Functions (CLFs) are widely used for controller synthesis, with quadratic CLF candidates being the most common due to their simplicity. Yet, we show that this class is highly restrictive, especially when the state dimension is large: under mild conditions, their existence implies stabilizability of the bilinear system by a constant input---that is, the control remains fixed over time. We establish this result by formulating a quadratically constrained quadratic program (QCQP) that exactly characterizes valid CLFs. Since QCQPs are NP-hard, we propose a convex semidefinite relaxation that offers a sufficient validity condition. For single-input systems, we prove that a quadratic CLF requires constant control stabilizability, and empirically demonstrate that this extends to high-dimensional multi-input systems in many cases.

Keywords: Nonlinear system theory, Model/Controller reduction, Computational methods
Abstract: The Koopman operator and its data-driven approximations, such as extended dynamic mode decomposition (EDMD), are widely used for analysing, modelling, and controlling nonlinear dynamical systems. However, when the true Koopman eigenfunctions cannot be identified from finite data, multi-step predictions may suffer from structural inaccuracies and systematic bias. To address this issue, we analyse the first and second moments of the multi-step prediction residual. By decomposing the residual into contributions from the one-step approximation error and the propagation of accumulated inaccuracies, we derive a closed-form expression characterising these effects. This analysis enables the development of a novel and computationally efficient algorithm that enforces unbiasedness and reduces variance in the resulting predictor. The proposed method is validated in numerical simulations, showing improved uncertainty properties compared to standard EDMD. These results lay the foundation for uncertainty-aware and unbiased Koopman-based prediction frameworks that can be extended to controlled and stochastic systems.

Keywords: Predictive control for linear systems, Robust control, Linear systems
Abstract: State-of-the-art approaches of Robust Model Predictive Control (MPC) are restricted to linear systems of relatively small scale, i.e., with no more than about 5 states. The main reason is the computational burden of determining a robust positively invariant (RPI) set, whose complexity suffers from the curse of dimensionality. The recently proposed approach of Deadbeat Robust Model Predictive Control (DRMPC) is the first that does not rely on an RPI set. Yet it comes with the full set of essential system theoretic guarantees. DRMPC is hence a viable option, in particular, for large-scale systems. This paper introduces a detailed design procedure for DRMPC. It is shown that the optimal control problem generated for DRMPC has exactly the same computational complexity as Nominal MPC. A numerical study validates its applicability to randomly generated large-scale linear systems of various dimensions.

Keywords: Predictive control for linear systems, Observers for linear systems, Sliding mode control
Abstract: A fundamental issue in control applications is ensuring performance and stability in the presence of modeling uncertainties and external disturbances. In this regard, in recent years the integration of Sliding Mode Control (SMC) with Model Predictive Control (MPC) has proven particularly effective, combining the intrinsic robustness of SMC with MPC’s ability to handle constraints and provide optimal performance. In this context, Integral Sliding Mode Predictive Control (ISMPC) has emerged as a promising approach. Despite this practical relevance, relatively few studies in the literature address discrete-time, observer-based implementations of ISMPC when the full state is not available and noisy output measurements must be used. In particular, application of ISMPC to collocated electromechanical systems presents the issue of dealing with invariant zeros. In this work, three discrete-time observer implementations, namely, a standard discrete-time Sliding Mode Observer (DSMO), a DSMO variant named Direct Output Injection Estimator (DOIE), and an Unknown Input Observer (UIO), are analyzed and compared, when used in combination with ISMPC to control a simple two-mass system with friction. Under noisy measurement conditions, both the UIO and DOIE significantly outperform the DSMO; while the UIO provides maximum robustness, the DOIE delivers comparable performance with minimal tuning effort.

Keywords: Delay systems, Distributed parameter systems, Predictive control for linear systems
Abstract: This article deals with the implementation of predictor feedback, analog to the Smith Predictor, for state feedback control in state space representation. The desired control law, obtained using partial differential equations and backstepping control, contains an integral term that has to be approximated for implementation. In this article, we propose a new way to implement this control law using a dynamic controller. The control law is composed of a state feedback term and a dynamic term that approaches the integral term that has to be estimated for implementation. Using a Lyapunov functional, we provide sufficient conditions, in terms of a linear matrix inequality, to guarantee that the closed-loop system is stable when the proposed control law is applied. We use three examples, taken from the literature, to show the benefits of the proposed approach.

Keywords: Predictive control for linear systems, Identification for control
Abstract: A recently developed data-driven Kalman predictor requires offline measurement of the process disturbance; a requirement that is often unmet for many practical applications. We propose a solution that parametrizes the Kalman predictor exclusively using measured input and output data. The key idea is to use the innovations form which naturally accounts for the process disturbance and measurement noise into a single orthogonal stochastic process. Unlike process disturbances, the innovations process can be estimated directly from input-output data via a numerically efficient projection step. The performance of the method is demonstrated using a benchmark simulation.

Keywords: Predictive control for linear systems, Uncertain systems, Adaptive control
Abstract: We propose a novel flexible-step model predictive control algorithm for unknown linear time-invariant discrete-time systems. The goal is to asymptotically stabilize the system without relying on a pre-collected dataset that reveals the system's behavior in advance. In particular, we aim to avoid a potentially harmful initial open-loop exploration phase for identification, since full identification is often not necessary for stabilization. Instead, the proposed control scheme explores and learns the unknown system online through measurements of inputs and states. The measurement results are used to update the prediction model in the finite-horizon optimal control problem. If the current prediction model results in an infeasible optimal control problem, then persistently exciting inputs are applied until feasibility is reestablished. The proposed flexible-step approach allows for a flexible number of implemented optimal input values in each iteration, which is beneficial for simultaneous exploration and exploitation. A generalized control Lyapunov function is included into the constraints of the optimal control problem to guarantee stability. This way, the problem of optimization is decoupled from the problem of stabilization. For an asymptotically stabilizable unknown control system, we prove that the proposed flexible-step algorithm can lead to global convergence of the system state to the origin.

Keywords: Computational methods, Linear systems, Predictive control for linear systems
Abstract: In this paper, we revisit the computation of controlled invariant sets for linear discrete-time systems through a trajectory-based viewpoint. We begin by introducing the notion of convex feasible points, which provides a new characterization of controlled invariance using finitely long state trajectories. We further show that combining this notion with the classical backward fixed-point algorithm allows for the computation of the maximal controlled invariant set. Building on these results, we propose a model predictive control (MPC) scheme that guarantees recursive feasibility without relying on precomputed terminal sets. Finally, we formulate the search for convex feasible points as an optimization problem, yielding a practical computational method for constructing controlled invariant sets. The effectiveness of the approach is illustrated through numerical examples.

Keywords: Coverage control, Decentralized control, Statistical learning
Abstract: We present a novel decentralized algorithm for coverage control in unknown spatial environments modeled by Gaussian Processes (GPs). To trade-off between exploration and exploitation, each agent autonomously determines its trajectory by minimizing a local cost function. Inspired by the GP-UCB (Upper Confidence Bound for GPs) acquisition function, the proposed cost combines the expected locational cost with a variance-based exploration term, guiding agents toward regions that are both high in predicted density and model uncertainty. Compared to previous work, our algorithm operates in a fully decentralized fashion, relying only on local observations and communication with neighboring agents. In particular, agents periodically update their inducing points using a greedy selection strategy, enabling scalable online GP updates. We demonstrate the effectiveness of our algorithm in simulation.

Keywords: Decentralized control, Stochastic control, Game theoretical methods
Abstract: In a team decision problem, knowledge of an agent's action can leak its private information to others. Prior work enforced a weak form of privacy by restricting the set of allowed strategies, thus trading off optimality for privacy. We study privacy preservation in a Linear Quadratic Gaussian (LQG) team problem using information garbling, without constraining the strategies of the agents. A garbler noisily transforms agents' observations, while obeying the no-signaling communication constraint, resulting in a garbled team problem, that enjoys privacy at a small expense of team performance. Our main contributions are as follows. We formalize privacy constraints, including those that allow bounded leakage, and show that each privacy constraint can be obtained by a corresponding privacy-enabling garbling. For linear-Gaussian garblings, we derive conditions for the garbling to be both no-signaling and privacy-enabling. We show that such garblings can not improve team performance and derive tractable bounds on the performance-privacy trade-off.

Keywords: Optimization, Decentralized control, Energy systems
Abstract: This paper proposes a distributed operating framework for Collective Self-Consumption (CSC) energy communities in Spain, based on a hierarchical Alternating Direction Method of Multipliers (ADMM). The framework coordinates two interlinked layers: a local suggestion layer and a community coordination layer. At the local level, prosumer and consumer agents autonomously propose time-varying allocation coefficients reflecting their self-consumption preferences and access priorities to the photovoltaic (PV) pool. These proposals are constrained by anti-egoism limits and shared only through aggregate signals to preserve data level privacy. At the Energy Community Manager (ECM) layer, a centralized yet lightweight reconciliation problem computes feasible coefficients and jointly schedules the community battery, including charging, discharging, grid exports, and wholesale market sales. The formulation enforces technical constraints such as state-of-charge limits, grid-tie capacity, and charge/discharge exclusivity, while incorporating receiver-specific surplus variables for transparent revenue allocation under Spanish CSC regulation. The framework is evaluated on a real municipal energy community in Barcelona using day-ahead forecasts and measured data. Compared to static allocation, the proposed controller reduces surplus energy by 50.2%, increases community self-consumption by 5.3%, and improves total income by about 15%, confirming its scalability, data level privacy preservation, and regulatory compliance for distributed CSC operation with shared storage.

Keywords: Agents and autonomous systems, Observers for nonlinear systems, Decentralized control
Abstract: We propose decentralized k-hop Prescribed Performance State and Input Observers for heterogeneous multi-agent systems subject to bounded external disturbances. In the proposed input/state observer, each agent estimates the state and input of agents located two or more hops away by communicating only with 1-hop neighbors, while guaranteeing that the estimation errors satisfy predefined performance bounds. Theoretical analysis demonstrates that if a closed-loop controller with full state knowledge achieves the control objective and the estimation-based closed-loop system is set–Input to State Stable (set-ISS) with respect to the goal set, then the estimated states can be used to achieve the system objective with an arbitrarily small worst-case error governed by the accuracy of the states estimates. Simulation results are provided to validate the proposed approach.

Keywords: Energy systems, Decentralized control, Optimization
Abstract: This paper presents a hierarchical, grid-aware control framework for load shaping through price-based demand-side management, enabling real-time coordination between grid-level objectives and local operations with minimal communication. The proposed hierarchical architecture consists of flexibility agents, which mediate between the aggregator and local controllers by generating optimal signals. Using dynamic models that characterize the relationship between the price signal and the demand of flexible assets, an optimization algorithm is implemented at the flexibility agent level to determine price signals that ensure the aggregate demand follows the purchased demand. This configuration can shape the demand by requiring information about the aggregated reference demand, without accessing the demand of individual flexible assets. Simulation results demonstrate that the proposed framework effectively achieves load shaping for a fleet of flexible assets.

Keywords: Modeling, Communication networks, Decentralized control
Abstract: In high-bandwidth delay networks with small buffers, the fluid model of FAST TCP closely approximates the rate-based protocol. The rate-based protocol offers an alternative to the conventional window-based Transmission Control Protocol (TCP). We derive sufficient conditions for local stability of the rate-based protocol for users having heterogeneous round-trip delays. The analysis highlights the essential aspect regarding the scalability of the sufficient conditions as the network grows from a single to multiple bottlenecks. Our model considers routers with drop-tail buffers. We focus on intermediate and small buffers, as they reduce queuing delays. With intermediate buffers, optimizing protocol gain could be challenging due to tighter stability bounds. However, if the buffers are small, the conditions are decentralized. For stability, each user must adjust the protocol gain inversely proportional to its round-trip delay. This holds regardless of the number of bottleneck links and round-trip delays in the network. Furthermore, we corroborate our analytical insights through packet-level simulations of a single bottleneck link with a small buffer.

Keywords: Markov processes, Optimal control, Electrical power systems
Abstract: This paper presents a mean-field control approach for Piecewise Deterministic Markov Processes (PDMPs), specifically designed for controlling a large number of agents. By modeling the interactions of a large number of agents through an aggregate cost function, the proposed method mitigates the high dimensionality of the problem by focusing on a representative agent. The contribution of this work is the application of a PDMP-based mean-field control framework to the coordination of a large population of Thermostatically Controlled Loads (TCLs). Adapting this framework to TCLs requires incorporating a quality-of-service constraint ensuring that each agent’s temperature remains with high probability within a specified comfort range. To achieve this, an additional jump intensity is introduced so that agents are very likely to switch between heating and cooling modes when they reach the boundaries of their temperature range. This extension to TCLs is demonstrated through Water Heaters (WHs) control, with a decentralized algorithm based on a dual formulation and stochastic gradient descent. The numerical results obtained illustrate this approach on two examples (signal tracking and taking into account energy price).

Keywords: Optimal control, Optimization algorithms
Abstract: We present a model-free policy gradient method (PGM) for the linear quadratic Gaussian (LQG) control with high-probability, finite-sample convergence to an approximate stationary point via an emph{input-output history} (IOH) representation of the closed-loop system. First, we show that any dynamic output-feedback controller is equivalent to a static partial state feedback gain within finite-length IOH dynamics, reducing the search to an optimization over that gain. Next, we relax the LQG problem by injecting small noise into the input and adding a regularization term to the LQG cost to ensure coerciveness. We then show that a vanilla PGM for the relaxed problem converges to overline K with |nabla J(overline K)|_F le mathcal O(|epsilon|), where J is the original LQG cost and epsilon represents the above two perturbations. Moreover, based on a zeroth-order approximation, we derive a model-free implementation and establish high-probability, finite-sample convergence to such an mathcal O(epsilon)-stationary point. Numerical simulations show that controllers learned from input-output data approach an optimal LQG controller.

Keywords: Optimal control, Emerging control theory, Energy systems
Abstract: There is little agreement on the social objective that should guide the intergenerational allocation of an exhaustible resource. Thus, remaining agnostic about a broad class of possible societal goals, we propose assessing feasible consumption paths by a fairness index that measures the worst relative performance across generations. Each generation's performance ratio compares its attained welfare with the maximum welfare achievable for that generation in isolation. The resulting fairness criterion is robust with respect to a broad canonical class of social objectives (symmetric, increasing, and concave). The fair consumption path, which maximizes the fairness index subject to basic economic consistency, is characterized by three primitive requirements: implementability (a nonincreasing path), exhaustiveness (full use of the resource), and balancedness (equalized ratios across cohorts). Together, these conditions yield a unique allocation that is Pareto-efficient, Lorenz-undominated, and guarantees strictly positive consumption for all future generations. In the benchmark case of constant absolute risk aversion (CARA), the model admits a semi-closed-form solution, including an explicit value function and a unique fairness level. The proposed criterion provides a transparent standard, without the need for society to endorse any particular social welfare function, while guaranteeing even-handed treatment to all cohorts.

Keywords: Quantum control, Optimal control, Computational methods
Abstract: This paper presents a class of structure-preserving numerical methods for quantum optimal control problems, based on commutator-free Cayley integrators. Starting from the Krotov framework, we reformulate the forward and backward propagation steps using Cayley-type schemes that preserve unitarity and symmetry at the discrete level. This approach eliminates the need for matrix exponentials and commutators, leading to significant computational savings while maintaining higher--order accuracy. We extend the formulation to nonlinear Schrödinger and Gross-Pitaevskii equations using a Cayley-polynomial interpolation strategy. Numerical experiments on state-transfer problems illustrate that the CF-Cayley method achieves the same accuracy as high-order exponential or Cayley-Magnus schemes at substantially lower cost, especially for longtime or highly oscillatory dynamics. In the nonlinear regime, the structure--preserving properties of the method ensure stability and norm conservation, making it a robust tool for large-scale quantum control simulations. The proposed framework thus bridges geometric integration and optimal control, offering an efficient and reliable alternative to existing exponential-based propagators.

Keywords: Optimal control, Large-scale systems, Modeling
Abstract: We study the sparse stabilization of nonlinear multi-agent systems within a mean-field optimal control framework. The goal is to drive large populations of interacting agents toward consensus with minimal control effort. In the mean-field limit, the dynamics are described by a Vlasov-type kinetic equation, and sparsity is enforced through an ell_1–ell_2 penalization in the cost functional. The resulting non-smooth optimization problem is solved via a three-operator splitting (TOS) method that separately handles smooth, non-smooth, and constraint components through gradient, shrinkage, and projection steps. A particle-based Monte Carlo discretization with random batch interactions enables scalable computation while preserving the mean-field structure. Numerical experiments on the Cucker-Smale model demonstrate effective consensus formation with sparse, localized control actions, confirming the efficiency and robustness of the proposed approach.

Keywords: Predictive control for linear systems, Optimal control, Stochastic systems
Abstract: The consideration of stochastic uncertainty in optimal and predictive control is a well-explored topic. Recently Polynomial Chaos Expansions (PCE) have received considerable attention for problems involving stochastically uncertain system parameters and also for problems with additive stochastic i.i.d. disturbances. While there exist a number of open-source PCE toolboxes, tailored open-source codes for the solution of OCPs involving additive stochastic i.i.d. disturbances in julia are not available. Hence, this paper introduces the toolbox PolyOCP.jl which enables to efficiently solve stochastic OCPs for linear systems subject to a large class of disturbance distributions. We explain the main mathematical concepts between the PCE transcription of stochastic OCPs and how they are provided in the toolbox. We draw upon two examples to illustrate the functionalities of PolyOCP.jl.

Keywords: Robotics, Optimization, Optimal control
Abstract: Singularities pose a major challenge in the trajectory optimization of robotic manipulators. This paper presents a polytope-based method to identify singularity-free regions in the configuration space of both fixed-base and free-floating robots. The nonlinear singularity condition, derived from the determinant of the Jacobian matrix, is reformulated as a semi-infinite optimization problem and iteratively approximated by convex polytopes. The developed method is first validated on manipulators with different degrees of freedom and with fixed or free-floating bases, for which the singularities are known. The method is then applied to a spatial 6 DOF free-floating robot, for which no description of the singularities exists to date. The resulting linear polytope constraints are directly integrated into optimal control-based trajectory planning, demonstrating computationally efficient singularity avoidance. Additionally, it is shown how trajectory planning can be achieved around non-convex singularity surfaces through the use of multiple polytopes.

Keywords: Robotics
Abstract: End-effector (EE) based robotic systems are cost-effective and versatile solutions for rehabilitation, but they often fail to reproduce human joint trajectories accurately, especially when patient and therapist limb morphologies differ. This paper presents a Human Joint Trajectory Tracking (HJTT) algorithm for EE-based rehabilitation, which uses the patient’s arm kinematics to generate offline a trajectory in the robot joint space that minimizes the discrepancies between therapist-recorded and patient-executed joint movements. The proposed method has been experimentally validated on healthy subjects with different arm morphologies using a Franka Emika Panda robot and Xsens inertial measurement units (IMUs). Statistical comparison with conventional Cartesian Trajectory Tracking (CTT) demonstrates improved joint-angle tracking.

