Fixed-Time, Robust and Fuzzy Adaptive Consensus Control with Event Trigger Mechanism Based on a Novel ZNN Scheme

Abstract

As a systematic approach, zeroing neural network (ZNN) is an elegant tool in control applications. However, the application of ZNNs in multi-agent systems still needs further research. Adaptive control schemes with adjustable convergence speed are important in practical application, but researchers mainly use explicit and direct rules to update the convergence parameter, which is less effective than using fuzzy logic system (FLS). In this study, combined with FLS, a novel event-triggered fuzzy adaptive ZNN (ET-FAZNN) model is proposed to solve the consensus problem in a fixed time. With a set of predefined fuzzy rules, the convergence rate is adaptively adjusted after an overall evaluation of the system state. We also boost computing efficiency by introducing the event-triggered mechanism. Specifically, we first present a detailed theoretical analysis to show that the novel ET-FAZNN model is fixed-time stable and robust. Then, we estimated the upper bounds of settling-time functions through a novel method based on the improper integral. Finally, four numerical experiments are presented to further verify the fixed-time convergence, adaptiveness, robustness, and the superior computing efficiency compared with conventional fuzzy adaptive ZNN.