Secure Consensus Of Multi-agent Systems Under Impulsive Control

In recent years,multi-agent systems have gained widespread use for their high reliability and strong robustness.Consensus is a central issue in the study of multi-agent systems and is a major concern for scholars.Due to the complexity of the network and the openness of the system,multi-agent systems are vulnerable to attacks from external sources,with typical attacks including deception attacks,denial-of-service(DoS)attacks.Therefore,research on secure consensus issues in multi-agent systems is urgently needed.Meanwhile,the complexity of topology in actual network environments is also an important factor to consider.This article focuses on single-layer and double-layer network structures,and studies the secure consensus issues of multi-agent systems under deception attacks and DoS attacks.The main research contributions are as follows:1.We investigate the secure consensus problem of nonlinear multi-agent systems under deception attacks,and propose an event-triggered impulsive control method with an upper bound on the triggering time,in combination with fixed-time interval impulsive control.We obtain sufficient conditions for the secure consensus of the multi-agent system and verify that the proposed event-triggered impulsive control method can automatically adjust the impulsive time interval to achieve fast convergence to secure consensus.2.We consider the secure consensus problem of nonlinear multi-agent systems under deception attacks in a two-layer network structure,consisting of a leader layer and a follower layer,with dynamical switching network topology.By using impulsive control methods,based on knowledge of graph theory,Lyapunov stability theory,and linear matrix inequalities,we analyze the secure consensus problem for the multi-agent system in leader layer and the secure consensus problem for the multi-agent system in two-layer network.3.We propose a hybrid attack model consisting of DoS attacks and deception attacks and study the secure consensus problem of nonlinear multi-agent systems.The network is composed of a leader agent and several follower agents,and the hybrid attack can cause network disconnection and false data injection.By using impulsive control methods,we obtain the secure consensus of the multi-agent system.