Keywords: Robotics
Abstract: In this paper, a bilateral position-position teleoperation framework is considered for the mechanical cutting task and a shared-control methodology is proposed for the incorporation of the cutting task constraints regarding the cutting tool’s motion. All task-related constraints are imposed directly to the follower’s tool via a proposed admittance dynamic model, prohibiting tool translational motions along the lateral to the heading direction and rotational motions about the heading direction, while preventing impact with the cutting object’s supporting surface. The constraints are satisfied autonomously via the follower’s control, thus minimizing the operator’s cognitive load. Haptic feedback is provided to the leader, informing the operator about actual tool forces and torques arising at the follower side from the tool’s contact with the environment and/or virtual forces and torques when the operator attempts to violate task constraints. Experiments validate the proposed approach in a teleoperation setup consisting of a 6-DOF Sigma haptic device (leader) and a 7-DOF KUKA robot (follower) equipped with a knife on its end-effector, for the teleoperated, mechanical cutting of a soft target object.

Keywords: Robotics, Mechatronics, Linear systems
Abstract: In large and heavy structures, vibrations arise during motion, posing significant challenges for precise manipulation. To accomplish the desired motion, control algorithms must effectively suppress these structural vibrations. In cutting-edge projects, such as remote maintenance of future fusion energy reactors (tokamaks), the manipulation of this type of structure is defined as a crucial task. This paper presents a control strategy to suppress transverse vibrations in flexible payloads during motion using a collaborative payload manipulation approach. Two different industrial robot arms are arranged in a leader–follower configuration for the manipulation strategy. The leader robot guides the motion with shaped velocity commands, while the follower robot ensures compliance with the estimated external forces applied by the leader on the payload through an admittance controller. Unlike existing methods, the proposed approach enables collaborative manipulation of heavier and larger flexible objects, addressing additional challenges such as vibration suppression and heterogeneous robot specifications. The dynamics of the leader–follower–payload system are modeled using an equivalent mass-spring-damper model, and it is shown that, with appropriate admittance parameters, the total energy of the system is passively dissipated. A stability proof is also provided. Numerical simulations validate the proposed method, and experimental results (see online video [1] ) demonstrate its effectiveness

Keywords: Robotics, Uncertain systems, Adaptive systems
Abstract: This paper proposes a multi-modal control scheme for soft robotic manipulators with hysteretic elasticity, which allows tracking time-varying reference signals. Long Short-Term Memory (LSTM) networks are employed to estimate hysteresis moments in configurable frequency bands of the reference signal. For this purpose, supervisory training data is obtained by smoothing differentiated acceleration and solving a moment balance equation offline. Subsequently, a multi-modal control scheme is developed, which automatically selects, based on the reference spectrogram, the suitable LSTM estimator to track a variable-frequency reference signal. The proposed controller demonstrates superior tracking performance with respect to a baseline controller with integral action, for high-frequency reference signals, and to its single-modal version, for low-frequency references.

Keywords: Flexible structures, Modeling, Sensor and signal fusion
Abstract: This paper presents a novel framework for estimating the position and orientation of the end-effector in flexible manipulators undergoing vertical motion using multiple inertial measurement units (IMUs), optimized and calibrated with ground truth data. The flexible links are modeled as a series of rigid segments, with joint angles estimated from accelerometer and gyroscope measurements acquired by cost-effective IMUs. A complementary filter is employed to fuse the measurements, with its parameters optimized through particle swarm optimization (PSO) to mitigate noise and delay. To further improve estimation accuracy, residual errors in position and orientation are compensated using radial basis function neural networks (RBFNN). Experimental results validate the effectiveness of the proposed intelligent multi-IMU kinematic estimation method, achieving root mean square errors (RMSE) of 0.00021~m, 0.00041~m, and 0.00024~rad for y, z, and theta, respectively.

Keywords: Linear systems, Behavioural systems, Computational methods
Abstract: In this paper, we study the intersection of several discrete time Linear Time Invariant (LTI) behaviors, which is autonomous. Our objective is to compute integer invariants (minimal number of states, lag, and structure indices) of this autonomous intersection behavior. Let Bi, i = 1, . . . , K be given LTI discrete time behaviors with kernel representations Bi(σ), i = 1, . . . , K with the intersection of Bi's = B being autonomous. Here, we propose a numerical algorithm based on the computation of a Greatest Common Right Divisor (GCRD) of polynomial matrices Bi(σ), i = 1, . . . , K to compute a full row rank kernel representation of B. This algorithm also facilitates to determine the minimal number of states and the lag of B directly from the computed GCRD. Further, using this computed GCRD, we propose an algorithm to generate trajectories of the restriction of B to finite length. These trajectories are then used to compute structure indices. The proposed algorithms are demonstrated using a few numerical examples.

Keywords: Linear systems, Reduced order modeling, Algebraic/geometric methods
Abstract: We derive upper and lower bounds for the eigenvalues of the controllability and observability Gramians, as well as for the Hankel singular values, of continuous-time single-input single-output LTI systems with diagonalizable state matrix. The analysis employs a Cauchy matrix formulation that yields explicit, easily computable bounds in terms of the state matrix eigenvalues. The results provide insight into how the Gramian and Hankel spectra decay with the system size. Illustrative examples on structured systems show how the derived bounds are able to capture distinct spectral-scaling regimes.

Keywords: Optimization, Linear systems, Computational methods
Abstract: In this manuscript, we present a framework for computing the nearest non-coprime polynomial set to a given set of univariate polynomials. Here, nearness is with respect to the Frobenius norm. First, we propose a closed-form, rank-one solution for the optimal perturbation required to make the polynomial set share a common specified root. This framework is then extended to structured problems of various kinds. We show that each of these structured problems can be solved by modifying the solution of the general case. Finally, we solve the more general problem of finding the smallest perturbation that moves a given set of univariate polynomials to a set of polynomials which are non-coprime. We extend these results to a variety of structured problems as well. Numerical examples are provided to demonstrate the efficacy of these methods.

Keywords: Uncertain systems, Reduced order modeling, Linear systems
Abstract: In this paper, we revisit the parametric/uncertain modeling through the lens of the interpolatory multivariate Loewner Framework (mLF). The mLF is exploited to generate models with unstructured and/or structured uncertainties, directly from data. We emphasize the versatility and interests of such approach by allowing uncertain modeling directly from (tensorized) data, without any prior knowledge on the underlying system (e.g. rational/irrational, order and complexity). The underlying ideas are illustrated on two examples: a rational and an irrational, time-delayed, one.

Keywords: Signal processing, Filtering, Linear systems
Abstract: A systematic approach for designing optimization problems to tune linear time-invariant (LTI) differentiators is presented. The key contribution is the formulation of optimization problems that preserve essential properties of continuous-time differentiators after discretization, while tailoring performance to specific application requirements using real-world measured signals. In contrast to existing methods, the proposed formulations do not require a reference measurement of the desired time derivative or numerical integration of filter outputs and can be applied to differentiators of arbitrary orders while explicity considering effects such as delays, phase distortion, and discretization errors. The approach encompasses differentiators with both finite-duration and infinite-duration impulse responses, including those defined by rational transfer functions, state-space representations, and convolution kernels realizing orthogonal projections. Various cost and constraint formulations are introduced to address different application scenarios. The effectiveness of the proposed formulations is demonstrated through a numerical study showing that well-designed optimization problems can be solved effectively using simple optimization algorithms, overcoming the limitations of previous manual tuning heuristics.

Keywords: Large-scale systems, Linear systems, Optimization
Abstract: Linear systems of equations are foundational in mathematics and widely used across many disciplines. Modern complexity makes solving these equations challenging for real-world problems. Decomposing large-scale static systems into smaller subsystems provides a beneficial approach for addressing challenges arising from system complexity. The efficacy of this strategy is further augmented when the resultant subsystems exhibit weak coupling, as such decomposition enables parallel processing, thereby contributing to an accelerated solution process. In this direction, an overlapping decomposition offers additional flexibility, especially when a weakly coupled decomposition is not feasible. This paper presents efficient methodologies for the weakly coupled and overlapping decompositions of linear equation systems. The methodology commences with the formulation of an optimization problem, grounded in graph theory, to achieve a weakly coupled decomposition. Subsequently, the proposed optimization problem is extended to accommodate an overlapping decomposition. Finally, the efficacy and performance of the proposed methodologies are validated by means of simulations performed on very large-scale integrated (VLSI) circuits.

Keywords: Distributed parameter systems, Fluid flow systems, Output regulation
Abstract: In this paper, we consider set-point output tracking for the Oseen linearisation of the incompressible Navier-Stokes equations around a stable equilibrium. We study the 2-D cylinder wake fluid model at low Reynolds numbers with boundary input and output, a model with mixed boundary conditions. We show that an integral controller tracks constant reference signals and rejects constant input disturbances. We illustrate the results using numerical simulations. These findings are a first step towards local output regulation of nonlinear flows.

Keywords: Distributed parameter systems
Abstract: The paper is concerned with the strict-feedback form of hyperbolic distributed-parameter systems. Such a system structure is well known to be the basis for the recursive backstepping control design for nonlinear ODEs and is also reflected in the Volterra integral transformation used in the backstepping-based stabilization of parabolic PDEs. Although such integral transformations also proved very helpful in deriving state feedback controllers for hyperbolic PDEs, they are not necessarily related to a strict-feedback form. Therefore, the paper looks at structural properties of hyperbolic systems in the context of controllability. By combining and extending existing backstepping results, exactly controllable heterodirectional hyperbolic PDEs as well as PDE-ODE systems are mapped into strict-feedback form. While stabilization is not the objective in this paper, the obtained system structure is the basis for a recursive backstepping design and provides new insights into coupling structures of distributed-parameter systems that allow for a simple control design. In that sense, the paper aims to take backstepping for PDEs back to its ODE origin.

Keywords: Distributed parameter systems
Abstract: The paper presents an approach to flatness-based control design for hyperbolic multi-input systems, building upon the hyperbolic controller form (HCF). The transformation into HCF yields a simplified system representation that considerably facilitates the design of state feedback controllers for trajectory tracking. The proposed concept is demonstrated for a Timoshenko beam and validated through numerical simulations, demonstrating trajectory tracking and closed-loop stability.

Keywords: Distributed parameter systems, Feedback linearization
Abstract: This paper introduces a novel approach to control a boundary-actuated semilinear parabolic system. To convey the core aspects of the design, an exemplary diffusion-reaction PDE is considered for simplicity. After a flatness-based state transformation into an infinite-dimensional system of ODEs in strict-feedback form, classical integrator backstepping is applied recursively. This enables the design of an exactly-linearizing state feedback of so-called flat coordinates, mapping the system into a linear target system with desired stability reserve. Based on the linear equations of the target system in closed loop, a controller is designed for trajectory tracking in the transformed coordinates. The flatness-based transformation and the resulting control law are traced back to the solution of an appropriate linear Cauchy problem that depends on the system state. An efficient numerical scheme is proposed to approximate the solution of the Cauchy problem and, thus, obtain an online approximation of the control law. Simulation results demonstrate the practicability of the design and illustrate the control performance.

Keywords: Distributed parameter systems, Stability of nonlinear systems, Lyapunov methods
Abstract: We study the control of a class of stable infinite-dimensional linear systems. Our main goal is to design a controller that makes the closed-loop system excitable in the sense of the system's sensitivity to external input pulses exceeding a given threshold. We achieve this with a dynamic feedback controller combining positive and negative feedback. As our main results we prove the well-posedness of the infinite-dimensional nonlinear closed-loop system and analyse the closed-loop stability properties for different parameter configurations of the controller. We illustrate the excitability properties in the case of one-dimensional beam and heat equations.

Keywords: Distributed parameter systems, Flexible structures, Linear systems
Abstract: This paper extends previous work on finite-difference schemes over staggered grids for infinite-dimensional port-Hamiltonian systems. In the one-dimensional setting, it generalizes the discretization approach originally developed for the wave equation to a broader class of systems characterized by interconnection operators that include both differential and non-differential terms, such as the Timoshenko beam equation. The paper then introduces a discretization strategy for the two-dimensional case that requires only two grids, thereby accommodating a wider range of systems, including those whose interconnection operators contain non-differential components, such as the Mindlin plate model.

Keywords: Computational methods, Power plants, Energy systems
Abstract: A ground-generation airborne wind energy system is considered. The device consists of a kite attached to an oscillating arm in the horizontal plane that generates electricity. The kite's trajectory is constrained to follow a prescribed lemniscate pattern. An implicit modeling approach is adopted where the control force required to maintain the figure-eight trajectory is computed. This system is reduced to a two-dimensional second-order ordinary differential equation. Its limit cycles are studied numerically, with specific equilibria serving as starting points to obtain physically admissible periodic motions. These cycles depend on several design parameters including line length, arm inertia, and braking coefficient, that are tuned to maximize the average power generated by the arm's oscillations. The system's behavior is further analyzed across a typical range of wind speeds, demonstrating robustness to wind variations while maintaining positive line tension throughout the cycle.

Keywords: Electrical machine control, Power electronics, Lyapunov methods
Abstract: An alternative doubly-fed induction generator (DFIG) wind turbine (WT) system formulation is introduced, that is fully compatible with the voltage modulated-direct power control (VM-DPC) approach. The presented model captures the complete dynamics of the system and enables the design of simple controllers that effectively regulate independent active and reactive power exchange with the main grid, without the need of a synchronization mechanism, i.e. a phase-locked-loop (PLL) block. Contrary to standard design procedures, the proposed power control design is based on PI-type regulators which are completely independent from system parameters and provide directly the controlled duty-ratio inputs to the converter configuration of the rotor circuit and thus avoiding a troublesome division by a dc-link voltage state variable. In addition, they feature an additional dynamic damping term in their structure that significantly enhances the WT system response during transient conditions. By utilizing important characteristics of the closed-loop system structure, an appropriate nonlinear stability analysis procedure is employed so as to verify stability and state convergence properties. The results obtained by conducting a thorough simulation procedure reveal an enhanced and stable dynamic response, completely in line with the theoretical analysis.

Keywords: Energy systems, Machine learning, Robust adaptive control
Abstract: We study heat pump control in buildings as a constrained reinforcement learning (RL) problem: minimize electrical energy while keeping indoor temperature within comfort bounds. We formulate the task as a constrained Markov decision process with temperature comfort constraints, argue and verify that optimal operation typically lies near the comfort boundary and that a smoothed log‑barrier variant of Soft Actor‑Critic (CSAC‑LB) exploits this structure, and report robustness to sensor noise and model mismatch. We release I4B, a lightweight simulator with a Gym‑style API and a built‑in MPC baseline to enable reproducible RL studies. On two building scenarios, CSAC‑LB balances comfort and energy at least on par with MPC while maintaining constraint satisfaction. Benchmarking against several baseline algorithms demonstrates CSAC-LB's efficiency in exploration and control performance.

Keywords: Energy systems, Optimization, Neural networks
Abstract: The increasing complexity in building energy systems, driven by renewables, fluctuating consumption, and real-time constraints, requires advanced control strategies. Traditional optimization-based energy management systems (EMS) are effective, however, they frequently have computational limitations mainly from cost function optimization each time step and increased modeling complexity, leading to restrict their application in real-time scenarios. To address this, a hybrid EMS framework is developed in this work, integrating physics-informed neural networks with recurrent deep leaning networks (PIDL) for optimal control actions. A rule-based (RB) adjustment layer is incorporated to ensure physical feasibility and adherence to system constraints. This method is compared to a non-linear controller. Both methods are deployed in residential buildings equipped with photovoltaic (PV), energy storage systems (ESS), and utility grid connection. The building can order and sell energy from both grid and ESS. The model considers real hourly PV and load data, and varying pricing conditions for the different energy sources including time of use tariffs for the grid energy. The evaluation considers three climate zones (hot-humid, hot-dry, and cold) and four seasons. As results, both methods met energy demand, while PIDL-RB achieved in most cases costs reduction, high ESS usage, and lower grid energy trading compared to non-linear controller, particularly in PV-rich seasons like spring and summer. Moreover, by embedding physical equations into the learning model, PIDL-RB simplifies system modeling (combining system model and control in one unit) and reduces computational time, enabling fast control without solving optimization problems at each step.

Keywords: Markov processes, Modeling, Energy systems
Abstract: This paper models the dynamics and levelized costs of utilizing a wave glider-based energy harvester for saltwater desalination and freshwater transport. The paper is motivated by the acute need for freshwater that often arises in coastal communities, especially in the aftermath of natural disasters such as hurricanes. Mobile wave energy harvesters, such as anchorless wave gliders, are well-suited for addressing this need. The literature already presents models of the dynamics of such wave gliders. However, the use of these models for assessing the levelized cost of water remains relatively unexplored. We address this gap by augmenting existing models of wave glider dynamics with a Markov chain representation of system construction, aging, repair, and retirement. The resulting Markov model captures both: (i) the system’s capital and operating expenses, as well as (ii) the associated rewards- namely, water production and transport. This approach makes it possible to compute the levelized costs of both water desalination and freshwater transport. Such computation is particularly valuable for determining the critical distance below which it is cheaper to transport readily available freshwater, and above which it is cheaper to desalinate ocean water.

Keywords: Optimization
Abstract: This work enhances the performance of a steady-state open-loop optimizer for the automatic operation of solar photo-Fenton plants, an advanced oxidation technology used for the removal of organic microcontaminants from urban wastewater. The proposed strategy dynamically adjusts the reagent dosage in response to real-time variations in solar irradiance, improving process robustness without requiring additional measurements. Simulations under both clear- and cloudy-day conditions showed reductions in process constraint violations of up to 78 % for microcontaminant removal, with only minor increases in reagent consumption (< 15 %) and operating cost (< 2 %). The strategy effectively mitigates the impact of irradiance fluctuations, demonstrating its potential as a practical and cost-efficient solution for the robust optimization of large-scale solar photo-Fenton systems.

Keywords: Fault detection and identification, Filtering, Delay systems
Abstract: Time-delays are known to have a detrimental effect on feedback systems. In the context of networked cyber-physical systems, delays can be injected by malicious adversaries. Detecting them early is an important challenge. This paper proposes a novel variation of Interacting Multiple Model filtering to detect delay injection attacks in feedback control systems, when hidden in an open loop setting. The detection scheme is formalised by treating delay as alternative modes of the system, and theoretical analysis of the stationary distribution informs a reduction to a three parameter model as well as the choices of hyper parameter values. The method is evaluated on a cruise control application, and shows detection within a few seconds and a low false alarm probability.

