Adaptive Periodic Event-Triggered Stabilization of Switched Neural Networks Under the Merging Signal Scheme

This article is concerned with the exponential stabilization of switched neural networks (SNNs) with asynchronous switching. To save the limited bandwidth effectively, an adaptive periodic event-triggered mechanism (APETM) with a novel adaptive rule is excogitated, in which the threshold function is updated at each sampling instant to quickly respond to system changes and a tuning parameter is introduced to increase the adjustable range of the threshold function. A merging signal is constructed and then the closed-loop system is established to carry out stability analysis for asynchronous and synchronous switching cases within a unified framework. Then, based on the merging signal, a corresponding Lyapunov functional that includes a looped functional is constructed. This looped functional helps to reduce conservatism of the stability criterion. Finally, examples, including relevant discussions and comparisons, are shown to illustrate the efficacy of our developed results.