Impulsive Consensus of One-Sided Lipschitz Multi-Agent Systems With Deception Attacks and Stochastic Perturbation

This article addresses the consensus of one-sided Lipschitz multi-agent systems (MASs) subject to deception attacks and stochastic perturbation by using impulsive protocols. First, the one-sided Lipschitz condition is considered because it has a wider application range and less conservatism than general Lipschitz condition. Second, the impact of stochastic perturbation is considered, which is more consistent with the practical situation. Third, deception attacks are considered in controller-to-actuator channels and modeled as the Bernoulli distribution. In addition, two kinds of deception attacks are considered, namely injection attacks and replacement attacks. Then, based on Lyapunov stability theory, some results are obtained to accomplish consensus of MASs with both deception attacks and stochastic perturbations where the mean-squared consensus errors are bounded. What is more, the relationship between impulsive interval, convergence rate, and error bound are presented in this article. Based on this relationship, this article gives the rules for selecting these parameters, which can make a better control effectiveness. Finally, the validity of obtained results are proved via simulation examples.

I. Introduction

In The past few decades, inspired by cooperative biological phenomena observed in nature, such as fish swarms and bird migration, multiagent systems (MASs) have obtained wide focus due to its advantages in group tasks [1], [2], [3]. As the basis of cooperative problem, the distributed consensus of MASs has become a popular research topic [4], [5].