Keywords: Aerospace, Observers for nonlinear systems, Adaptive systems
Abstract: Magnetic Anomaly Navigation (MagNav) is a promising navigation method that uses the Earth’s crustal magnetic field as a position reference, offering a resilient alternative for aircraft positioning in environments where satellite navigation is not available. A key challenge for MagNav is performance degradation and potential navigation filter divergence when traversing magnetically flat regions (small or non-existent anomalies) where position observability is weak. This paper presents a robustified navigation algorithm based on an error-state Kalman filter (ESKF) on the subgroup of the special Euclidean group of three (SE2(3)). We compare static and innovation-based measurement covariance tuning strategies with a novel gradient-adaptive scheme that dynamically adjusts the assumed magnetometer measurement’s covariance based on the local magnetic field variability, which is known a priori from the magnetic anomaly map. Through Monte Carlo simulations, we demonstrate that while an optimistic static tuning gives the best results for a high-observability scenario, it fails in challenging low-gradient, low-velocity scenarios. In contrast, both dynamic noise adaptation methods and conservative static noise could improve convergence rates. A conservative fixed measurement noise assumption showed consistently superior robustness and low position error. These insights are paving the way for more resilient full-scale deployments of airborne MagNav systems.

Keywords: Neural networks, Fault tolerant systems, Aerospace
Abstract: Deep Reinforcement Learning (DRL) algorithms like the Twin Delayed Deep Deterministic Policy Gradient (TD3) are becoming popular in aerospace Attitude Determination and Control Subsystems (ADCS) for their high adaptability. However, deploying these complex controllers on safety-critical space hardware demands a rigorous analysis of their robustness and fault resilience. This paper investigates the trade-offs between model compression and fault tolerance for a TD3-based spacecraft attitude controller. We compare the effects of Magnitude-based and Discrete Cosine Transform (DCT) pruning, followed by a full quantization pipeline to generate deployable models. Resilience is assessed via fault injection campaigns modeling Single Event Upsets (SEUs) common in harsh space environments. Our results establish a clear framework for deploying robust DRL controllers, demonstrating that a combined pruning and quantization pipeline can reduce model size by over 95% while simultaneously making the policy up to 4x more resilient to transient hardware faults than its uncompressed counterpart.

Keywords: Fault detection and identification, Aerospace
Abstract: This paper proposes a novel robust enhancement to the nullspace-based fault detection filter design method. Robustness against model uncertainties is achieved by minimizing the H_infty norm of the residual filter. The problem is formulated as a semidefinite program with linear matrix inequalities based on the Bounded Real Lemma, enabling a systematic and computationally efficient solution. The resulting robust filters reduce the effects of model uncertainties in the residual generation while preserving fault sensitivity. This ultimately increases the performance of the fault detection filters, as either lower detection thresholds can be implemented to enable faster detection times, or robustness against false alarms is increased when the original threshold values remain unchanged. The effectiveness of the approach is demonstrated first for a set of randomly generated linear systems. Subsequently, the method is applied to a realistic small fixed-wing unmanned aircraft model. Both analyses confirm the improved robustness of the fault detection filter designs.

Keywords: Fault estimation, Observers for linear systems, Uncertain systems
Abstract: This paper investigates fault estimation for systems with intermittent measurements and unmatched faults. To handle output interruptions, an intermittent intermediate observer (IIO) is developed by combining intermittent measurements with an intermediate observer structure. A switching Lyapunov framework is established to analyze the estimation error under nonperiodic measurement loss. An input-to-state stability (ISS) bound is derived in the presence of bounded disturbances and faults, and exponential convergence is obtained in the disturbance- and fault-free case. Moreover, linear matrix inequality (LMI)-based conditions are provided for observer gain computation without requiring the matching condition. A quadrotor UAV example is used for validation. Simulation results show that the proposed observer can accurately estimate both states and faults despite intermittent output availability, with performance comparable to that of a continuous-output observer.

Keywords: Observers for linear systems, Fault estimation, Neural networks
Abstract: Connectivity in connected and autonomous vehicles (CAVs) introduces vulnerability to cyber threats such as false data injection (FDI) attacks, which can compromise system reliability and safety. To ensure resilience, this paper proposes a control framework that combines a nonlinear controller with an interval observer for robust state estimation under measurement noise. The observer bounds leader’s states, while a neural network-based estimator estimates the unknown FDI attacks in real time. These estimates are then used to mitigate FDI attack effects and maintain safe inter-vehicle spacing. The proposed approach leverages an idea of interval observer-based estimation and merges model-based and learning-based methods to achieve accurate estimations and real-time performance. MATLAB/Simulink results confirm resilient tracking, precise FDI attack estimation, and robustness to noise, demonstrating potential for real-world applications of cooperative adaptive cruise control (CACC) under cyber and physical attacks, disturbance, and bounded measurement noise.

Keywords: Adaptive control, Transportation systems, Intelligent systems
Abstract: E-bikes have the potential to reshape urban micro-mobility, making it more sustainable and agile. However, a key limitation must be addressed: the lack of assistance controllers capable of simultaneously accounting for the rider's fatigue status and the battery consumption.

Therefore, this work presents a novel mixed control strategy specifically designed to meet these two often conflicting objectives. Specifically, it consists of two layers; the lower one is composed of two independent controllers: one minimizing the rider’s heart rate, considered a reliable proxy for fatigue, and another aimed at reducing energy consumption. The higher layer, on the other hand, consists of a strategy to combine the two proposed control actions based on the usage scenario, thereby providing adaptability to this optimization approach. Preliminary evaluation in simulation and real-world scenarios demonstrates that the proposed strategy can reduce the rider’s heart rate with a negligible increase in energy consumption compared to actually employed fixed-assistance control, proving its potential to enhance user comfort and extend battery life.

Keywords: Linear systems, Optimization algorithms, Identification
Abstract: Analog electronics have gained interest in recent years as they have the potential to perform computations that are substantially faster and more energy-efficient when compared to conventional digital computers. Energy-based learning algorithms, such as contrastive learning, can be used to train analog computing systems using local learning rules. In this paper, a contrastive learning algorithm is developed for a network of adjustable linear resistors, ideal capacitors, and current sources. The dynamics of this network are governed by an ``energy'' function. Based on this energy function, we define a cost function as the difference between two energy states: a free state and a state associated with the training data. The cost function is proven to be convex and its gradient Lipschitz continuous, guaranteeing convergence of the contrastive learning algorithm for any step-size belonging to a specified set.

Keywords: Optimization, Energy systems
Abstract: Vapor compression cycles (VCCs) are characterized by nonlinear cyclic thermodynamic models and complex two-phase fluid behavior. We propose a gradient-based optimization approach for the evaluation of stationary VCC solutions in the presence of constraints. We solve the problem of integrating the nonlinear and discontinuous fluid properties and parameters by means of smooth hat functions which allow for twice continuously differentiable approximation, compact support, and non-oscillatory behavior. We validate the accuracy of the approximation by comparison with a high-fidelity model. Additional numerical results demonstrate the framework’s ability to generate optimal solutions that incorporate operational constraints.

Keywords: Process control, Mechatronics, Control over communication
Abstract: Directional drilling involves guiding a borehole along a predefined path with high precision. This process forms a distributed cyber-physical system in which surface and downhole subsystems are connected via the drill string. Automating this process is challenging because surface and downhole components communicate over a low-bandwidth, time-delayed channel. Together with process uncertainties, these constraints hinder accurate modeling and limit conventional model-based control. Existing automation focuses on downhole steering, while surface inputs, despite their influence on steerability, are typically adjusted manually. We propose an application-oriented, hierarchical dual-actuator control architecture that coordinates communication-limited steering control with fast but hard-to-predict surface actuation. The approach combines a conventional steering controller with a data-driven, surrogate-based two-degree-of-freedom surface controller and an event-based, communication-aware decision logic to determine when steering updates should be downlinked and when surface adjustments suffice. High-fidelity simulations demonstrate improved reference tracking and fewer steering downlinks compared to steering-only automation, highlighting the benefits of coordinated surface–steering actuation under realistic drilling constraints.

Keywords: Flexible structures, Delay systems
Abstract: Absolute Vibration Suppression (AVS) is a control method for flexible structures governed by the wave equation. Such system may be a rotating shaft, a rod in tension or a taught string. AVS is based on an accurate, infinite dimension, transfer function, relating arbitrary actuation and measurement points, with general boundary conditions. Those transfer function consists of time delays due to the wave motion and low order rational terms which correspond to the reflection from the boundary. That compact mathematical representation, which also has a clear physical interpretation, was utilized to develop the AVS controller which is a collocated, rate to force feedback that completely eliminates the vibration. The current paper extends the AVS strategy in several directions. First it examines the disturbance rejection capabilities. Then it considers the MIMO case where several controllers are applied to the structure, raising questions such as where to place and how to design them. Finally, the realistic case of distributed control over a finite area of the structure is investigated.

Keywords: Maritime, Aerospace
Abstract: Gravity-aided navigation enables autonomous underwater positioning by exploiting the stability of gravity field signatures and the covert nature of passive gravimetric sensing. Fusing inertial navigation with gravity map matching yields an integrated solution that mitigates drift accumulation over extended missions. However, conventional Terrain Contour Matching (TERCOM) entails excessive computational cost and limited matching precision, constraining its practical deployment. This paper presents a two-stage gravity matching algorithm that combines particle swarm optimization (PSO) with affine transformation. PSO accelerates global search in the coarse matching stage, while affine transformation refines localization accuracy and corrects geometric distortion in the fine stage. Hardware-in-the-loop experiments demonstrate that the proposed method improves positioning accuracy by 20.2% and reduces computation time by 43.9% compared to traditional algorithm. The result validates its effectiveness for error compensation in adaptive search regions, supporting reliable passive navigation in complex underwater environments.

Keywords: Observers for linear systems, Uncertain systems, Identification for control
Abstract: We develop an interpolation-based framework for noisy linear systems with unknown system matrix with bounded norm (implying bounded growth or non-increasing energy), and bounded process noise energy. The proposed approach characterizes all trajectories consistent with the measured data and the prior bounds in a purely data-driven manner. This characterization enables data-consistency verification, inference, and one-step-ahead prediction, which can be leveraged for safety verification and cost minimization. Ultimately, this work represents a preliminary step toward exploiting interpolation conditions in data-driven control, offering a systematic way to characterize trajectories consistent with a dynamical system within a given class and enabling their use in control design.

Keywords: Observers for linear systems, Adaptive systems, Uncertain systems
Abstract: In practical applications, the efficacy of a control algorithm relies critically on the accurate knowledge of the parameters and states of the underlying system. However, obtaining these quantities in practice is often challenging. Adaptive observers address this issue by performing simultaneous state and parameter estimation using only input-output measurements. While many adaptive observer designs exist for continuous-time systems, their discrete-time counterparts remain relatively unexplored. This paper proposes an initial excitation (IE)-based adaptive observer for discrete-time linear time-invariant systems. In contrast to conventional designs that rely on the persistence of excitation condition, which requires continuous excitation and infinite control effort, the proposed method does not require excitation for infinite time, thus making it more suitable for stabilization tasks. We employ a two-layer filtering structure and a normalized gradient descent-based update law for learning the unknown parameters and the initial state. We also introduce modified regressors to enhance information extraction from input-output data. Rigorous theoretical analysis guarantees bounded and exponentially converging estimates of both states and parameters under the IE condition, and simulation results validate the efficacy of the proposed design.

Keywords: Observers for linear systems, Linear systems
Abstract: We develop observer design over hypercomplex quaternions in a characteristic-polynomial-free framework. Using the standard right-module convention, we derive a right observable companion form and companion polynomial that encode error dynamics through right-eigenvalue similarity classes. We also give an Ackermann-type formula for real-coefficient target polynomials, where polynomial evaluation is similarity-equivariant. The resulting recipes place observer poles directly over quaternions and clarify when companion-coordinate updates and one-shot Ackermann formulas remain valid.

Keywords: Observers for nonlinear systems, Algebraic/geometric methods, Robotics
Abstract: This work addresses the problem of designing an equivariant observer for a first order dynamical system on the unit-sphere. Building upon the established case of unit bearing vector dynamics with angular velocity inputs, we introduce an additional linear velocity input projected onto the unit-sphere tangent space. This extended formulation is particularly useful in image-based visual servoing scenarios where stable bearing estimates are required and the relative velocity between the vehicle and target features must be accounted for. Leveraging lifted kinematics to the Special Orthogonal group, we design an observer for the bearing vector and prove its almost global asymptotic stability. Additionally, we demonstrate how the equivariant observer can be expressed in the original state manifold. Numerical simulation results validate the effectiveness of the proposed algorithm.

Keywords: Observers for linear systems, Stability of linear systems, Linear systems
Abstract: The paper shows that positive linear systems can be stabilized using positive Luenberger-type observers. This is achieved by structuring the observer as monotonically converging upper and lower bounds on the state. Analysis of the closed-loop properties under linear observer feedback gives conditions for positive observation that cover a larger class than previous observer designs. The results are applied to nonpositive systems by enforcing positivity of the dynamics using feedback from the upper bound observer. The setting is expanded to include stochastic noise, giving conditions for convergence in expectation using feedback from positive observers.

Keywords: Stochastic filtering, Sensor and signal fusion, Filtering
Abstract: Reliable outlier detection is essential for robust state estimation in systems that fuse inertial and loosely coupled GNSS position measurements. However, traditional chi-squared-based outlier tests perform poorly when applied to state estimates on the extended special Euclidean group, especially under relatively high attitude uncertainties. This shortcoming arises from their neglect of the nonlinear coupling between orientation and position uncertainties. To address these limitations, this work introduces a novel outlier detection method that employs a Laplace approximation of the innovation distribution, using the linearization point naturally obtained in the iterated Lie-group-based Extended Kalman Filter (EKF) update. Through both planar and spatial simulation studies, the method achieves outlier detection accuracy comparable to a previously developed, theoretically more accurate Gaussian mixture-based method, while being significantly faster and more numerically stable.

Keywords: Aerospace, Optimization algorithms
Abstract: We propose a differentiable programming-based control allocation framework for a single-chamber solid-propellant Divert and Attitude Control System. The method explicitly captures the nonlinear coupling between chamber pressure, nozzle throat areas, and thrust using physics-based differentiable models. Desired force and torque commands are mapped to actuator inputs via a constrained Levenberg–Marquardt optimization, where physical bounds are enforced through proximal projection, and gradients are computed using automatic differentiation. Closed-loop six-degree-of-freedom simulations indicate that the proposed framework is computationally efficient and tractable for onboard implementation and achieves precise attitude stabilization and minimal miss distance even under strong nonlinear coupling.

Keywords: Aerospace, Observers for nonlinear systems, Algebraic/geometric methods
Abstract: Rendezvous and docking between a chaser spacecraft and an uncooperative target, such as an inoperative satellite, require synchronization between the chaser spacecraft and the target. In these scenarios, the chaser must estimate the relative attitude and angular velocity of the target using onboard sensors, in the presence of gyroscope bias. In this work, we propose a cascade of Equivariant Filters (EqF) to address this problem. The first stage of the cascade estimates the chaser’s attitude and the bias, using measurements from a star tracker, while the second stage of the cascade estimates the relative attitude and the target’s angular velocity, using observations of two known, non-collinear vectors fixed in the target frame. The stability of the EqF cascade is theoretically analyzed and simulation results demonstrate the filter cascade’s performance.

Keywords: Aerospace, Machine learning, Neural networks
Abstract: This paper addresses the challenge of intelligent evasion for high-speed flight vehicles under limited measurements (only partial observations such as line-of-sight angles). This restriction transforms the problem into a Partially Observable Markov Decision Process (POMDP), where conventional reinforcement learning policies struggle to generalize to unseen scenarios. This paper proposes a novel meta-learning solution based on Algorithm Distillation (AD). The resulting AD model exhibits in-context learning, enabling zero-shot generalization to new threat scenarios without gradient updates. Simulations show the AD method achieves substantially higher success rates across diverse test scenarios than both the original PPO and an expert-distillation baseline, narrowing the gap between static RL policies and the demands of real-world evasion tasks.

Keywords: Aerospace, Linear time-varying systems, Optimal control
Abstract: A linear optimal control law for orbital station-keeping in the Earth-Moon Restricted Three Body Problem (R3BP) is developed via Linear Quadratic Regulator (LQR) theory. First, the cost function is established considering a periodic state-weight matrix, leveraging stability information of the target orbits retrieved through Floquet theory. Then, the resulting periodic Riccati differential equation is solved and local asymptotic stability guarantees are shown. Finally, the performance of the proposed LQR when tracking periodic orbits in the circular and elliptic R3BPs is analyzed numerically.

Keywords: Aerospace, Autonomous systems
Abstract: This paper presents an integrated guidance and control (IGC) framework for intercepting stationary targets at a prescribed impact time. A dual-controlled interceptor equipped with canard and tail aerodynamic surfaces is considered, in which fin deflection commands are directly designed to achieve precise target interception at a user-specified time. A dual-layer sliding manifold is employed to design the IGC strategy, ensuring that the relevant error variables converge to zero at a preassigned time instant. Consequently, impact-time–constrained interception is achieved regardless of the initial engagement geometry between the interceptor and the target. Furthermore, a weighted control allocation scheme is utilized to distribute the control effort between the canard and tail actuators. Numerical simulations demonstrate that the proposed method achieves accurate impact-time interception against stationary targets.

Keywords: Uncertain systems, Identification for control, Aerospace
Abstract: The rise of orbital debris demands autonomous proximity operations with uncooperative targets with uncertain orientation and inertia. This work proposes a reachability analysis framework for orientation dynamics on the Lie group SO(3) under parametric uncertainty. The method integrates sensitivity equations to compute exact flow Jacobians, enabling numerical estimates of the Lipschitz constant to characterize local trajectory expansion. The resulting reachable-sets over-approximations are provably correct up to the estimation error on the Lipschitz constant, avoiding the rapid bound inflation typical of classical conservative methods—particularly challenging on compact spaces like SO(3) where they can cover the entire domain over short time horizons. Tightness increases as the inter-sample distance decreases, a refinement achievable with minimal runtime impact through GPU-based parallelization. Validation on two ENVISAT case studies with uncertain inertia matrices shows stable Lipschitz estimates, accurate reachable-set characterization, and robustness over long horizons, demonstrating suitability for real-time analysis of tumbling rigid bodies in orbit.

Keywords: Predictive control for nonlinear systems, Uncertain systems, Autonomous systems
Abstract: This introductory talk presents how safe-by-design control is achieved through robust model predictive control (MPC). Designed to be accessible with minimal mathematical complexity, it focuses on the fundamental components of robust MPC, including robust predictions, optimization, and receding-horizon implementation. The talk covers basic theoretical results, highlighting how safety and stability are ensured with a proper design. Examples from diverse robotics applications illustrate the practical relevance and motivate the audience. The presentation concludes by highlighting the main challenges in the design. This sets the stage for the rest of the tutorial, explaining the structure and emphasizing the importance of each upcoming talk to achieve safe-by-design robot control.

