European Journal of Control最新文献

This paper presents a conceptual framework for addressing the consensus problem in multi-agent systems with unknown nonlinear dynamics using reinforcement learning (RL) based nearly optimal sliding mode controller (SMC). The agents’ dynamics are assumed to have uncertainty and mismatched disturbance. An adaptive fixed-time estimator is introduced to estimate uncertain dynamics and disturbances for each agent at a certain time. The paper proposes two control strategies. In the first strategy, a controller is designed, incorporating adaptive SMC and an optimal controller. Adaption law in SMC estimates the bound of fixed time estimator error before convergence, ultimately achieving asymptotic convergence to the sliding surface and converting the agent’s dynamics to linear. This enables solving a linear consensus problem using an RL-based adaptive optimal controller through an on-policy critic–actor method. The second control strategy enhances the adaptive SMC into a fixed-time controller, reducing the time to reach the sliding surface regardless of initial conditions. Consequently, the convergence time of the consensus error to zero is diminished. This reduction in reaching time results in faster convergence of the consensus error to zero. The effectiveness of both strategies is validated through numerical experiments on two real system models, aligning with the theoretical proofs.

This paper deals with the stabilization of aperiodic sampled-data switched affine systems to a predetermined hybrid limit cycle using a hybrid dynamical system representation and a control Lyapunov function approach. Some preliminaries on the hybrid dynamical system formalism provide the framework for modeling switched affine systems followed by definitions on hybrid limit cycles and related notions. The main result, based on simple Linear Matrix Inequalities (LMI), guarantees that the solutions to the closed-loop system converge to a hybrid limit cycle defined by the states, functioning modes with their corresponding dwell times. The theoretical results are evaluated on academic examples and demonstrate the potential and the originality of the method over the recent literature.

This paper proposes model-based and model-free policy gradient methods (PGMs) for designing dynamic output feedback controllers for discrete-time partially observable deterministic systems without noise. To fulfill this objective, we first show that any dynamic output feedback controller design is equivalent to a state-feedback controller design for a newly introduced system whose internal state is a finite-length input–output history (IOH). Next, based on this equivalence, we propose a model-based PGM and show its global linear convergence by proving that the Polyak–Łojasiewicz inequality holds for a reachability-based lossless projection of the IOH dynamics. Moreover, we propose a model-free implementation of the PGM with a sample complexity analysis. Finally, the effectiveness of the model-based and model-free PGMs is investigated through numerical simulations.

This article deals with the robustness of large-scale structured systems in terms of controllability when subject to failure of links from the inputs to the state variables (i.e., input-links). Firstly, we consider a deletion problem of determining the minimum number of input-links, if removed, lead to a structurally uncontrollable system. This problem is known to be NP-hard. We prove that it remains NP-hard even for strongly connected systems. We develop efficient polynomial time methods to solve this problem optimally/suboptimally under suitable assumptions imposed on the generic rank of the state matrix. The assumptions imposed are often satisfied by a large class of systems. These methods mainly use the notion of Dulmage–Mendelsohn decomposition of bipartite graphs and minimum vertex cover problem for undirected graphs. Secondly, we consider an addition problem whose goal is to identify a set of input-links of minimum cardinality to be added between the existing inputs and the state variables in order to preserve structural controllability with respect to failure of an arbitrary input-link. We establish that this particular problem is NP-hard and even inapproximable to a multiplicative factor of $logp$, where $p$ is the number of critical input-links in the system. Additionally, we identify several practically relevant tractable cases associated with this problem. Finally, an example illustrating the usefulness of the methods developed is given in this article.

This paper presents the design of a Lyapunov-based Luenberger-like observer to simultaneously estimate the non-available angular velocity and the unknown constant load torque of rotating machines. This proposal assumes that the only measurable output is the rotor angle, which is considered to be defined in the circle, its natural topology. This provides nonlinear dynamics where the Lyapunov-based Luenberger-like observer globally converges. Moreover, a trajectory tracking controller for the motor angular velocity is also designed by feeding back the observer estimations. Convergence analysis of the observer-controller system is performed using Lyapunov-like theory. The effectiveness and performance of the system are validated through tests performed on an experimental platform.

This research work is devoted to boundary control of a catalytic cracking reactor using state and error feedback regulators. The process mathematical model is governed by a set of partial differential equations (PDEs), for which infinite-dimensional representation and spectral properties of the system generator are employed to solve the regulation problems. The primary objective is to track a desired output reference trajectory in the presence of disturbances that are generated by a distributed parameter exosystem. Initially, a state feedback stabilizing regulator is designed to drive the process output towards the reference trajectory. The second objective is to develop a dynamic controller that employs the tracking error as input. The closed-loop plant is shown to be exponentially stable and the tracking error asymptotically approaches zero. The performances of the designed regulators are shown through numerical simulations.

This paper presents a methodology to simultaneous estimate state variables and actuator fault for descriptor nonlinear parameter-varying systems. This class of systems is composed of an interpolated parameter-varying linear part and Lipschitz nonlinear one. The proposed observer structure allows some robustness and steady state accuracy in face to parametric uncertainties. The conditions of existence and stability are given in terms of LMIs (Linear Matrix Inequalities). The obtained results are illustrated on the heat exchanger system with two countercurrent cells model with actuator fault.

For nonlinear systems with asymmetric input saturation, an adaptive fixed-time control strategy via multi-dimensional Taylor network (MTN) is proposed to effectively solve the tracking control problem. Firstly, the Gaussian error function and the differential mean value theorem are used to simulate the asymmetric input saturation model, that is, the nonlinear model is transformed into a linear model with a bounded perturbation. Secondly, MTNs are applied to control the design process, which approximate the nonlinear structures with linear combination of polynomials. Thirdly, in addition to the tracking error converging to an arbitrarily small neighborhood around the origin within a fixed time, the proposed control scheme not only guarantees that any signal in the controlled system remains bounded, but also avoids the problem of finite-time control relying on the initial state. Finally, three examples are used to validate the feasibility and superiority of the scheme.

Flow control aims at modifying a natural flow state to reach an other flow state considered as advantageous. In this paper, active feedback flow separation control is investigated with two different closed-loop control strategies, involving a reference signal tracking architecture. Firstly, a data-driven control law, leading to a linear (integral) controller is employed. Secondly, a phenomenological/model-driven approach, leading to a non-linear positive (integral) control strategy is investigated. While the former benefits of a tuning simplicity, the latter prevents undesirable effects and formally guarantees closed-loop stability. Both control approaches were validated through wind tunnel experiments of flow separation control over a movable NACA 4412 plain flap. These control laws were designed with respect to hot-film measurements, performed over the flap for different deflection angles. Both control approaches proved efficient in avoiding flow separation. The main contribution of this work stands in providing practitioners, simple but yet efficient control design methods for the flow separation phenomena. Equivalently important, a complete validation campaign data-set is also provided.

This paper is devoted to the design of adaptive nonlinear prescribed performance controllers for balancing constraint and robustness. Firstly, a shift function is introduced to enhance robustness in the initial stage. Secondly, a new finite-time performance function and a regulation of dynamic adjustments are developed for the first time. Due to the regulation, it can be prevented by the proposed scheme that tracking errors approach or exceed performance boundaries during operation, which ensures both robustness and constraint, and the performance function is differentiable of order n, which can be used in backstepping design for higher-order systems without approximators. Finally, the designed controller is more robust in both transient and stable states, whose effectiveness and superiority are validated by stability analysis, matlab simulation, and an actual test on a real robotic arm.