A unified neural event-triggered control approach of high-order switched uncertain systems with time-varying state constraints
Chunyu Chu, Youqian He
Abstract
Purpose This work proposes a unified adaptive event-triggered tracking control scheme for high-order switched uncertain nonlinear systems, subject to asymmetric time-varying state constraints. The purpose of this paper is to guarantee closed-loop stability and boundedness under average dwell time switching while reducing communication load. Design/methodology/approach System uncertainties are approximated via neural networks, and finite-time differentiators are used to prevent the “explosion of complexity” in backstepping. High-order terms are handled with separable functions and auxiliary lemmas instead of power integrators. Furthermore, a universal-constrained nonlinear function and coordinate transformations flexibly manage both constrained and unconstrained cases without feasibility conditions. Findings The proposed adaptive event-triggered controller guarantees strict stability of the closed-loop system and boundedness of all signals under switching signal. Furthermore, the event-triggered mechanism effectively reduces unnecessary control updates, thereby saving communication resources. Simulation results validate the effectiveness and robustness of the proposed control scheme. Originality/value Unlike conventional designs, the method avoids iterative embedding of virtual control gains, which simplifies controller construction. It unifies constrained and unconstrained cases and integrates separable-based tools with event-triggered control, offering a novel and efficient solution for high-order switched nonlinear systems.