Keywords: Predictive control for nonlinear systems, Uncertain systems, Autonomous systems
Abstract: This forms the core of the tutorial, providing a unified exposition on diverse techniques for uncertainty propagation of nonlinear robot dynamics. We cover trajectory-specific sum-of-squares approaches, contraction metrics, linearization-based techniques, system level synthesis (SLS), and sampling-based methods. We first provide intuition by explaining uncertainty propagation for linear dynamics and highlight the challenges arising due to nonlinear dynamics. he methods are contrasted based on which design decisions are made offline and which are optimized online during execution Robotics examples demonstrate the effect of these differences in practice. For simplicity, we focus primarily on the robust case with bounded model error, but also briefly highlight how methods naturally extend to stochastic noise and probabilistic guarantees.

Keywords: Predictive control for nonlinear systems, Uncertain systems, Autonomous systems
Abstract: This forms the core of the tutorial, providing a unified exposition on diverse techniques for uncertainty propagation of nonlinear robot dynamics. We cover trajectory-specific sum-of-squares approaches, contraction metrics, linearization-based techniques, system level synthesis (SLS), and sampling-based methods. We first provide intuition by explaining uncertainty propagation for linear dynamics and highlight the challenges arising due to nonlinear dynamics. he methods are contrasted based on which design decisions are made offline and which are optimized online during execution Robotics examples demonstrate the effect of these differences in practice. For simplicity, we focus primarily on the robust case with bounded model error, but also briefly highlight how methods naturally extend to stochastic noise and probabilistic guarantees.

Keywords: Uncertain systems, Machine learning, Predictive control for nonlinear systems
Abstract: This talk focuses on retrieving the model and the characterization of the model accuracy from data, which is a key step in the robust design. A variety of methods will be studied, highlighting their difference in setup (stochastic, robust, partial measurement), computational complexity, and theoretical guarantees (asymptotic vs. exact finite-sample). This includes set-membership estimation, maximum likelihood estimation and Gaussian process models. The talks highlights how data-driven uncertainty quantification enables an automated and modular design of safe-by-design controllers. It will also highlight some of the specific challenges for uncertainty quantification that arise from perception-based control, i.e., when online processing of image data provides a critical measurement signal in the control loop.

Keywords: Statistical learning, Uncertain systems, Predictive control for nonlinear systems
Abstract: Accelerated by rapid advances in machine learning and AI, there has been tremendous success in the design of learning-enabled autonomous systems in areas such as autonomous driving and robotics. These exciting developments are accompanied by new fundamental challenges that arise regarding the safety and reliability of these increasingly complex systems due to imperfect learning, system unknowns, and uncertain environments. Conformal prediction (CP) — a statistical tool for uncertainty quantification — has gained popularity due to its ability to deal with these challenges. However, CP-based safety guarantees assume i.i.d. data, an assumption that is violated when system changes induce shifts in the underlying data distribution. In this tutorial, I will provide new insight to design safe controllers under distribution shifts using robust CP. I will begin by advocating for the use of CP due to its simplicity, generality, and efficiency as opposed to existing optimization-based verification techniques. I will then provide an introduction to CP and summarize existing work that uses CP to design probabilistically safe controllers in dynamic environments. Subsequently, we will look into interactive settings where the system’s behavior may change the environment's behavior, and vice versa. This circular dependency creates an interaction-driven distribution shift that invalidates existing safety guarantees. To deal with this chicken-and-egg problem, we propose an iterative framework that episodically updates the controller while robustly maintaining safety guarantees by quantifying the potential impact of a controller update on the environment's behavior. We realize this via adversarially robust CP where we perform a regular CP step in each episode using observed data under the current controller, but then transfer safety guarantees across controller updates by analytically adjusting the CP result to account for distribution shifts.

Keywords: Predictive control for nonlinear systems, Autonomous robots, Uncertain systems
Abstract: This talk investigates the application of MPC for mobile robot motion planning in human-centric environments. As autonomous systems---such as self-driving cars, vessels, and drones---increasingly operate alongside people, they must effectively manage uncertainties arising from both external sources (e.g., human behavior) and internal constraints (e.g., hardware and software limitations). We first present two approaches to handle human-induced uncertainties: a scenario-based design combined with MPC, and geometrical approaches to define the MPC navigation envelope. Finally, we demonstrate how robust MPC arguments can be used to overcome implementation constraints on resource-limited robots, such as drones. We will conclude with a discussion on the limitations of current methods and how to overcome them with tools taken from game theory and machine learning.

Keywords: Energy systems
Abstract: Demand response programs have traditionally focused on controlling electric loads to follow price signals, provide ancillary services to the grid, or reduce demand during periods of scarcity. While these programs are essential for operating power systems with high shares of renewable energy, it is difficult to explicitly link participation in demand response to measurable reductions in carbon emissions. At the same time, a growing group of electricity consumers—from hyperscale computing companies to individual households—can be described as “carbon-sensitive.” These consumers are willing to adapt their real-time electricity use based on the carbon intensity of the grid, much like price-sensitive consumers respond to fluctuations in electricity prices. We refer to this practice as carbon-aware load control. In this talk, we examine the carbon intensity signals commonly used to guide carbon-aware load control and demonstrate that they can sometimes lead to counterintuitive or even counterproductive impacts on overall grid emissions. We then propose an approach for integrating consumer carbon preferences into electricity markets to better align individual actions with overall grid emissions outcomes.

Keywords: Autonomous robots, Machine learning
Abstract: Autonomous robots that can assist humans in situations of daily life have been a long standing vision of robotics, artificial intelligence, and cognitive sciences. A first step towards this goal is to create robots that can learn tasks triggered by environmental context or higher level instruction. However, learning techniques have yet to live up to this promise as only few methods manage to scale to high-dimensional manipulator or humanoid robots. In this talk, we investigate a general framework suitable for learning motor skills in robotics which is based on the principles behind many analytical robotics approaches. To accomplish robot reinforcement learning learning from just few trials, the learning system can no longer explore all learn-able solutions but has to prioritize one solution over others – independent of the observed data. Such prioritization requires explicit or implicit assumptions, often called ‘induction biases’ in machine learning. Extrapolation to new robot learning tasks requires induction biases deeply rooted in general principles and domain knowledge from robotics, physics and control. Empirical evaluations on a several robot systems illustrate the effectiveness and applicability to learning control on an anthropomorphic robot arm. These robot motor skills range from toy examples (e.g., paddling a ball, ball-in-a-cup) to playing robot table tennis, juggling and manipulation of various objects.

Keywords: Adaptive control, Machine learning, Intelligent systems
Abstract: Model-based reinforcement learning (MBRL) with autoregressive world models suffers from compounding errors, whereas diffusion world models mitigate this by generating trajectory segments jointly. However, existing diffusion guides are either policy-only—discarding value information—or reward-based, which becomes myopic when the diffusion horizon is short. We introduce Advantage-Guided Diffusion for MBRL (AGD-MBRL), which steers the reverse diffusion process using the agent’s advantage estimates so that sampling concentrates on trajectories expected to yield higher long-term return beyond the generated window. We develop two guides: (i) Sigmoid Advantage Guidance (SAG) and (ii) Exponential Advantage Guidance (EAG). We prove that a diffusion model guided through SAG or EAG allows us to perform reweighted sampling of trajectories with weights increasing in state–action advantage—implying policy improvement under standard assumptions. Additionally, we show that the trajectories generated from AGD-MBRL follow an improved policy (that is, with higher value) compared to an unguided diffusion model. AGD integrates seamlessly with PolyGRAD-style architectures by guiding the state components while leaving action generation policy-conditioned, and requires no change to the diffusion training objective. On MuJoCo control tasks (HalfCheetah, Hopper, Walker2D and Reacher), AGD-MBRL improves sample efficiency and final return over PolyGRAD, an online Diffuser-style reward guide, and model-free baselines (PPO/TRPO), in some cases by a margin of 2x. These results show that advantage-aware guidance is a simple, effective remedy for short-horizon myopia in diffusion-model MBRL.

Keywords: Optimization, Markov processes, Iterative learning control
Abstract: In this paper, we propose a second-order deterministic actor--critic framework in reinforcement learning. The method extends the classical deterministic policy gradient method to exploit curvature information of the performance function. Building on the concept of compatible function approximation for the critic, we introduce a quadratic critic that simultaneously preserves the true policy gradient and an approximation of the performance Hessian. A least-squares temporal difference learning scheme is then developed to estimate the quadratic critic parameters efficiently. This construction enables a quasi-Newton actor update using information learned by the critic, yielding faster convergence compared to first-order methods. The main contribution is the theoretical construction of a quasi-Newton compatible quadratic critic with improved convergence demonstrated through simulation studies.

Keywords: Machine learning, Autonomous robots, Neural networks
Abstract: Autonomous exploration of cluttered environments requires efficient exploration strategies that guarantee safety against potential collisions with unknown random obstacles. This paper presents a novel approach combining a graph neural network-based exploration greedy policy with a safety filter to ensure safe navigation goal selection. The network is trained using reinforcement learning and the proximal policy optimization algorithm to maximize exploration efficiency while reducing the safety filter interventions. However, if the policy selects an unfeasible action, the safety filter intervenes to choose the best feasible alternative, ensuring system consistency. Moreover, this paper proposes a reward function that includes a potential field based on the agent's proximity to unexplored regions and the expected information gain from reaching them. Overall, the approach investigated in this paper merges the benefits of the adaptability of reinforcement learning-driven exploration policies and the guarantee ensured by explicit safety mechanisms. Extensive evaluations in simulated environments, along with comparisons against established benchmarks, demonstrate that the proposed approach achieves efficient and safe exploration in cluttered settings.

Keywords: Constrained control, Neural networks, Robust control
Abstract: This work proposes a two-layered control scheme for constrained nonlinear systems represented by a class of recurrent neural networks and affected by additive disturbances. In particular, a base controller ensures global or regional closed-loop l_p-stability of the error in tracking a desired equilibrium and the satisfaction of input and output constraints within a robustly positive invariant set. An additional control contribution, derived by combining the internal model control principle with a stable operator, is introduced to improve system performance. This operator, implemented as a stable neural network, can be trained via unconstrained optimisation on a chosen performance metric, without compromising closed-loop equilibrium tracking or constraint satisfaction, even if the optimisation is stopped prematurely. In addition, we characterise the class of closed-loop stable behaviours that can be achieved with the proposed architecture. Simulation results on a pH-neutralisation benchmark demonstrate the effectiveness of the proposed approach.

Keywords: Machine learning, Fuzzy systems, Intelligent systems
Abstract: This paper introduces a fuzzy reinforcement learning framework, Enhanced-FQL(lambda), that integrates novel Fuzzified Eligibility Traces (FET) and Segmented Experience Replay (SER) into fuzzy Q-learning with the Fuzzified Bellman Equation (FBE) for continuous control. The proposed approach employs an interpretable fuzzy rule base instead of complex neural architectures, while maintaining competitive performance through two key innovations: a fuzzified Bellman equation with eligibility traces for stable multi-step credit assignment, and a memory-efficient segment-based experience replay mechanism for enhanced sample efficiency. Theoretical analysis proves the proposed method convergence under standard assumptions. On the Cart--Pole benchmark, Enhanced-FQL(lambda) improves sample efficiency and reduces variance relative to n-step fuzzy TD and fuzzy SARSA(lambda), while remaining competitive with the tested DDPG baseline. These results support the proposed framework as an interpretable and computationally compact alternative for moderate-scale continuous control problems.

Keywords: Agents and autonomous systems, Machine learning, Autonomous robots
Abstract: In this paper, a combined control structure is presented integrating a classical Linear Quadratic Regulator (LQR) with a machine learning-based nonlinear compensator implemented through Kolmogorov-Arnold Networks (KANs). The main goal of the presented method is to ensure stability guarantees and computational efficiency, while the control performance level is increased. Firstly, a Reinforcement Learning (RL) framework is used to learn an optimal nonlinear policy over the operational range of the nonlinear system. Then, the trained agent is used as a teacher network for supervised knowledge distillation into a KAN. Finally, a stability condition is derived using the maximum Lipschitz constant of the KAN to guarantee the stability of the closed-loop LQR-KAN system. The results show that the KAN-based control structure achieves performance levels nearly the same as the RL policy, while the stability analysis is significantly simplified and the resulting neural network is less complex than the original teacher network. The whole approach is demonstrated through a double inverted pendulum problem.

Keywords: Predictive control for linear systems, Robust control, Uncertain systems
Abstract: Tube-based model predictive control (TBMPC) ensures constraint satisfaction under uncertainties, which is particularly useful for safety-critical applications. While conventional TBMPC methods primarily address additive uncertainties, dealing with parametric uncertainties has recently received more attention. However, applying TBMPC to high-dimensional systems often incurs significant computational complexity, underscoring the need for more scalable solutions. In this work, we propose a scalable TBMPC approach for discrete-time linear systems affected by both additive and parametric uncertainties by leveraging interval matrices and zonotopes. By fixing the order of the zonotopes, the online computational complexity of solving the proposed TBMPC problem scales polynomially with the system dimension. Numerical evaluations demonstrate the effectiveness of the proposed method in comparison with an alternative approach using ellipsoids.

Keywords: Predictive control for linear systems, Optimization algorithms, Computational methods
Abstract: The term constraint removal refers to methods that detect constraints that cannot become active in an MPC problem. Online constraint removal goes beyond finding constraints that can never become active (redundant constraints) by removing constraints online depending on the current state of the system. We extend a variant of online constraint removal that is based on comparing the current optimal cost function value to precomputed bounds. Essentially, we here extend this method from bounds that are independent of the current state to new bounds that depend on the current state and thus are far less conservative. Results are illustrated with the double integrator for illustration purposes and with computational experiments for a larger problem. The MPC computation time is reduced by up to 70% for these two cases.

Keywords: Predictive control for linear systems, Randomized algorithms, Optimization algorithms
Abstract: This paper presents a library-free, approximated Model Predictive Control (MPC) for systems with fast dynamics and limited computational resources. Conventional MPC relies on optimisation solvers, which can be computationally demanding and often require commercial solver licenses, making them less suitable for embedded systems. The proposed approach replaces solvers with a procedure that randomly samples a set of control sequences. The samples are evaluated for primal feasibility, and the best-performing suboptimal sequence is applied. Two improvements, variance-decaying and variance-adaptive approximated MPC, are introduced to direct the sampling procedure toward promising regions of the feasible solution space by sampling from a multivariate normal distribution. The variance-adaptive approximate MPC is designed to increase response to disturbances. Another advantage is the predictable and bounded computational effort, as the number of samples per control iteration can be directly conditioned by the available sampling time. Closed-loop stability and recursive feasibility are ensured through an auxiliary support controller. The resulting method remains solver-free, lightweight, and suitable for embedded implementations, while offering tunable performance–complexity trade-offs. Validation on a multivariable quadrotor model shows reliable control performance even under disturbances, and real-time experiments on Flexy2 PC–Arduino setup confirm its robustness on physical hardware

Keywords: Predictive control for linear systems
Abstract: This paper reviews the principle of the classical predictive functional control (PFC). For SISO processes, all possible closed-loop transfer functions are given to highlight why it is not suitable for unstable systems and is not even optimal in the input regulation response of stable systems. The paper proposes a decomposition of the prediction model into two stable systems, suitable for any SISO unstable system, and which also improves the input regulation response for SISO stable systems. No pre-stabilisation is necessary here, and the decomposition formulation is clearly stated. The strategy is illustrated by simulation results on first-order systems only.

Keywords: Stochastic control, Predictive control for nonlinear systems, Energy systems
Abstract: The main objective for wind turbine control in high winds is to keep the rotational speed within specified limits while keeping fatigue loads low. Until now this has been done by following a speed set point sufficiently close to stay within limits even in high turbulence. This paper presents a stochastic model predictive control (SMPC) approach where the speed does not follow a set point but is only kept within limits. The idea is that this leaves more flexibility for the controller to reduce loads. This approach is tested using the NREL5MW virtual turbine simulated with OpenFAST. Compared to the benchmark controller ROSCO it is demonstrated that a reduction of tower fatigue, fore aft and side side, of 15% and 21% can be obtained.

Keywords: Uncertain systems, Predictive control for linear systems, Robust control
Abstract: This paper presents a time-optimal Model Predictive Control (MPC) scheme for linear discrete-time systems subject to multiplicative uncertainties represented by interval matrices. To render the uncertainty propagation computationally tractable, the set-valued error system dynamics are approximated using a matrix-zonotope-based bounding operator. Recursive feasibility and finite-time convergence are ensured through an adaptive terminal constraint mechanism. A key advantage of the proposed approach is that all the necessary bounding sets can be computed offline, substantially reducing the online computational burden. The effectiveness of the method is illustrated via a numerical case study on an orbital rendezvous maneuver between two satellites.

Keywords: Aerospace, Cooperative control, Cooperative autonomous systems
Abstract: This paper proposes a cooperative integrated estimation-guidance framework for simultaneous interception of a moving non-maneuvering target using a team of uncrewed autonomous vehicles, assuming only a subset of vehicles are equipped with dedicated sensors to measure the target's states. Unlike earlier approaches that focus solely on either estimation or guidance design, the proposed framework unifies both within a cooperative architecture. To circumvent the limitation posed by heterogeneity in target observability, sensorless vehicles estimate the target's state by leveraging information exchanged with neighboring agents over a directed communication topology through a prescribed-time observer. The proposed approach employs true proportional navigation guidance (TPNG) as the baseline, which provides an exact time-to-go formulation against moving non-maneuvering targets and remains applicable across a wide spectrum of target motions. Furthermore, prescribed-time observer and controller are employed to achieve convergence to the target's true state and ensure consensus in time-to-go within set predefined times, respectively. Simulations demonstrate the effectiveness of the proposed framework under various engagement scenarios.

Keywords: Agents and autonomous systems, Optimization algorithms, Control over networks
Abstract: Aggregative cooperative optimization problems arise in distributed decision-making scenarios where each agent’s objective depends on its own decision as well as on an aggregate variable representing the collective system's behavior. Motivated by practical settings in which gradient information is unavailable, this paper proposes a randomized gradient-free algorithm, named ARGFree, for solving such problems. We establish that ARGFree converges in expectation to an approximate optimizer, where the approximation error originates from the use of a randomized gradient estimator. To the best of our knowledge, ARGFree is the first method in the literature capable of solving aggregative cooperative optimization problems without requiring gradient information. The effectiveness of the proposed algorithm is validated through robotic formation control experiments, including an implementation on a team of embedded systems based on Segway-type robots.

Keywords: Concensus control and estimation, Decentralized control, Autonomous systems
Abstract: Consensus algorithms enable cooperation in distributed systems, yet convergence often slows down when information is shared with a limited number of neighbors only. The problem is particularly relevant for UAV swarms, where broadcast links are unreliable and communication resources are limited. In this paper, we propose a lightweight share-and-relay consensus protocol in which, at each iteration, agents broadcast their states and neighbors re-broadcast the received packets with a one-step relay depth (RD), allowing two-hop information to propagate with a one-step delay while preventing flooding. This strategy enriches the standard consensus update with delayed two-hop information, effectively embedding an extended interaction topology on the original graph while maintaining fully distributed operation. We derive stability and convergence conditions that also account for stochastic packet dropouts, showing that the proposed scheme accelerates convergence compared to classical consensus while preserving bounded communication load and robustness to link failures.

