Reinforcement learning-based dynamic event-triggered prescribed performance control for nonlinear systems with input delay
Hao Xu, Ning Zhao, Ning Xu, Ben Niu, Xudong Zhao
Abstract
This article addresses the event-triggered optimised prescribed performance control problem for nonlinear strict feedback systems under input delays and actuator faults. First, an identifier-actor-critic architecture is employed to implement the adaptive reinforcement learning (RL) algorithm, ensuring optimal control performance. To address input delays, this article incorporates auxiliary compensation variables, enabling the conversion of the entire system into un-delayed system. A prescribed performance error transformation is constructed to guarantee that the tracking error converges to an arbitrarily small residual set, thereby achieving the desired tracking performance. Furthermore, an adaptive control strategy is designed to mitigate system instability caused by actuator faults. A dynamic event-triggered mechanism is also proposed, where threshold parameters are adaptively adjusted based on real-time tracking performance. Finally, the effectiveness of the proposed strategy is validated through a simulation example.