Further results on event-triggered H∞ networked control for neural networks with stochastic cyber-attacks

Abstract

This paper is concerned with decentralized event-triggered H∞ networked control for neural networks (NNs) subject to two types of stochastic cyber-attacks. Firstly, a new dynamic event-triggered scheme is introduced to monitor the sampled data transmissions, and two independent Bernoulli distributed variables are used to describe the randomly occurring cyber-attacks. Secondly, based on the networked control, the closed-loop system is constructed under the stochastic cyber-attacks and limited network bandwidth. Thirdly, by the Lyapunov-Krasovskii functional (LKF) approach, an improved stability criterion is established to ensure the closed-loop system is mean-square asymptotical stability with a prescribed H∞ performance. Based on the criterion, desired control gain is determined. Finally, the effectiveness of the obtained result is illustrated by two numerical examples.