Keywords: Concensus control and estimation, Cooperative control, Network analysis and control
Abstract: This paper addresses the problem of identifying leader nodes in semi-autonomous consensus networks from observed agent dynamics. Using the grounded Laplacian formulation, we derive graph-theoretic conditions that ensure the components of the Fiedler vector associated with leader and follower nodes are distinct. Building on this foundation, we employ the notion of relative tempo from [1] as an observable quantity that relates agents’ steady-state velocities to the Fiedler vector. This relationship enables the development of a data-driven algorithm that reconstructs the Fiedler vector, and consequently identifies the leader set, using only steady-state velocity measurements and without requiring knowledge of the network topology. The proposed approach is validated through numerical examples.

Keywords: Aerospace, Cooperative control, Lyapunov methods
Abstract: Protection of an aircraft using an active defending interceptor, when faced with a highly sophisticated attacking interceptor, is being popularly studied in the research community. In this paper, we propose guidance commands for the defender-aircraft team utilizing the geometric constraints imposed through a generalization of the triangle guidance. The design is carried out by considering the nonlinear engagement dynamics, and hence, problems arising from using the linearized dynamics are averted. We present simulation examples of various scenarios to validate the proposed guidance commands and compare the proposed acceleration requirement with the existing case.

Keywords: Aerospace, Linear systems, Cooperative control
Abstract: In this paper, we consider the coupling between the translational and rotational motions of a spacecraft under the influence of aerodynamic lift and drag for a spacecraft formation flying mission. Under the assumption of a circular reference orbit, we derive a set of coupled linear equations for the relative motion between two spacecraft, considering the variations in aerodynamic forces due to changes in attitude and relative velocity. The rotational dynamics are linearized about a reference attitude using Modified Rodrigues parameters for orientation representation. With control torque as the input, the controllability of the system is demonstrated in both in-plane and out-of-plane relative motions. The design of a Linear-Quadratic-Regulator-based controller, using the derived linearized dynamics, is presented to control the relative motion dynamics while simultaneously stabilizing the attitude dynamics. The performance of the proposed controller is demonstrated using the full set of nonlinear coupled orbit-attitude dynamics.

Keywords: Constrained control, Lyapunov methods, LMI's/BMI's/SOS's
Abstract: This paper presents a novel approach for the design of globally stabilizing linear static anti-windup compensators for input saturated closed-loop systems with exponentially stable plants. The proposed solution leverages a non-quadratic Lyapunov function involving sign-indefinite quadratic forms, which introduces additional degrees of freedom for synthesizing linear static anti-windup gains. The design is performed directly under semilinear global stability conditions, eliminating the need for iterative algorithms and initial solution guesses. Furthermore, an optimization criterion is included to minimize the control effort, expressed as the squared norm of the control signal, and penalize the cumulative saturation events, thereby providing some metrics on the anti-windup performance. Numerical examples demonstrate the effectiveness of the proposed strategy in achieving stability and improved performance under input constraints.

Keywords: Constrained control, Neural networks, Stability of nonlinear systems
Abstract: We study a control architecture for nonlinear constrained systems that integrates a performance-boosting (PB) controller with a scheduled Predictive Safety Filter (PSF). The PSF acts as a pre-stabilizing base controller that enforces state and input constraints. The PB controller, parameterized as a causal operator, influences the PSF in two ways: it proposes a performance input to be filtered, and it provides a scheduling signal to adjust the filter's Lyapunov-decrease rate. We prove two main results: (i) Stability by design: any controller adhering to this parametrization maintains closed-loop stability of the pre-stabilized system and inherits PSF safety. (ii) Trajectory-set expansion: the architecture strictly expands the set of safe, stable trajectories achievable by controllers combined with conventional PSFs, which rely on a pre-defined Lyapunov decrease rate to ensure stability. This scheduling allows the PB controller to safely execute complex behaviors, such as transient detours, that are provably unattainable by standard PSF formulations. We demonstrate this expanded capability on a constrained inverted pendulum task with a moving obstacle.

Keywords: Constrained control, System reconfiguration, Predictive control for linear systems
Abstract: This paper addresses the problem of inverting a static affine mapping subject to input constraints, i.e. finding all inputs in a prescribed input set that are mapped to a given output. A extensive review of the literature first reveals that this problem arises in many areas of control, which are generally considered as disconnected. Next, a comprehensive solution to the problem is presented in didactic manner. Finally, fully constructive results are proposed in the case where the input set is a polyhedron.

Keywords: Constrained control, Safety critical systems, Nonlinear system theory
Abstract: We study the control of nonlinear constrained systems via over-approximations. Our key observation is that the over-approximation error, rather than being an unknown disturbance, can be exploited as input-dependent preview information. This leads to the notion of informed policies, which depend on both the state and the error. We formulate the concretization problem -recovering a valid input for the true system from a preview-based policy- as a fixed-point equation. Existence of solutions follows from the Brouwer fixed-point theorem, while efficient computation is enabled through closed-form, linear, or convex programs for input-affine systems, and through an iterative method based on the Banach fixed-point theorem for nonlinear systems.

Keywords: Linear systems, Optimal control
Abstract: This paper investigates the guidance problem of missiles equipped with an off-axis seeker, focusing on satisfying non-standard field-of-view (FOV) constraints. The engagement scenario is reformulated in a relative coordinate frame, and the guidance law is derived based on a time-to-go polynomial approach without any linearization assumptions. Exact closed-form solutions are obtained for the guidance law, from which the lock-on conditions and the characteristics of the acceleration command are analytically examined. Extensive simulations are conducted to validate the performance of the proposed methods.

Keywords: Aerospace, Optimal control, Autonomous systems
Abstract: This paper proposes a Physics-Informed Interior Point Method (PI-IPM) for inequality-constrained optimal control. The method combines a PINN-based indirect solver for the barrier-augmented two-point boundary value problem with feasible-set mappings and an interior point homotopy for inequality path constraints. To improve practical convergence, we introduce a data-driven warm-start strategy by pretraining the network on trajectories generated using Sequential Convex Programming (SCP). The proposed framework is demonstrated on a representative spacecraft deorbit problem, and it was observed that the warm-started PI-IPM reduces solve time relative to a trapezoidal SCP baseline while producing feasible continuous-time trajectories of competitive quality. These results indicate the potential of PI-IPM as a warm-started indirect framework in repeated-solve settings over related problem instances.

Keywords: Robotics
Abstract: We present a geometry-driven embedded controller for quadruped locomotion that blends short step vectors into continuous trajectories and biases the center of mass (CoM) toward a geometry-derived reference point. The framework unifies stanceswing scheduling, safety envelopes, and a static CoM-based offset mechanism, allowing smooth transitions across standing, posture editing, and locomotion. Implementation on a microcontrollerdriven quadruped platform shows that the controller consistently converges the CoM towards geometric target and preserves phase continuity during both linear and turning motion, demonstrating embedded feasibility and phase-continuous locomotion under a lightweight geometric CoM guidance scheme.

Keywords: Observers for nonlinear systems, Mechatronics, Lyapunov methods
Abstract: This work presents the design and the corresponding stability analysis of a self-tuning external load observer for an electro-hydraulic actuator system. The proposed observer does not require the exact knowledge of the dynamical parameters of the actuator and ensures the asymptotic convergence of the external load estimate to the actual value, removing the need of an extra force sensor. An adaptive observation gain algorithm is also introduced to ease the tuning process. The stability and convergence of the observer is ensured via Lyapunov based arguments. Extensive numerical studies are performed to demonstrate the viability and effectiveness of the proposed method.

Keywords: Maritime, Mechatronics, Sensor and signal fusion
Abstract: Residual payload pendulation on offshore cranes can arise from diverse sources: load pickup, slewing, residual motion-compensation error and unmodeled structural dynamics that must be arrested before final placement. This paper presents an on-demand, vision-based swing suppression system that activates on operator request. A tip-mounted camera coupled with an IMU detects a hook-mounted marker, and a lightweight pendulum state estimator fuses geometry-based propagation with filtered vision updates at 200Hz, achieving >99% detection availability across four field sessions (4h of active operation), to recover the payload's swing amplitude, phase, and energy partition. A phase-aware planner schedules a single, quarter-period velocity pulse toward the next swing apex. Onshore field trials on a full-scale motion-compensated crane (Seaonics ECMC C25) with two payloads (500kg and 1500kg) across 51 maneuvers yielded a median sway amplitude reduction of 12.8dB (77%). The method is positioned relative to input shaping and continuous energy damping to clarify its role among existing approaches.

Keywords: Robotics, Mechatronics, Emerging control applications
Abstract: Controlling humanoid robot locomotion is in general challenging, while biological systems demonstrate adaptive and robust movement with minimal control effort. In that sense, human kinematics hold potential for locomotion controller design. Among various human movements, squatting serves as a fundamental behavior that integrates both stance and balance subfunctions of locomotion. This study investigates how observed human kinematics can be leveraged to control squatting motions in a humanoid robot.

We propose a bioinspired control scheme in which recorded human joint angles parameterize a reference trajectory that is tracked by simple low-level PD controllers. To isolate the effect of human kinematic structure, we deliberately avoid learning-based or optimization-heavy controllers and examine how far simple reference generation can reproduce human-like squatting. This scheme was implemented on a detailed simulation model of our humanoid robot. We explored the parameter space of joint gains to evaluate three key performance metrics: Stability, Efficiency, and Similarity.

Simulation results show that the kinematics-based reference generation, combined with simple closed-loop control tracking, is effective in producing human-like squatting behaviors. By tuning the gains, trade-offs among stability, efficiency, and similarity can be achieved to obtain balanced performance. This work demonstrates the feasibility of transferring human squatting principles to robotic systems with simple, yet efficient control schemes, and systematically evaluates the resulting behaviors within a structured performance framework.

Keywords: Servo control, Robotics, Optimization algorithms
Abstract: This research presents a practical, scalable frame- work for systematically characterizing and correcting spatial errors in cost-effective multi-axis Robots. With the growing industrial demand for deploying such systems in applications such as dimensional measurement, precise welding operations, and quality control, enhancing positional accuracy and trajectory fidelity has become increasingly vital. The proposed methodology integrates a high-precision optical measurement system (Keyence XM-5000) to assess geometric deviations over the Robot’s workspace while replicating realistic industrial measurement conditions. Spatial errors are quantified through the construction of an error array, measured using a high-accuracy (<3 μm) optical measurement system. Building upon this characterization, a two-layer compensation strategy is introduced: a grid-based interpolation algorithm first estimates and corrects static positioning errors across the Robot’s workspace, enabling high-speed open-loop motion. A high-precision optical measurement system (Keyence XM-5000) is used to perform real-time fine servoing to achieve micron-level accuracy. This research tries to bridge the performance gap between low-cost collaborative Robots and conventional precision metrology systems.

Keywords: Robotics, Adaptive control, Constrained control
Abstract: Robot manipulators increasingly operate in settings where fast motion, uncertain dynamics, and strict actuator limits must be handled simultaneously. While adaptive control improves trajectory tracking performance under uncertainties, it may violate input constraints during transients or high-speed motion. In this paper, the problem of input-constrained adaptive control of manipulators is addressed. Motivated by the observation that adaptive robot control and path-velocity control can be integrated for path-tracking problems, a combined controller is proposed to ensure constraint satisfaction during adaptation. Simulation and experimental results demonstrate that the proposed controller achieves higher path tracking performance than either approach alone, while maintaining performance comparable to unconstrained adaptive control. The results further indicate that the proposed method also improves the trajectory tracking capability of path-velocity control.

Keywords: Nonlinear system theory, Nonlinear system identification, Complex systems
Abstract: This paper proposes a novel formulation of frequency response for nonlinear systems in the Koopman operator framework. This framework is a promising direction for the analysis and synthesis of systems with nonlinear dynamics based on (linear) Koopman operators. We show that the frequency response of a nonlinear plant is derived through the Laplace transform of the output of the plant, which is a generalization of the classical approach to LTI plants and is guided by the resolvent theory of Koopman operators. The response is a complex-valued function of the driving angular frequency, allowing one to draw the so-called Bode plots, which display the gain and phase characteristics. Sufficient conditions for the existence of the frequency response are presented for three classes of dynamics.

Keywords: Emerging control theory, Nonlinear system theory, Robust control
Abstract: The purpose of this paper is to study phases of tensors under the Einstein product. Firstly, by defining the numerical range and sectorial tensors, the sectorial tensor decomposition is derived, which leads to the definition of phases for square tensors. Secondly, the maximin and minimax expressions of tensor phases are given and inequalities involving the phases of a sectorial tensor and those of its compressions are established. Finally, a tensor version of the small phase theorem is presented, which can be regarded as a natural generalization of the matrix case, recently proposed in Ref. [1].

Keywords: Nonlinear system theory, Optimization
Abstract: We introduce a generalization of the scaled relative graph (SRG) to pairs of operators, enabling the visualization of their relative incremental properties. This novel SRG framework provides the geometric counterpart for the study of nonlinear resolvents based on paired monotonicity conditions. We demonstrate that these conditions apply to linear operators composed with monotone mappings, a class that notably includes NPN transistors, allowing us to compute the response of multivalued, nonsmooth and highly nonmonotone electrical circuits.

Keywords: Nonlinear system theory, Feedback linearization, Algebraic/geometric methods
Abstract: Following the recent approach of the authors [Nicolau et. al 2024], this paper further analyzes relationships between mechanical flatness, the underlying mechanical structure of the system, and different variants of differential flatness, in particular, configurational flatness. For the class of mechanical control systems with n degrees of freedom and n-1 inputs, we establish a general classification result in terms of normal forms, emphasizing the connections between mechanical flatness, configuration flatness, and static mechanical feedback linearization. Moreover, in the case where the distribution, spanned on the configuration manifold by control vector fields, is involutive, we provide geometric verifiable characterizations of both configuration flatness and mechanical flatness.

Keywords: Constrained control, Nonlinear system theory, Algebraic/geometric methods
Abstract: This paper investigates small-time local controllability (STLC) in nonlinear control-affine systems under zero-excluding asymmetric input constraints, such as strictly positive or strictly negative bounded intervals. Such systems arise naturally in applications including spacecraft with unidirectional thrusters, cable-driven robots, and biomechanical actuators. Classical STLC criteria rely on symmetric control domains to recover local reversibility through Lie bracket approximations, and therefore fail in these irreversible settings. Focusing on a controlled equilibrium, we examine the feasible velocity set and show that when the origin lies outside this set, a supporting hyperplane defines a covector that induces a strictly monotonic scalar functional along admissible trajectories. This yields a scalar quantity that is strictly increasing along every admissible trajectory in a neighborhood of the equilibrium, thereby ruling out STLC while remaining compatible with finite-time local controllability (FTLC). By contrast, when the origin belongs to the feasible velocity set, no such uniform monotonicity can be inferred from first-order bounds alone, so STLC remains undecidable without higher-order analysis. The proposed framework complements existing results for symmetric and zero-including unilateral input domains by providing a geometric criterion for irreversibility. An application to satellite attitude control with unidirectional thrusters illustrates the approach and provides explicit time bounds.

Keywords: Nonlinear system theory, Stability of nonlinear systems
Abstract: In this short note, we establish the Lipschitz continuity of solutions to parameterized nonlinear Roesser models with respect to both the parameters and the boundary conditions.

Keywords: Differential algebraic systems, Distributed parameter systems
Abstract: This paper investigates the solvability of infinite-dimensional differential--algebraic equations (DAEs) on Hilbert spaces, focusing on systems with irregular structure where sE - A is not surjective onto the full state space. Instead, surjectivity holds only on a smaller subspace determined by the algebraic constraints. Referring to Showalter’s framework for implicit evolution equations with dissipative operators, we first show that the conditions therein yield contraction semigroup generation, thereby ensuring existence and uniqueness of solutions under this partial surjectivity. These conditions in fact serve as an extension of the Lumer--Phillips theorem to dissipative infinite-dimensional DAEs. Moreover, focusing on a special class of infinite-dimensional DAEs, we illustrate the additional conditions needed for Showalter’s framework to yield E-radiality of the system with a certain choice of the spaces, resulting in a decomposition of the state space parallel to the Weierstrass form for finite-dimensional DAEs. Such decomposition offers a foundation for applications in control and observer design. The theoretical framework is illustrated by a class of coupled equations.

Keywords: Distributed parameter systems, Delay systems, Linear systems
Abstract: This contribution develops an algebraic approach to obtain a controller form for a class of linear hyperbolic MIMO systems, bidirectionally coupled with a linear ODE system at the unactuated boundary. After a short summary of established controller forms for SISO and MIMO ODE as well as SISO hyperbolic PDE systems, it is shown that the approach to state a controller form for SISO systems cannot easily be transferred to the MIMO case as it already fails for a very simple example. Next, a generalised hyperbolic controller form with different variants is proposed and a new flatness-based scheme to compute said form is presented. Therein, the system is treated in an algebraic setting where quasipolynomials are used to express the predictions and delays in the system. The proposed algorithm is then applied to the motivating example.

Keywords: Distributed parameter systems, Robust control, Observers for linear systems
Abstract: We develop a mu-analysis and synthesis framework for infinite-dimensional systems that leverages the Integral Quadratic Constraints (IQCs) to compute the structured singular value's upper bound. The methodology formulates robust stability and performance conditions jointly as Linear Partial Integral Inequalities within the Partial Integral Equation framework, establishing connections between IQC multipliers and mu-theory. Computational implementation via PIETOOLS enables computational tools that practically applicable to spatially distributed infinite dimensional systems. Illustrations with the help of Partial and Delay Differential Equations validate the effectiveness of the framework, showing a significant reduction in conservatism compared to unstructured methods and providing systematic tools for stability-performance trade-off analysis.

Keywords: Delay systems, Switched systems
Abstract: We develop a predictor-feedback control design for a class of linear systems with state-dependent switching. The main ingredient of our design is a novel construction of an exact predictor state. Such a construction is possible as for a given, state-dependent switching rule, an implementable formula for the predictor state can be derived in a way analogous to the case of nonlinear systems with input delay. We establish uniform exponential stability of the corresponding closed-loop system via a novel construction of multiple Lyapunov functionals, relying on a backstepping transformation that we introduce. We validate our design in simulation considering a switching rule motivated by communication networks.

Keywords: Observers for nonlinear systems, Distributed parameter systems
Abstract: The dual Koopman observer is extended in order to accommodate initial conditions associated with bounded density functions that represent a probability distribution of the true initial condition. The initial condition of the observer is then defined as the image of this density function under a specific kernel-integral operator. It is proved that the resulting problem is well-posed and that the dual Koopman observer is exponentially convergent. The fact that this approach offers greater flexibility in selecting the observer initial condition is highlighted. The reported results are illustrated with numerical simulations.

