Adaptive critic designs for decentralised robust control of nonlinear interconnected systems via event-triggering mechanism

Abstract

In this article, a novel decentralised event-triggered control scheme is developed for the robust stabilisation of interconnected systems. The existing literature about decentralised robust control by adaptive critic designs (ACDs) is mostly time-triggered which requires extensive communication and computation. To extend the ACD framework and save the communication and computational resources, the event-triggered decentralised control using ACDs is investigated. First, we utilise a series of auxiliary subsystems to restructure the interconnected systems in a distributed way, so as to transform the robust control problem into the corresponding optimal control problem. Then, adaptive critic designs with event-triggering mechanism are investigated to solve the Hamilton–Jacobi–Bellman equations via local policy iteration algorithm. The single-critic networks are employed to achieve the approximation of optimal value functions. The Lyapunov stability proof ensures the asymptotic stability of the whole system and the uniform ultimate boundedness of the critic neural networks. Finally, an interconnected plant is employed to demonstrate the effectiveness of proposed method.