Dynamic Event-Triggered Adaptive Neural Network Control for Stochastic Nonholonomic Uncertain Systems

Abstract

In this article, we construct a dynamic event-triggered adaptive control strategy for a class of stochastic nonholonomic uncertain systems. The dynamic event-triggered mechanism can make the threshold adjustable, which is firstly considered for stochastic nonholonomic uncertain systems to conserve communication resources. We propose a state-input scaling transformation that converts the stochastic nonholonomic uncertain systems into a new form that facilitates controller design. By introducing a novel auxiliary dynamic variable to design a dynamic event-triggered mechanism (DETM) and defining a suitable parameter, we propose a new dynamic event-triggered adaptive neural network controller, which contains only one adaptive law. It is shown that the proposed control strategy can greatly reduce the computational complexity and communication burden, and the input-to-state stability (ISS) assumption is no longer needed. Simultaneously, all signals in the closed-loop system are ensured to be uniformly ultimately bounded (UUB) in probability. Then, the uncontrollability phenomenon is eliminated by constructing an adaptive event-triggered control-based switching strategy. In addition, the efficacy of the proposed controller is demonstrated through simulation results.