Keywords: Machine learning, Game theoretical methods, Optimization algorithms
Abstract: Federated learning is an emerging distributed paradigm that addresses challenges posed by heterogeneous, privacy-sensitive data. It enables multiple clients to collaboratively train a model by aggregating their local updates on the server. However, conventional aggregation schemes typically use fixed weights that fail to reflect unequal and time-varying client contributions, leading to biased and unstable learning. To improve fairness and stability, we propose the trajectory Shapley value (TSV), a contribution metric that evaluates how each client influences the optimization trajectory of the global model using a validation-based, temporally consistent utility. Building on TSV, we design FedTSV, an adaptive aggregation method that converts per-round evaluations into dynamic client weights, allowing the server to respond to heterogeneous and adversarial participation in real time. Experiments on benchmark datasets show that FedTSV accelerates convergence, improves robustness, and yields more equitable contribution assessments, thereby providing a principled foundation for fairness-aware federated optimization.

Keywords: Energy systems, Electrical power systems, Optimization
Abstract: This paper extends the recently proposed dynamic model for local energy markets (LEM) by introducing a control-based formulation that ensures faster convergence and improved transient performance. The original model captures the interactions between producers and consumers in LEMs through monotonic and bijective supply and demand functions, leading to a unique market-clearing price (MCP) as a global attractor. However, asymmetric behavior of price and quantity dynamics can result in oscillations and slow convergence. To overcome these limitations, we propose a control law that preserves the equilibrium while reshaping the transient response. Analytical conditions are presented, providing sufficient guarantees of global convergence. Numerical results demonstrate that the proposed controller effectively damps oscillations, accelerates convergence, and maintains the autonomy properties of the original model, making it a promising approach for the dynamic coordination of LEMs.

Keywords: Energy systems, Predictive control for linear systems, Optimization
Abstract: This paper presents a comparative analysis of model predictive control (MPC) strategies for managing an experimental photovoltaic microgrid with battery energy storage and hybrid inverters. The hybrid inverters enable battery charging from both solar generation and grid imports, providing enhanced operational flexibility. Two optimization objectives are evaluated: minimizing the economic cost of imported energy and minimizing the total imported energy. The economic optimization employs a variable prediction horizon aligned with electricity tariff periods, while the energy minimization uses a fixed 24-hour horizon. The primary contribution of this work lies in the use of a detailed battery model that captures SOC- and power-dependent charge/discharge efficiencies and the integration of hybrid inverter dynamics within the MPC optimization framework, along with an efficient formulation method to incorporate these variable efficiencies into the optimization problem. This approach enables a more accurate representation of real battery behavior compared to conventional constant-efficiency models. Simulation results across multiple operating scenarios demonstrate the impact of detailed efficiency modeling on control performance and provide insights for selecting appropriate strategies based on system objectives and operational priorities.

Keywords: Energy systems, Electrical power systems, Optimal control
Abstract: Service stations for fast charging of plug-in electric vehicles (PEVs) will need to operate energy storage systems (ESS) in order to balance in real time the aggregated PEV load, which can easily reach MW levels. In this paper, a stochastic optimal ESS control strategy is presented, to flatten the power flow at the point of connection of the service station with the grid. Chance constraints on the violation of area power limits are included and a deterministic counterpart of the original problem formulation is derived leveraging Cantelli's inequality. The resulting ESS control is based on the expected value and the variance of the aggregate PEV power demand over the time, which in turn are derived from analysing the service station modelled as a queuing system, with Poissonian arrivals and exponentially distributed dwelling times. Numeric simulations show that the proposed method is effective in properly balancing the PEV load, leading to a lower power profile at the point of connection with the grid, which brings economic benefits for both the service station operator and the PEV drivers, due to lower connection fees and investment costs.

Keywords: Energy systems, Optimization, Large-scale systems
Abstract: The Unit Commitment (UC) problem consists in controlling a large fleet of heterogeneous electricity production units in order to minimize the total production cost while satisfying consumer demand. Electric Vehicles (EVs) are used as a source of flexibility and are often aggregated for problem tractability. We develop a new approach to integrate EV flexibilities in the UC problem and exploit the generalized polymatroid structure of aggregated flexibilities of a large population of users to develop an exact optimization algorithm, combining a cutting-plane approach and submodular optimization. We show in particular that the UC can be solved exactly in a time which scales linearly, up to a logarithmic factor, in the number of EV users when each production unit is subject to convex constraints. We illustrate our approach by solving a real instance of a long-term UC problem, combining open-source data of the European grid (European Resource Adequacy Assessment project) and data originating from a survey of user behavior of the French EV fleet.

Keywords: Concensus control and estimation, Electrical power systems, Energy systems
Abstract: Islanded DC microgrids relying on Renewable Energy Sources (RESs), such as photovoltaic and wind systems, face significant challenges in maintaining stable operation due to the intermittent nature of renewable generation and the uncertainty of demand. To address this issue, energy storage systems (battery stacks) are typically integrated to balance supply and demand over different time scales. Effective coordination of these components is therefore essential for reliable long-term operation. This paper proposes a consensus-based control approach that coordinates in-service batteries to i) compensate the power imbalances when RES units do not match the demanded load, and ii) achieve consensus among the batteries’ State of Charge (SoC). By driving the SoCs to consensus, the proposed method helps improving their performance and lifespan, while ensuring that the load demand is satisfied. In scenarios where all the batteries are fully charged, power redistribution among RES units is employed to preserve the supply–demand balance. Simulation results confirm the effectiveness of the presented method in achieving SoCs consensus under different operating conditions. In addition, a realistic converter transfer function model is incorporated to validate the proposed control strategy.

Keywords: Game theoretical methods, Optimization algorithms, Distributed control
Abstract: This work proposes a distributed peer-to-peer (P2P) energy market scheme to optimize energy transactions among agents in an energy community (EC). The proposed framework integrates a welfare-based optimization approach that maximizes overall community benefit while penalizing excessive generation costs, and long distribution distances to account for electric grid losses and congestion. The model relies on an underlying communication graph, where trading dynamics are induced as agents change their roles between sellers or buyers according to their generation–demand profiles. A replicator dynamics (RD) mechanism is employed to evolve the sellers’ strategies over time toward a welfare-optimal equilibrium. The methodology incorporates Lagrangian relaxation (LR) and a consensus protocol to manage coupling constraints related to generation limits, demand satisfaction, local information exchange, and the pursuit of community self-sufficiency. Simulation results validate the proposed distributed approach, showing convergence to optimal welfare with low relative errors compared to a conventional centralized sequential least squares programming (SLSQP) solution.

Keywords: UAV's, Fault diagnosis, Aerospace
Abstract: This paper presents a GNSS spoofing detection method based on the attitude estimation anomaly of an IMU-GNSS fusion algorithm in case of GNSS spoofing pointed out in a previous work of the author. To support detection, a GNSS independent attitude estimator is also constructed which gives proper attitude information for acceptably long time. Two error measures are defined, one for the difference of estimated attitudes of the two algorithms and one for the difference of IMU-GNSS fusion magnetic vector estimate and the measured magnetic vector. Both estimators are tuned on real flight data. Then GNSS trajectories with sophisticated spoofing are considered, the two estimators run and the error measures evaluated again with real flight data. Thresholding and updown counters are applied for GNSS spoofing detection giving acceptable results with the attitude error-based detection and unacceptable ones with the magnetic estimate-based.

Keywords: UAV's, Fault tolerant systems, Switched systems
Abstract: This paper presents a novel dual Markov jump linear system (DMJLS)-based fault-tolerant control (FTC) framework for a quadrotor unmanned aerial vehicle (QUAV) under single rotor faults. The method employs a dual switching control strategy: stochastic jumps modeled by a Markov chain capture unpredictable rotor faults, while deterministic switches governed by a scheduling signal enable fault compensation. A cascaded structure is designed, where the outer loop includes a PID controller for position tracking, and the inner loop uses the DMJLS for attitude stabilization. A co-design approach is further introduced to jointly optimize the feedback gains and switching strategy, ensuring the system mean-square (MS) stability and satisfying H_∞ performance under rotor faults and disturbances. Simulation results show that the proposed method achieves accurate trajectory tracking and strong robustness against both faults and external disturbances.

Keywords: Electrical power systems, Fault detection and identification, Observers for nonlinear systems
Abstract: This paper highlights a low frequency model for a synchronous generator, as typically used in a hydroelectric power plant. This model is chosen so that it can be used to retrieve the internal variables of interest, like magnetic fluxes, voltages, currents, on the direct and quadrature axes, in addition to the excitation gain. After analyzing observability of the system, the well-known Extended Kalman Filter is performed on a real industrial case. Finally, taking advantage of the modeling, an observer-based sensor fault detection application is presented.

Keywords: Fault detection and identification, Identification, Adaptive systems
Abstract: Modern control systems require real-time monitoring to ensure safety and reliability, yet detecting changes in system behavior remains challenging in nonlinear, high-dimensional, and time-varying environments. We introduce an interpretable change-point detection framework based on truncated online Dynamic Mode Decomposition with control (toDMDc). The method combines optimal rank truncation with online system identification, enabling real-time adaptation to evolving dynamics while maintaining numerical stability. By comparing reconstruction errors between reference and test windows, the framework detects changes in system behavior. We demonstrate convergence and validate the approach on three case studies: synthetic step changes, nonlinear two-tank system with input delays, and industrial battery energy storage system. Results show that toDMDc-based detection achieves accurate change-point identification with interpretable statistics, bounded detection delays, and computational efficiency suitable for real-time deployment in safety-critical control applications.

Keywords: Fault diagnosis, Chemical process control, Signal processing
Abstract: This work examines how different approaches to probability density estimation affect causality analysis based on transfer entropy (TE). TE quantitatively describes cause-effect relationships in time series. It captures not only statistical correlations, but also the direction of information flow. The basis of TE calculations is the estimation of probability density functions (PDF). In practical applications, true properties are unknown and must be approximated. One of the most common techniques for this purpose is kernel density estimation (KDE). Results demonstrate that there is no single, universally optimal estimation method, as performance depends on statistical properties of data. Cross-validation methods may lead to numerically unstable results for time series dominated by control error dynamics. The research is conducted in three simulation environments: a synthetic model with known structure, a complex multi-loop dynamic model, and the Tennessee Eastman Process benchmark. The effectiveness is compared for four kernel functions (Gaussian, Laplace, Student's t, and alpha-stable) and three bandwidth selection strategies (Silverman, likelihood cross-validation (LCV), and an original hybrid method).

Keywords: Fault tolerant systems, Fault estimation, System reconfiguration
Abstract: Diverse control techniques have been developed to maximize the energy captured from the sea by wave energy converters (WECs). As an alternative to optimization-based control strategies for WECs, which rely on numerical optimization of an energy-based cost function, non-optimization-based controllers (NOBCs) are characterized by lower computational complexity. Implementation of NOBCs for WECs, however, generally requires additional mechanisms to satisfy motion constraints. Since position and velocity constraint handling are essential to guarantee safe operation of the WEC, this paper proposes a fault-tolerant control (FTC) strategy to tackle the presence of faults in position and velocity sensors. Specifically, the considered control strategy combines a "simple and effective" NOBC with a multiplicative setpoint conditioning method, based on a Gaussian modulating envelope, to handle position and velocity constraints. The performance of the proposed FTC is evaluated through simulation with a realistic wave profile in the presence of diverse fault scenarios.

Keywords: Transportation systems, Aerospace, Predictive control for linear systems
Abstract: This paper presents a Model Predictive Control (MPC) scheme for flight scheduling and energy management of electric aviation networks, where electric aircraft transport passengers between electrified airports equipped with sustainable energy sources and battery storage, with the goal of minimizing grid dependency. Specifically, we first model the aircraft flight and charge scheduling problem jointly with the airport energy management problem, explicitly accounting for local weather forecasts. Second, we frame the minimum-grid-energy operational problem as a mixed-integer linear program and solve it in a receding horizon fashion, where the route assignment and charging decisions of each aircraft can be dynamically reassigned to mitigate disruptions. We showcase the proposed MPC scheme on real-world data taken from a conventional flight network and weather conditions in the US American North East. The proposed framework saves between 10 and 37% of grid energy requirements when compared to a baseline without re-routing. Hence, results show that MPC can effectively guarantee operation of the network by efficiently re-assigning flights and rescheduling aircraft charging, while maximizing the efficiency of the on-site energy systems.

Keywords: Sliding mode control, Stability of nonlinear systems, Model validation
Abstract: In this paper, an assistance system with pneumatic artificial muscles for a manually guided payload handling is presented. One operation mode requires the stabilization of the end effector position during payload changes, where two alternative position control approaches are designed and investigated: flatness-based and integral sliding mode control. The proposed control structure is a cascaded one, where in the inner loop the internal muscle pressures are controlled. Two polynomial functions are identified experimentally to approximate the characteristics for the nonlinear muscle forces and volumes, whereas the equations of motion for the mechanical structure are derived using Lagrange’s equations. Both control designs are validated and compared to each other with varying payloads on a prototype of the assistance system. Experimental results indicate a good control performance.

Keywords: Traffic control, Machine learning, Decentralized control
Abstract: Effective traffic control is essential for mitigating congestion in transportation networks. Conventional traffic management strategies, including route guidance and ramp metering, often rely on state feedback controllers, which are used for their simplicity and reactivity; however, they lack the adaptability required to cope with complex and time-varying traffic dynamics. This paper proposes a multi-agent reinforcement learning (RL) framework in which each agent adaptively tunes the parameters of a state feedback traffic controller, combining the reactivity of state feedback controllers with the adaptability of RL. By tuning parameters at a lower frequency rather than directly determining control inputs at a high frequency, the RL agents achieve improved training efficiency while maintaining adaptability to varying traffic conditions. The multi-agent structure further enhances system robustness, as local controllers can operate independently in the event of partial failures. The proposed framework is evaluated on a simulated multi-class transportation network under varying traffic conditions. Results show that the proposed multi-agent framework outperforms the no-control and fixed-parameter state feedback control cases, while performing on par with the single-agent RL-based adaptive state feedback control, but with much greater resilience to disturbances.

Keywords: Emerging control applications, Identification
Abstract: We study whether politicians adapt the sentiment of their messages in response to public engagement, a question relevant to understanding online political communication and to informing interventions against polarization and negativity amplification on social platforms. Using over 1.5 million tweets from Members of Parliament in the United Kingdom, Spain, and Greece during 2021, we identify sentiment dynamics through a simple yet interpretable linear model. The results are consistent with a closed-loop mechanism: engagement with positive and negative messages predicts the sentiment of subsequent posts. The learned coefficients also reveal systematic differences across political roles: opposition members are more reactive to negative engagement, whereas government officials respond more to positive signals. These results provide a quantitative, control-oriented view of behavioral adaptation in online politics, showing how feedback principles help interpret self-reinforcing dynamics in social media discourse.

Keywords: Electrical machine control, Computer networks, Linear systems
Abstract: The focus of this paper is control of vibrations in a mechanical structure using wireless sensors and contactless actuators.

The mechanical system consists of three coupled, under-damped spring-mass systems and two electromagnetic actuators. A complete mathematical model of the mechanical system is developed and validated using data from the real system.

Three wireless sensors are constructed to measure the acceleration of the three masses in the mechanical structure. The wireless sensors are built around a EPS32 micro-controller together with an accelerometer module. The EPS32 micro-controller is selected because it offers a wireless connection with a fast dual-core CPU. Together with the three sensors, digital IIR filters are designed and applied for filtering out measurement noise and unwanted frequencies outside the control frequency range.

The system is controlled by using a cascade controller architecture. A controller is designed for each of the three measured accelerations using standard frequency design methods. The designed controller is validated on both the simulation model as well as on the real system.

Keywords: Manufacturing processes, Process control, Optimal control
Abstract: The phenomenon of sloshing, i.e., the oscillation of liquids inside partially filled containers during motion, represents a significant challenge in transportation and process engineering. This paper addresses the design of an optimized motion law to reduce unwanted liquid oscillations in automated liquid handling processes. A constrained optimization problem based on a linearized sloshing model is designed with the aim of minimizing liquid oscillation both during movement and in the subsequent resting phase. To experimentally validate the proposed method, a dedicated experimental setup is designed, including a controlled motion system and a visual tracking system. Experimental tests confirm the reliability of the optimized motion law.

Keywords: Differential algebraic systems, Observers for nonlinear systems, LMI's/BMI's/SOS's
Abstract: This paper presents a novel observer design for a class of singular nonlinear systems. Unlike other methods that require a set of nonlinear transformations, the proposed approach avoids them by introducing an auxiliary dynamic based on a chain of integrators. Furthermore, algebraic constraints are incorporated as auxiliary outputs within the observer dynamics to improve estimation accuracy. The observer synthesis is formulated as an optimization problem subject to linear matrix inequality (LMI) constraints. Numerical simulations are provided to demonstrate the effectiveness and robustness of the proposed methodology.

Keywords: Observers for nonlinear systems, Neural networks, Uncertain systems
Abstract: This paper presents a Physics-Informed Neural Network (PINN) framework for jointly estimating state, parameters, and disturbance for a class of nonlinear dynamical autonoumous systems in a lower triangular form, subject to unknown disturbances. First, we introduce a bank-of-observers approach for recursive state, parameter, and disturbance estimation, where the objective (loss) function incorporates previous estimations, allowing the bank of estimators to recursively learn from the last estimations, and to retro-propagate the estimation error through the previous neural networks. We later address the issue of error propagation through the serial approximators where an upper bound on the estimation error is derived. The effectiveness of the proposed method is demonstrated through numerical simulations.

Keywords: Autonomous systems, Observers for nonlinear systems, Neural networks
Abstract: Accurate full-state estimation of a rigid body is essential in many navigation and control tasks, but may be challenging with low-cost sensors. This work addresses the problem using only Inertial Measurement Unit (IMU) and Global Positioning System (GPS) data. We propose a hybrid architecture based on the Extended Kalman Filter (EKF), en- hanced by a neural network that adaptively estimates process and measurement noise covariances. This Neural Enhanced Extended Kalman Filter (NEEKF) improves robustness under time-varying and nonlinear noise conditions. Standard EKFs assume fixed covariances (Q, R), limiting adaptability in real- world scenarios—e.g., autonomous tractors on irregular terrain. To overcome this limitation, we train neural models to estimate Qk and Rk from internal filter signals such as innovations and uncertainty metrics. The approach is validated in a high- fidelityUnity simulation replicating realistic disturbances includ- ing rain, dropouts, and soil-induced vibrations. Results show that adaptive noise modeling significantly enhances estimation accuracy. Among the tested architectures, feedforward networks showed the best trade-off between performance and simplicity, while Gated Recurrent Units (GRUs) and Transformers offered improvements in dynamic and temporally correlated conditions.

