Observer-Based Event-Triggered Tracking Control of Multi-Player Nonlinear Systems via Adaptive Dynamic Programming

Abstract

This paper investigates event-triggered optimal tracking control (OTC) problems of unknown nonlinear multiplayer systems by using adaptive dynamic programming. To begin with, a neural network-based observer is constructed to obtain the unknown system dynamics. By establishing an augmented system and designing a discount performance index function, the OTC problem is transformed to an optimal regulation problem. Subsequently, critic-only structure is adopted to solve the event-triggered tracking Hamilton-Jacobi-Bellman (HJB) equation. Moreover, for the purpose of reducing the computing and communication burdens, a novel triggering condition which is suitable for multiple controllers is designed via Lyapunov’s direct method, and tracking controllers are renovated at triggering moments only. Theoretical analysis shows that the tracking error is guaranteed to be uniform ultimate bounded. Finally, simulation example is provided to validate the effectiveness of the developed method.