International Journal of Robust and Nonlinear Control最新文献

Model predictive control (MPC) based on the Koopman operator is an effective data-driven control method for nonlinear systems. However, modeling errors are almost inevitable when applying the Koopman operator to model the nonlinear systems and influence the control performance. It is an important problem to guarantee the robustness of Koopman MPC under modeling approximation errors. Aiming at the above problem, this paper proposes a data-driven approach to handle modeling errors and develops robust min–max MPC solutions with the Koopman operator for nonlinear discrete-time dynamical systems. The data-driven method is proposed to identify the modeling error as the polytopic uncertainty within the Koopman operator framework, and its effectiveness analysis is given. Based on the identified modeling error, the paper presents two design schemes for min-max robust predictive controllers. The first scheme is formulated as the feedback control computed by minimizing the upper bound of the worst-case value of an infinite-horizon quadratic objective function at each sampling time. To formulate the other scheme, two algorithms are presented to estimate the current model based on the identified polytope. With the help of the estimated model, the scheme allows the inclusion of the first control move separately from the rest of the control sequence governed by the feedback law, which can achieve better performance, which is an improved version of the first one. Recursive feasibility and closed-loop stability guarantees are provided. Numerical simulations verify the modeling and control effectiveness of the proposed algorithms.

This research investigates the design of state feedback controllers for Networked Control Systems (NCSs) subject to actuator failures alongside hybrid cyber-attacks, specifically Denial-of-Service (DoS) and deception attacks. To reduce the use of network resources, two distinct dynamic event-triggered schemes (DETSs) are proposed. These DETSs govern the data transmission scheduling for both measurement outputs and control inputs across the communication network. A key contribution lies in the simultaneous consideration of actuator malfunctions and the aforementioned hybrid cyber-attack scenarios. Utilizing Lyapunov stability theory, conditions for the mean square asymptotic stability of the resultant augmented closed-loop system are established. Subsequently, controller gains and event-triggering parameters are co-designed through the application of linear matrix inequality (LMI) techniques. Finally, a simulation example validates the theoretical results, demonstrating the proposed method's efficacy in mitigating the adverse impact of such cyber-attacks and actuator failures on networked control systems' performance.

In this article, the delay margin problem of linear systems with multiple time delays is investigated. First, estimations for the delay margin of uncontrolled system are provided by the argument principle. Then, in order to achieve wider delay range for robust stability of the control system, an output feedback controller is designed to improve the delay margins. The output feedback gain matrices can be computed by solving the optimization problems with respect to the delay margin. Finally, numerical examples to illustrate the effectiveness of the approach are given in this article.

This paper studies the issue of trajectory tracking control of quadrotor unmanned aerial vehicles in the presence of external disturbances. A prescribed performance control method with more general performance functions is proposed to guarantee the predesigned performance, in which Chebyshev neural networks are used to estimate and compensate for external disturbances, and the norm approximation approach is applied to minimize the computational complexity of neural networks. Moreover, an event-triggered mechanism with general formal threshold functions is introduced to reduce controller update frequency. Finally, simulation results of the proposed methods demonstrate their effectiveness and superiority.

Stabilizing nonlinear distributed parameter systems within the framework of the late lumping approach is a challenging problem. Diffusion-reaction systems, described by semilinear partial differential equations, represent a common class of distributed parameter systems encountered in real applications. In this work, the zeroing dynamics method is used to design a stabilizing infinite-dimensional controller for this important class of systems. Both distributed and boundary actuations are investigated. The design of the controller relies on the concept of the characteristic index. In the case of distributed control, as the characteristic index is finite, the design is straightforward. However, in the case of boundary control, the characteristic index is infinite, thus an equivalent pointwise control form is derived to reduce it to a special case of distributed control, which enables the controller design. In the disturbance-free case, the developed controllers ensure exponential stability. In the presence of disturbances, the closed-loop system remains stable, and the steady-state error can be made arbitrarily small by tuning the controller parameters. Four common application examples from the literature are provided to demonstrate the effectiveness of the developed controllers: catalytic rod, monoenzyme system, heated rod, and FitzHugh-Nagumo equation.