Keywords: Observers for nonlinear systems, Observers for linear systems
Abstract: In this paper, we propose a sample-based moving horizon estimation (MHE) scheme for general nonlinear systems to estimate the current system state using irregularly and/or infrequently available measurements. The cost function of the MHE optimization problem is suitably designed to accommodate these irregular output sequences. We also establish that, under a suitable sample-based detectability condition known as sample-based incremental input/output-to-state stability (i-IOSS), the proposed sample-based MHE achieves robust global exponential stability (RGES). Additionally, for the case of linear systems, we draw connections between sample-based observability and sample-based i-IOSS. This demonstrates that previously established conditions for linear systems to be sample-based observable can be utilized to verify or design sampling strategies that satisfy the conditions to guarantee RGES of the sample-based MHE. Finally, the effectiveness of the proposed sample-based MHE is illustrated through a simulation example.

Keywords: Observers for nonlinear systems, Algebraic/geometric methods, Nonlinear system theory
Abstract: In this paper the problem of transforming the state equations of a single-input single-output nonlinear discrete-time control system into the generalized observer form is studied. The latter form is linear up to some injection terms depending on system input, output and a finite number of their past values. This allows to construct easily the state observers with linear error dynamics for the equations in the generalized observer form. Previous studies have assumed system reversibility and/or observability. In this work less restrictive assumptions of submersivity and backward observability (generalization of constructibility) are made allowing to expand the class of equations which can be transformed into the generalized observer form. The main result gives necessary and sufficient conditions for finding the minimal number of past values of the input and the output necessary to transform the state equations into the generalized observer form.

Keywords: Observers for nonlinear systems, Nonlinear system theory, LMI's/BMI's/SOS's
Abstract: This paper addresses the design of observers for nonlinear Lipschitz systems using Linear Matrix Inequalities (LMIs). The primary contribution lies in proposing a novel technique for decomposing nonlinear functions. We demonstrate that the choice of decomposition significantly impacts the feasibility of the resulting LMI conditions. Specifically, the LMIs are affected by how the nonlinearity is decomposed, enabling the incorporation of certain nonlinear terms into the observer structure. To validate the proposed method, illustrative examples is provided, highlighting the critical role of accurate nonlinearity decomposition in observer design.

Keywords: Aerospace, Constrained control, Optimization
Abstract: Free-space optical communications on CubeSats demands compact and precise pointing systems that minimize dependence on satellite body attitude. This paper presents the design and validation of a gimbaled-prism coarse pointing assembly for the Cube1G mission, enabling hemispherical pointing on CubeSat scale. The mechanism uses direct-drive brushless DC motors with absolute encoders for azimuth (356 deg) and elevation (80 deg) control, integrated with a compact launch-lock system. A hierarchical control strategy combines a cascaded servo loop for current, velocity, and position control with an s-curve profiler for smooth ephemeris-based tracking. During acquisition, a pre-deformed spiral scan compensates for the gimbals spherical coordinates, and once the beacon is detected, a quadrant photodiode-based PI loop closes the optical feedback. Experimental validation shows the emission of low mechanical vibration (dominant at 100 Hz) during a full rotation and stable closed-loop performance. The coarse pointing assembly maintains open loop pointing within the field of view for accelerations up to 1 rev/s² and velocities up to 0.1 rev/s. For the closed-loop tracking phase, a total optical mean pointing error of 11.1 µrad is maintained.

Keywords: Aerospace, UAV's, Nonlinear system theory
Abstract: We present a new stability proof for cascaded geometric control used by aerial vehicles tracking time-varying position trajectories. Our approach uses sliding variables and a recently proposed quaternion-based sliding controller to demonstrate that exponentially convergent position trajectory tracking is theoretically possible. Notably, our analysis reveals new aspects of the control strategy, including how tracking error in the attitude loop influences the position loop, how model uncertainties affect the closed-loop system, and the practical pitfalls of the control architecture.

Keywords: Aerospace, Sliding mode control, Autonomous systems
Abstract: Autonomous ground operations for aircraft demand precise lateral control to ensure safe navigation in constrained environments. This paper focuses on the control aspect of obstacle avoidance maneuvers, addressing the design of a robust lateral guidance strategy based on Integral Sliding Mode control (ISM). The controller is developed on a reduced-order dynamic model with virtual control inputs and integrated within a realistic control-allocation framework that distributes effort among rudder, nose wheel steering, and braking systems, in connection with a high-fidelity multibody simulation environment. Comparative tests against a parallel controller and an H-infinity benchmark demonstrate that the ISM achieves the smallest lateral tracking errors and remains nearly unaffected by model uncertainties. Despite a more aggressive control action, the method ensures superior robustness and accuracy, making it particularly suitable for safety-critical scenarios such as obstacle avoidance and precision hangar parking.

Keywords: Aerospace, Sliding mode control, Robust control
Abstract: This paper presents a coupled and decoupled sliding mode based controller for the line-of-sight (LOS) stabilization of a two degree-of-freedom gimbal system. The governing nonlinear dynamical model incorporates static and dynamic mass imbalance, external disturbance, motor friction, and torque ripple effect, along with the mechanical coupling between inner gimbal and outer gimbal axes. The coupled controller only requires limited information about the bounds on the coupled dynamic unbalance torques to tackle the coupling of the gimbal systems. Furthermore, we also propose a decoupled controller using a decoupled dynamics obtained by eliminating the cross inertia and mass terms from both the inner and outer gimbal dynamics. The controller's robustness and tracking performance are evaluated under complete coupled dynamics using a simulation study. The simulation study compares the performance of the proposed coupled an decoupled controller using two reference test signals designed to test LOS stabilization against high or smooth maneuvering targets. The results validate the controller's effectiveness in maintaining precise LOS pointing toward the target under external disturbance.

Keywords: Machine learning, Aerospace, Iterative learning control
Abstract: This paper presents a data-driven control framework for active Gust Load Alleviation (GLA), on a flexible wing demonstrator, using the Reinforcement Learning based Soft Actor-Critic (SAC) algorithm. This approach extends RL methods to the domain of aeroelastic control, where traditional model based control methods are limited by the difficulty of obtaining accurate linear Aeroservoelastic (ASE) representations for highly flexible structures. The control law is defined by a neural network architecture that maps sensor inputs to actuator commands. This architecture is trained offline to achieve an optimal control logic using the SAC RL algorithm. The data required for this training process is generated from a highfidelity linear ASE model of the flexible wing demonstrator. Rather than employing this model explicitly within the control design, it serves as a digital twin to provide the necessary training data. The study details the actor critic neural network structures, the reward model formulation, and the training procedure for controller optimisation. The resulting control law is evaluated in simulation and subsequently validated through wind tunnel experiments. Results demonstrate that the SAC based controller effectively mitigates gust induced loads, highlighting the potential of RL based methods for future active aeroelastic control applications.

Keywords: UAV's, Aerospace, Modeling

Keywords: Nonlinear system theory, Lyapunov methods, Emerging control theory
Abstract: Over recent years, Koopman operator theory has been the focus of increasing attention in the context of nonlinear systems analysis and control. In this talk, we will cover the fundamentals of Koopman operator theory for autonomous systems, highlighting both advantages and limitations of the approach as well as common pitfalls. We will first review the definitions and main properties of the Koopman operator, with a specific focus on its spectral properties. Next, we will provide a broad overview of the finite-dimensional approximation techniques that are leveraged in numerical implementations. In the last part of the talk, concepts and methods from Koopman operator theory will be illustrated with recent results in stability analysis and state estimation.

Keywords: Nonlinear system theory, Predictive control for nonlinear systems, Lyapunov methods
Abstract: Koopman operator theory considers nonlinear dynamical systems and constructs an infinite-dimensional linear system representation whose solutions are consistent with the solutions of the underlying nonlinear system. The original Koopman composition operator was defined as a measure-preserving, linear bounded operator acting on Hilbert spaces. Generalizing the Koopman operator to nonlinear systems with control input has gain much interest in recent years, due to relevant implications for nonlinear control design. In this tutorial talk we will provide a compact overview of possible approaches to the Koopman operator for systems with inputs, ranging from linear and bilinear in control Koopman models to the more recent, generalized bilinear Koopman models defined on tensor products of Hilbert spaces. We will discuss theoretical assumptions that guarantee exact Koopman models and data-driven methods for computing finite-dimensional approximate Koopman models. Furthermore, we will also show a comparison of various Koopman models in terms of closed-loop performance when utilized as prediction models in MPC algorithms. The developed software implementations will be shared, and participants will be invited to apply the tested methods to applications of interest.

Keywords: Nonlinear system theory, Predictive control for nonlinear systems, Lyapunov methods
Abstract: Extended dynamic mode decomposition, embedded in the Koopman framework, is a widely-applied data-driven technique to predict the evolution of an observable along the flow of a nonlinear dynamical system. In the first part of the talk, we consider the approximation error, which results from using only finitely many observable functions (projection) and finite data (estimation). Then, we indicate extensions towards reproducible kernel Hilbert spaces to establish pointwise error bounds using kernel EDMD. In the second part of the talk, we extend our findings to systems with inputs and demonstrate the applicability of Koopman data-driven surrogate models for (predictive) control with stability guarantees.

Keywords: Iterative learning control, Output feedback
Abstract: In this paper, we present an observer-based output feedback Iterative Learning Control (ILC) scheme for single input, single output discrete-time Lur'e-type systems with sector-bounded nonlinearities in the state equation. The design is achieved by concatenating trials and casting the trial-by-trial memory of ILC as a delay system. A pair of synthesis Linear Matrix Inequalities is formulated to solve the observer and controller design problems, respectively. The proposed design is tested on an unstable nonlinear system.

Keywords: Automotive, Optimization, Mechatronics
Abstract: Controller parameter tuning in automotive development typically requires extensive real-world testing by application engineers, which is time-consuming, costly, and risks evaluating parameter configurations that lead to poor or unsafe closed-loop behavior. In industrial practice, controller design follows a two-stage workflow – initial design and tuning in simulation (low-fidelity), followed by application and refinement on the real vehicle (high-fidelity) – placing additional requirements on tuning methods. Two key challenges arise in this setting: reducing the number of costly real-world experiments, and avoiding unsafe parameter configurations. In this work, we propose a framework for cautious learning of controller parameters based on constrained multi-fidelity Bayesian optimization, explicitly combining multi-fidelity modeling with safety-related constraints. The proposed framework guides the optimization toward promising regions while avoiding parameter choices that may degrade performance or violate operational limits. This enables efficient and risk-averse exploration even when high-fidelity evaluations are scarce and expensive. Simulation results show that the framework achieves an effective trade-off between learning efficiency and avoidance of undesirable configurations, offering a practical pathway toward automated, data-efficient controller tuning in industrial applications.

Keywords: Distributed cooperative control over networks, Machine learning, Neural networks
Abstract: Multi-agent reinforcement learning (MARL) has shown promise for large-scale network control, yet existing methods face two major limitations. First, they are typically designed for systems in which local interactions decay exponentially with network distance, which limits their ability to capture long-range dependencies such as cascading power failures or epidemic outbreaks. Second, most approaches lack generalizability across network topologies, requiring retraining when applied to new graphs. We introduce STACCA (Shared Transformer Actor-Critic with Counterfactual Advantage), a transformer-based MARL framework that addresses both challenges. STACCA employs a centralized Graph Transformer Critic to model long-range dependencies and provide system-level feedback, while its shared Graph Transformer Actor learns a generalizable policy capable of adapting across diverse network structures. To improve credit assignment during training, STACCA integrates a novel counterfactual advantage mechanism that isolates individual agent contributions and is compatible with state-value critic estimates. We evaluate STACCA on epidemic containment and rumor-spreading network control tasks, demonstrating improved performance, zero-shot generalization across diverse network topologies, and scalability to networks 20 times the size of the training network. These results highlight the potential of transformer-based MARL architectures to achieve scalable and generalizable control in large-scale networked systems.

Keywords: Statistical learning, Optimization, Automotive
Abstract: Tuning control policies manually to meet high-level objectives is often time-consuming. Bayesian optimization provides a data-efficient framework for automating this process using numerical evaluations of an objective function. However, many systems — particularly those involving humans — require optimization based on subjective criteria. Preferential Bayesian optimization addresses this by learning from pairwise comparisons instead of quantitative measurements, but relying solely on preference data can be inefficient. We propose a multi-fidelity, multi-modal Bayesian optimization framework that integrates low-fidelity numerical data with high-fidelity human preferences. Our approach employs Gaussian process surrogate models with both hierarchical, autoregressive and non-hierarchical, coregionalization-based structures, enabling efficient learning from mixed-modality data. We illustrate the framework by tuning an autonomous vehicle's trajectory planner, showing that combining numerical and preference data significantly reduces the need for experiments involving the human decision maker while effectively adapting driving style to individual preferences.

Keywords: Machine learning, Markov processes, Stochastic control
Abstract: Restless Multi-Armed Bandits (RMABs) are powerful models for decision-making under uncertainty, yet classical formulations typically assume fixed dynamics, an assumption often violated in nonstationary environments. We introduce MARBLE (Multi-Armed Restless Bandits in a Latent Markovian Environment), which augments RMABs with a latent Markov state that induces nonstationary behavior. In MARBLE, each arm evolves according to a latent environment state that switches over time, making policy learning substantially more challenging. We further introduce the Markov-Averaged Indexability (MAI) criterion as a relaxed indexability assumption and prove that, despite unobserved regime switches, under the MAI criterion, synchronous Q-learning with Whittle Indices (QWI) converges almost surely to the optimal Q-function and the corresponding Whittle indices. We validate MARBLE on a calibrated simulator-embedded (digital twin) recommender system, where QWI consistently adapts to a shifting latent state and converges to an optimal policy, empirically corroborating our theoretical findings.

Keywords: Distributed cooperative control over networks, Neural networks, Automotive
Abstract: In many control systems, tracking accuracy can be enhanced by combining (data-driven) feedforward (FF) control with feedback (FB) control. However, designing effective data-driven FF controllers typically requires large amounts of high-quality data and a dedicated design-of-experiment process. In practice, relevant data are often distributed across multiple systems, which not only introduces technical challenges but also raises regulatory and privacy concerns regarding data transfer. To address these challenges, we propose a framework that integrates Federated Learning (FL) into the data-driven FF control design. Each client trains a data-driven, neural FF controller using local data and provides only model updates to the global aggregation process, avoiding the exchange of raw data. We demonstrate our method through simulation for a vehicle trajectory-tracking task. Therein, a neural FF controller is learned collaboratively using FL. Our results show that the FL-based neural FF controller matches the performance of the centralized neural FF controller while reducing communication overhead and increasing data privacy.

Keywords: Predictive control for linear systems, Reduced order modeling, Robust control
Abstract: We address the design of a model predictive control (MPC) scheme for large-scale linear systems using reduced-order models (ROMs). Our approach uses a ROM, leverages tools from robust control, and integrates them into an MPC framework to achieve computational tractability with robust constraint satisfaction. Our key contribution is a method to obtain guaranteed bounds on the predicted outputs of the full-order system by predicting a (scalar) error-bounding system alongside the ROM. This bound is then used to formulate a robust ROM-based MPC that guarantees constraint satisfaction and robust performance. Our method is developed step-by-step by (i) analysing the error, (ii) bounding the peak-to-peak gain, an (iii) using filtered signals. We demonstrate our method on a 100-dimensional mass-spring-damper system, achieving over four orders of magnitude reduction in conservatism relative to existing approaches.

Keywords: Predictive control for linear systems, Stochastic control, Output feedback
Abstract: We propose a stochastic model predictive control (MPC) framework for linear systems subject to joint-in-time chance constraints under unknown disturbance distributions. Unlike existing approaches that rely on parametric or Gaussian assumptions, or require expensive offline computation, the method uses conformal prediction to construct finite-sample confidence regions for the system's error trajectories with minimal computational effort. These probabilistic sets enable relaxation of the joint-in-time chance constraints into a deterministic closed-loop formulation based on indirect feedback, ensuring recursive feasibility and chance constraint satisfaction. Further, we extend to the output feedback setting and establish analogous guarantees from output measurements alone, given access to noise samples. Numerical examples demonstrate the effectiveness and advantages compared to existing approaches.

Keywords: Predictive control for nonlinear systems, Stability of nonlinear systems, Nonlinear system identification
Abstract: This paper derives conditions under which Model Predictive Control (MPC) with terminal conditions, using a data-driven surrogate model as a prediction model, asymptotically stabilizes the plant despite approximation errors. In particular, we prove recursive feasibility and asymptotic stability if a proportional error bound holds, where proportional means that the bound is linear in the norm of the state and the input. For a broad class of nonlinear systems, this condition can be satisfied using data-driven surrogate models generated by kernel Extended Dynamic Mode Decomposition (kEDMD) using the Koopman operator. Last, the applicability of the proposed framework is demonstrated in a numerical case study.

Keywords: Predictive control for nonlinear systems, Adaptive control, Optimal control
Abstract: This paper presents a dual model predictive control (MPC) framework for nonlinear systems with nonparametric uncertainties. We propose a novel Gaussian process-based MPC that leverages the posterior state covariance to explicitly capture the dual effect, balancing identification and control. By incorporating the covariance into the MPC objective, the resulting controller enables active learning, steering the system towards informative regions or to the desired setpoint. We establish robust constraint satisfaction of the novel dual MPC through a contingency plan. Numerical results demonstrate the effectiveness of the proposed dual MPC on a nonlinear system.

Keywords: Predictive control for nonlinear systems, Optimization algorithms, Uncertain systems
Abstract: We present Riccati-ZORO, an algorithm for tube-based optimal control problems (OCP). Tube OCPs predict a tube of trajectories in order to capture predictive uncertainty. The tube induces a constraint tightening via additional backoff terms. This backoff can significantly affect the performance, and thus implicitly defines a cost of uncertainty. Optimizing the feedback law used to predict the tube can significantly reduce the backoffs, but its online computation is challenging.

Riccati-ZORO jointly optimizes the nominal trajectory and uncertainty tube based on a heuristic uncertainty cost design. The algorithm alternates between two subproblems: (i) a nominal OCP with fixed backoffs, (ii) an unconstrained tube OCP, which optimizes the feedback gains for a fixed nominal trajectory. For the tube optimization, we propose a cost function informed by the proximity of the nominal trajectory to constraints, prioritizing reduction of the corresponding backoffs. These ideas are developed for ellipsoidal tubes under linear state feedback. In this case, the decomposition into the two subproblems yields a substantial reduction of the computational complexity with respect to the state dimension from bigO(n_x^6) to bigO(n_x^3), i.e., the complexity of a nominal OCP.

We investigate the algorithm in numerical experiments, and provide two open-source implementations: a prototyping version in CasADi and a high-performance implementation integrated into the acados OCP solver.

