Decentralized Robust Tracking Control of Interconnected Nonlinear-Constrained Systems: A Dynamic Event-Sampled Method

Abstract

This paper presents a decentralized robust dynamic event-sampled tracking (EST) control law for interconnected nonlinear-constrained systems. The core of developing such a control law is to convert the original EST control problem into the event-sampled decentralized stabilization problem of augmented interconnected systems. To address the transformed decentralized stabilization problem, an indirect approach relying on the optimal control methodology is proposed. Initially, a group of cost functions are constructed for the nominal subsystems related to the augmented interconnected systems. Then, the dynamic event-sampling mechanisms are introduced for lessening the computational burden. Meanwhile, the event-sampled Hamilton–Jacobi–Bellman equations (ES-HJBEs) are proposed for the augmented interconnected systems. To approximately solve the ES-HJBEs, the critic approximators are used with their parameters tuned under the reinforcement learning framework. After that, the uniform ultimate boundedness of the tracking errors and the approximators' parameter estimation errors are assured based on the Lyapunov theorem. Finally, a nonlinear plant is provided to validate the decentralized robust dynamic EST control law.