Unmanned vehicle teams are envisaged to have broad application prospects and play a key role in future various cooperative engineering projects. This paper studies the formation control protocol design problem via input-output system data for multiple underwater vehicle networks involving highly nonlinear dynamics, strong couplings between the rotational motion and the translational motion, and switching topologies. A distributed data-driven formation control protocol based on reinforcement learning is developed for multiple underwater vehicles to achieve the translational formation and rotational synchronization with the leader, without the requirement of the vehicle dynamics information. The optimality of the proposed formation controller for UUVs is achieved via the reinforcement learning approach. Simulation results for a multi-vehicle network demonstrate the effectiveness of the developed data-driven formation control protocol.

This article investigates the state estimation problem (SEP) for complex networks (CNs) with switching topologies subject to denial-of-service (DoS) attacks, utilizing an event-triggering strategy to reduce data transmissions. It is noteworthy that DoS attacks may occur not only on topology channels but also on output channels. Different from previous works in which DoS attack behavior follows Bernoulli distribution, this article employs a more general deterministic DoS model. To achieve state estimation under these conditions, we develop an extended Kalman filter (EKF) that switches between two different dynamics and derive a piecewise Riccati-like difference equation whose solution provides an upper bound matrix for the estimation error covariance (EEC). Subsequently, a criterion is established to ensure that the estimation error (EE) remains bounded in mean square in the presence of DoS. We derive a duty cycle for DoS attacks below which the boundedness of the EE is maintained. Finally, the effectiveness of our algorithm is certified by a simulation example.

This paper is devoted to studying the formation and tracking problem of an unknown target utilizing only partial output information from the target. Each agent is modeled as an ellipse, serving to simulate the agent's shape and create safety zones to prevent collisions during the movement. Specifically, by utilizing the partial output information of the target, a distributed reduced-order observer is created to estimate the target's state. Furthermore, a formation control algorithm is proposed based on the separation conditions of ellipses and the estimated state to track the target system. Leveraging LaSalle's invariance principle, it is demonstrated that the unknown mobile target can be tracked in formation without collisions. Finally, the effectiveness of the proposed algorithm is ultimately validated through simulation examples.

This article proposes a new fault detection approach for polynomial fuzzy systems with parametric uncertainties using $��{�H�}_{�\infty �}�$ filter design and zonotopic analysis. First, a polynomial fuzzy $��{�H�}_{�\infty �}�$ filter that is robust against uncertainties is designed by solving a set of sum-of-square constraints. Second, zonotopic analysis of residual is presented to obtain the guaranteed adaptive residual thresholds. After that, zonotopic residual evaluation schemes, including zonotope-based and interval-based evaluation schemes, are presented to achieve fault detection. Then, a numerical simulation is presented, which shows that the proposed fault detection approach can improve the performance of fault detection compared with the existing $��{�H�}_{�-�}�/�{�L�}_{�\infty �}�$ ellipsoid-based method. Finally, we apply the method to a tunnel diode circuit system to validate the effectiveness of the presented polynomial fuzzy approach.

In response to the challenges of high energy consumption, difficulty in detecting maneuvering targets, and limited interceptor maneuver overload in the large airspace and long-time domain, a data-driven optimal coverage formation and maintenance strategy is proposed in three-dimensional space. First, the nonlinear dynamic model of heterogeneous interceptors is linearized through feedback linearization. Then, considering the time-varying characteristics of the communication topology among heterogeneous interceptors, a distributed observer is proposed to estimate the expected positions of multiple interceptors. Second, based on integral reinforcement learning, an offline policy iteration control algorithm is proposed to overcome the constraints of overload and dependence on target models. It can learn the optimal control strategy online and prove the optimality of the strategy as well as the convergence and stability of the algorithm. Finally, the effectiveness of the strategy is verified through numerical simulation experiments, demonstrating its ability to achieve energy-optimal coverage formation tracking of high-speed maneuvering targets in the large airspace and long-time domain.