Keywords: Robust adaptive control, Predictive control for linear systems, Linear parameter-varying systems
Abstract: Robust data-driven controllers typically rely on datasets from previous experiments, which embed information on the variability of the system parameters across past operational conditions. Complementarily, data collected online can contribute to improving the feedback performance relative to the current system’s conditions, but are unable to account for the overall—possibly time-varying—system operation. With this in mind, we consider the problem of stabilizing a time-varying linear system, whose parameters are only known to lie within a bounded polytopic set. Taking a robust data-driven approach, we synthesize the control law by simultaneously leveraging two sets of historical state and input measures: an offline dataset ---which covers the extreme variations of the system parameters--- and an online dataset consisting of a rolling window of the latest state and input samples. Our approach relies on the data informativity framework, allowing a direct data-to-feedback design based on standard Lyapunov arguments. The procedure is implemented via semi-definite optimization: this also yields an upper bound on the cost-to-go for the class of systems that are consistent with the online data, while guaranteeing a decreasing cost for all systems compatible with the offline data. Numerical experiments are presented to illustrate the effectiveness of the proposed controller.

Keywords: Cooperative control, UAV's, Aerospace
Abstract: We present a method for the automatic derivation of runtime monitoring specifications for unmanned aerial vehicles (UAVs) based on symbolic control behavior.

Starting from a symbolic control policy, which is synthesized from formal mission specifications, we simulate the controller together with a reduced-order plant model to generate reference trajectories.

These are transformed into geometry based monitors composed of tolerance tubes and consistency metrics that compare observed and reference behavior.

The specifications are computed online, allowing situation dependent adaptation.

Monitors operate externally, require no access to internal controller states, and enable real time detection of deviations.

The approach is evaluated on physical experiments with micro UAVs.

Keywords: Distributed cooperative control over networks, Network analysis and control
Abstract: This paper presents a distributed, gradient-based deployment strategy for maximizing coverage in heterogeneous wireless sensor networks, particularly in environments with obstacles. Our approach extends a framework that leverages Green's theorem to transform the computationally expensive 2D area coverage integral into a tractable 1D line integral over the coverage boundary. The central contribution is the proof that this formulation allows the exact gradient of the total coverage function to be computed distributedly, where each sensor only requires information from its immediate, physically overlapping neighbors. In this work, we advance this method by seamlessly incorporating obstacles, modeling their blocking effect as additional line segments in the boundary integration. Unlike traditional Voronoi-based methods, our strategy avoids complex geometric partitioning, resulting in significantly lower computational overhead. The algorithm is proven to converge to a local maximum of the coverage objective, and its effectiveness in navigating complex environments is demonstrated through simulations.

Keywords: Game theoretical methods, Agents and autonomous systems, Transportation systems
Abstract: It is well-known that selfish routing policies from individual agents in a transportation network can generate inefficient congestion across the network. It is also understood that when all agents are coordinated, and route to minimize travel time for the entire fleet, optimal traffic flows emerge. However, the effects when a fraction of traffic is selfish and remaining traffic is coordinated is less studied. Thus, we seek to understand the worst-case harm caused when coordinated fleets and selfish traffic are simultaneously present in a transportation network (which we refer to as heterogeneous coordination), relative to complete selfishness. Specifically, we study the perversity index, which measures the worst-case ratio of the equilibrium congestion cost in the presence of coordinated traffic to the equilibrium congestion cost when all agents route selfishly. First, for the class of two-terminal parallel networks, we provide a strict upper bound on the perversity index and explore lower bounds via numerical experiments. Then, for any two-terminal network, we provide sufficient conditions in which coordinated fleets do not cause perversity (i.e., coordination causes no harm relative to selfishness). Our results are a novel direction in establishing the effectiveness of heterogeneous coordination relative to nominal agent interactions.

Keywords: Aerospace, Cooperative control, Cooperative autonomous systems
Abstract: This paper proposes a cooperative guidance strategy enabling multiple pursuers, interacting over a ring digraph, to intercept a target simultaneously. The approach models agents as double integrators arranged in a ring digraph with homogeneous emph{macro-vertices}. Conditions for network stability and positional consensus are derived. Using only two design parameters, independent of network size, the method ensures closed-loop stability. This flexibility aligns with the goals of cooperative simultaneous interception strategies, where the derived theoretical insights are applied. Simulation results validate the proposed framework, demonstrating coordinated and reliable simultaneous interception.

Keywords: UAV's, Cooperative control, Agents and autonomous systems
Abstract: This paper presents a decentralized trajectory-planning and cooperative control framework for multi-UAV navigation in cluttered 3D environment. The proposed method combines Rapidly-exploring Random Trees (RRT) with a virtual potential-field controller and a proximity-triggered communication policy to coordinate agents. A token-based prioritization scheme regulates inter-UAV interactions, promoting deadlock-free passage through confined areas and maintaining collision avoidance. We present extensive numerical simulation results that validate the proposed framework.

Keywords: Distributed cooperative control over networks, Concensus control and estimation, Cooperative autonomous systems
Abstract: This paper investigates distributed consensus in networked systems subject to input saturation and malicious nodes. To detect both abrupt and strategically gradual nodes' misbehavior, a hybrid reputation mechanism is introduced, combining long-term deviation analysis with short-term state comparison. Based on reputation calculation and classification, communication weights are adaptively adjusted accordingly, thereby suppressing or even isolating low-reputation nodes and amplifying the influence of high-reputation ones. Then, a distributed control law is designed to achieve semi-global leader-following consensus while ensuring control inputs remain within the linear region of the saturation function. Numerical simulations validate the effectiveness of the proposed reputation-driven detection and control method. Overall, our work enhances the security and resilience of networked systems, particularly in adversarial and constrained environments.

Keywords: Stochastic control, Safety critical systems, Constrained control
Abstract: We propose a simple safety filter design for stochastic discrete-time systems based on piecewise affine probabilistic control barrier functions, providing an appealing balance between modeling flexibility and computational complexity. Exact evaluation of the safety filter consists of solving a mixed-integer quadratic program (MIQP) if the dynamics are control-affine, (or a mixed-integer nonlinear program in general). We propose a heuristic search method that replaces this by a small number of small-scale quadratic programs (QPs), or nonlinear programs (NLPs) respectively. The proposed approach provides a flexible framework in which arbitrary (data-driven) quantile estimators can be used to bound the probability of safety violations. Through extensive numerical experiments, we demonstrate improvements in conservatism and computational cost with respect to existing methods, and we illustrate the flexibility of the method for modeling complex safety sets.

Keywords: Constrained control, Stability of nonlinear systems, Safety critical systems
Abstract: Control barrier functions (CBFs) are a flexible and efficient approach for enforcing dynamic safety constraints. We consider a class of nonlinear systems with singular perturbations and develop a CBF framework for such systems to enforce safety constraints on both the slow and fast subsystems in the overall singularly perturbed nonlinear system. We first develop a robustness property showing that if a controller for the slow subsystem satisfies the slow CBF condition with a safety margin, then for sufficiently small values of the singular perturbation parameter, the slow subsystem remains within a neighborhood of its safe set. We then consider the general case of safety constraints on both fast and slow subsystems and provide conditions under which joint slow-fast safety is achieved. We provide simulation studies for two example systems.

Keywords: Safety critical systems, Constrained control, Autonomous systems
Abstract: We study feasibility guarantees for safety filters developed using Control Barrier Functions (CBFs) when a safe set is defined using the pointwise minimum of continuously differentiable functions, a construction that is common for the backup CBF (BCBF) method and typically nonsmooth. We replace the minimum by its log-sum-exp (soft-min) smoothing and show that, under a strict safety condition, the smooth function becomes a CBF (or extended CBF) for a range of the smoothing parameter. For compact safe sets, we derive an explicit lower bound on the smoothing parameter that makes the smooth function a CBF and hence renders the corresponding safety constraint feasible. For unbounded sets, we introduce tail conditions under which the smooth function satisfies an extended CBF condition uniformly. Finally, we apply these results to BCBFs. We show that safety of a compact (terminal) backup set under a backup controller, together with a condition ensuring safety of the backup trajectories on the relevant boundary of the safe set, is sufficient for constraint feasibility for BCBFs. These results provide a recipe for a priori feasibility guarantees for smooth inner approximations of nonsmooth safe sets without the need for additional online certification.

Keywords: Safety critical systems, Constrained control, Optimal control
Abstract: In safety-critical control systems, ensuring both safety and feasibility under sampled-data implementations is crucial for practical deployment. Existing Control Barrier Function (CBF) frameworks, such as High-Order CBFs (HOCBFs), effectively guarantee safety in continuous time but may become unsafe when executed under zero-order-hold (ZOH) controllers due to inter-sampling effects. Moreover, they do not explicitly handle finite-time reach-and-remain requirements or multiple simultaneous constraints, which often lead to conflicts between safety and reach-and-remain objectives, resulting in feasibility issues during control synthesis. This paper introduces Sampling-Aware Control Barrier Functions (SACBFs), a unified framework that accounts for sampling effects and high relative-degree constraints by estimating and incorporating Taylor-based upper bounds on barrier evolution between sampling instants. The proposed method guarantees continuous-time forward invariance of safety and finite-time reach-and-remain sets under ZOH control. To further improve feasibility, a relaxed variant (r-SACBF) introduces slack variables for handling multiple constraints realized through time-varying CBFs. Simulation studies on a unicycle robot demonstrate that SACBFs achieve safe and feasible performance in scenarios where traditional HOCBF methods fail.

Keywords: Constrained control, Autonomous systems, Maritime
Abstract: This paper presents a collision avoidance method for a path-following controller for a system with high relative degree and constant velocity along the path. High-order control barrier functions and a coordinate transformation are employed to safely avoid static and dynamic obstacles by modifying the nominal control output. A novel proof formulation for high-order control barrier functions simplifies existing derivations. Feasibility of the optimization problem is studied and a low computational effort method for hazard detection is presented. The method is implemented for a path-following controller in nominal simulation and on a real-time hardware-in-the-loop simulator.

Keywords: Cooperative autonomous systems, Adaptive control, Safety critical systems
Abstract: This paper investigates the distributed safety critical control for multi-agent systems (MASs) in the presence of uncontrollable agents with uncertain behaviors. To ensure system safety, the control barrier function (CBF) is employed in this paper. However, a key challenge is that the CBF constraints are coupled when MASs perform collaborative tasks, which depend on information from multiple agents and impede the design of a fully distributed safe control scheme. To overcome this, a novel reconstructed CBF approach is proposed. In this method, the coupled CBF is reconstructed by leveraging state estimates of other agents obtained from a distributed adaptive observer. Furthermore, a prescribed performance adaptive parameter is designed to modify this reconstruction, ensuring that satisfying the reconstructed CBF constraint is sufficient to meet the original coupled one. Based on the reconstructed CBF, we design a safety-critical quadratic programming (QP) controller and prove that the proposed distributed control scheme rigorously guarantees the safety of the MAS, even in the uncertain dynamic environments involving uncontrollable agents. The effectiveness of the proposed method is illustrated through a simulation.

Keywords: UAV's, Autonomous systems, Intelligent systems
Abstract: This paper proposes a vision-based imitation learning framework for autonomous steering of fixed-wing unmanned aerial vehicles (UAVs). Departing from traditional pipelines that depend on explicit state estimation or decoupled perception-control modules, the presented method learns an end-to-end policy to directly map raw image sequences into steering commands. The approach formulates the control problem as a conditional flow matching task, enabling effective capture of the expert’s state-action distribution from demonstration data and resulting in smooth, stable control trajectories. Evaluation is conducted in a high-fidelity co-simulation environment that combines photorealistic visuals with high-fidelity flight dynamics. Compared to baseline methods such as Behavioral Cloning and Diffusion Policy, the proposed flow matching-based steering policy demonstrates improved trajectory accuracy, higher success rates under varying thresholds, and enhanced control smoothness, confirming its capability for reliable vision-based navigation in the absence of global navigation satellite systems (GNSS).

Keywords: UAV's, Machine learning, Neural networks
Abstract: This article introduces a novel sample-efficient curriculum learning (CL) approach for training an end-to-end reinforcement learning (RL) policy for robust stabilization of a Quadrotor. The learning objective is to simultaneously stabilize position and yaw-orientation from random initial conditions through direct control over motor RPMs (end-to-end), while adhering to pre-specified transient and steady-state specifications. This objective, relevant in aerial inspection applications, is challenging for conventional one-stage end-to-end RL, which requires substantial computational resources and lengthy training times. To address this challenge, this article draws inspiration from human-inspired curriculum learning and decomposes the learning objective into a three-stage curriculum that incrementally increases task complexity, while transferring knowledge from one stage to the next. In the proposed curriculum, the policy sequentially learns hovering, the coupling between translational and rotational degrees of freedom, and robustness to random non-zero initial velocities, utilizing a custom reward function and episode truncation conditions. The results demonstrate that the proposed CL approach achieves superior performance compared to a policy trained conventionally in one stage, with the same reward function and hyperparameters, while significantly reducing computational resource needs (samples) and convergence time. The CL-trained policy's performance and robustness are thoroughly validated in a simulation engine (Gym-PyBullet-Drones), under random initial conditions, and in an inspection pose-tracking scenario. A video presenting our results is available at https://youtu.be/9wv6T4eezAU.

Keywords: UAV's, Hybrid systems, Maritime
Abstract: This paper presents a phase-aware hybrid con- troller for autonomous quadrotor landing on an oscillating platform. The key idea is to exploit the instantaneous phase of the deck motion – rather than predicting it – by switching between eight discrete control rules that encode physically meaningful behaviors. The controller partitions the deck heave into descending, crest, trough, and ascending phases, and assigns a rule-dependent altitude bias that shapes the reference relative to the offset-compensated deck height. The controller operates at 100 Hz using only VRPN pose feedback and a causal sliding-window estimate of deck velocity. Experiments conducted on a motion platform across seven excitation levels and one static case demonstrate repeatable phase-locked landings with touchdown velocity differences be- low 0.07 m/s and landing durations between 6 and 13 s under the tested sinusoidal heave motions.

Keywords: Robotics
Abstract: Soft robots are an emerging and fast-growing field of research with applications in various areas. With the rise of new soft robotic applications, control requirements increase. Thus, control of dynamic motion is becoming more and more important. The main challenges in controlling soft robots stem from their large, continuous deformations, which lead to many degrees of freedom, as well as the difficulty of sensor integration, which leads to a limited availability of control feedback. Additionally, unknown material properties and comparatively large manufacturing inaccuracies hinder reliable modeling of soft robots. So far, in soft robotics mainly so-called quasi-static control approaches, which neglect the dynamics of the system, are used. However, modern soft robotic applications demand increasingly faster and more accurate movements. In these cases, quasi-static controllers are insufficient, and controllers which account for the robots dynamics are needed.

In this work different feedforward and feedback control approaches for dynamic trajectory tracking control of soft robots are implemented and examined. A neural-network-based quasi-static feedforward controller, a dynamic feedforward controller based on the inversion of a dynamic piecewise constant curvature (PCC) model using servo-constraints, and the combination of the two different feedforward controllers with a linear quadratic regulator with integral action (LQRi) and gain scheduling are considered. For feedback control, only sensor feedback obtained from a curvature sensor integrated into the soft robot is used. The performance of these approaches is compared in simulations and experiments.

Keywords: UAV's, Lyapunov methods, Predictive control for nonlinear systems
Abstract: In this paper, we develop a unified Pontryagin-based Nonlinear Model Predictive Control (PMP–NMPC) framework that embeds Pontryagin’s Minimum Principle directly into the predictive control architecture, providing structured optimality conditions for nonlinear optimal control. As a motivating application, we consider close-proximity aerial maneuvers, particularly quadrotor landing, where downwash-induced aerodynamic disturbances create highly nonlinear residual forces. To address this challenge, a Deep Neural Network is integrated into the PMP–NMPC loop to learn residual downwash dynamics and incorporate them into the control law for disturbance rejection. A key theoretical contribution is a rigorous local finite-time stability analysis establishing closed-loop stability in the presence of learned disturbances. Simulations of landing scenarios demonstrate stable descent, rapid adaptation to downwash, and improved robustness compared with standard NMPC, highlighting the effectiveness of the proposed framework for disturbance-rich close-proximity flight.

Keywords: Robotics, Observers for nonlinear systems, Feedback linearization
Abstract: The end-effector tracking accuracy of rigid manipulators can be degraded by base-induced vibrations. This paper presents a robust control framework that explicitly accounts for model uncertainties, unmodeled dynamics, and external disturbances, including base vibrations. The framework first applies partial feedback linearization to simplify the base–manipulator model, resulting in a lumped disturbance term that captures these combined effects. A nonlinear disturbance observer is then employed to estimate this disturbance, and its effect on the output is compensated through output-feedback linearization. Stability analysis ensures the uniform ultimate boundedness of the closed-loop system. The proposed method was validated through physics-based simulations. Results show enhanced robustness in tracking performance, with higher observer gains enabling the system to recover performance closer to the disturbance-free case. This demonstrates the potential of the proposed method for manipulators operating on moving or vibration-prone bases.

Keywords: Control over communication, Optimal control of communication networks, Network analysis and control
Abstract: We consider event-triggered linear-quadratic-Gaussian (LQG) control when sensor updates are transmitted over an i.i.d. packet–erasure channel. Although the optimal controller in a standard LQG setup is available in closed form, choosing when to transmit remains computationally and analytically difficult because packet drops randomize packet delivery and couple scheduling decisions with the estimation-error dynamics, making direct dynamic-programming solutions impractical. By certainty equivalence, the co-design problem becomes choosing a binary send/skip sequence that balances control performance and communication cost. We derive a closed-form expansion of the error covariance as precomputable Gramian terms scaled by a survival factor that depends only on the number of transmission attempts on each interval. This converts the problem into an unconstrained binary program that we linearize exactly via running attempt counters and a one-hot encoding, yielding a compact mixed-integer linear program (MILP) well-suited to receding-horizon implementation. On the linearized Boeing-747 benchmark, a model predictive control (MPC) scheduler lowers cost while attempting far fewer transmissions than a one-shot baseline across channel success rates.

Keywords: Control over networks, Linear systems, Safety critical systems
Abstract: This paper is concerned with safety preserving scheduling for networked control systems (NCSs) whose shared communication networks have a limited communication capacity. Given the plant dynamics, the controller dynamics, the time horizon of interest, the capacity of the communication network, the set of possible initial states and the set of unsafe states, we present an algorithm that designs a scheduling logic under which all state trajectories of all the plants in the NCS avoid the unsafe states for the entire time horizon. The primary apparatuses for our analysis are graph theory and multiple barrier functions. A numerical example is presented to demonstrate our results.

Keywords: Control over networks, Predictive control for linear systems, Constrained control
Abstract: We study the design of an offloaded model predictive control (MPC) operating over a lossy communication channel. We introduce a controller design that utilizes two complementary bandwidth-reduction methods. The first method is a multi-horizon MPC formulation that decreases the number of optimization variables, and therefore the size of transmitted input trajectories. The second method is a communication-rate reduction mechanism that lowers the frequency of packet transmissions. We derive theoretical guarantees on recursive feasibility and constraint satisfaction under minimal assumptions on packet loss, and we establish reference-tracking performance for the rate-reduction strategy. The proposed methods are validated using a hardware-in-the-loop setup with a real 5G network, demonstrating simultaneous improvements in bandwidth efficiency and computational load.