Event-triggered Control For Consensus Of Multi-agent Systems Subject To Deception Attacks

Multi-agent systems are favored by many researchers due to their wide application in areas such as neural networks,transportation,drones,and wireless communications.In the study of multi-agent systems,consensus control of multi-agent systems is an important research direction,with commonly used control methods including sample-data control,impulse control,and intermittent control.The key of consensus control is to design effective control strategies,so that the dynamic characteristics of all agents in multi-agent systems with or without leaders can achieve consensus after sufficient information interaction.Continuous information interaction can improve the timeliness and the amount of information.However,this will inevitably places high demands on communication equipment performance and may result in unnecessary resource consumption.Therefore,the event-triggered control strategy is proposed to save network communication resources effectively.On the other hand,during operation,event-triggered controllers may be subjected to different degrees of external attacks,causing changes in control intensity and ultimately affecting the stability of the controlled system.Therefore,it is necessary to analyze common external attacks and study consensus control problems based on this analysis.Research on external attacks can reduce or even eliminate the adverse effects of attacks,thus ensuring control effectiveness to a large extent.Common external attacks mainly include deception attacks and denial-of-service attacks,with the concealment feature of deception attacks making them more threatening than other attacks.In summary,this paper focuses on the study of consensus problems in multi-agent systems based on event-triggered mechanisms under deception attacks.Based on Lyapunov stability theory,graph theory,and matrix theory,effective event-triggered control protocols are designed to obtain sufficient conditions for achieving consensus in multi-agent systems under attack backgrounds,while proving that the proposed event-triggered rules will not cause the Zeno phenomenon.The main contents of this paper are as follows:Chapter 1 explains the research background and significance of the study on consensus problems in multi-agent systems based on event-triggered mechanisms under deception attacks.It introduces the current domestic and foreign research status of consensus problems in multi-agent systems based on event-triggered mechanisms.Chapter 2 discusses the main theoretical knowledge related to the research content of this paper,and provides some necessary definitions and concepts.Chapter 3 presents the design of an event-triggered control protocol for a class of nonlinear multi-agent systems.Based on Lyapunov function method and inequality scaling skill,several stability criteria for multi-agent system consensus are obtained.At the same time,it is proved that the designed event-triggered function can avoid the Zeno phenomenon.Chapter 4 introduces a class of deception attacks as external interference to the controlled multi-agent system.Based on constraints on attack time and reasonable event-triggered mechanism settings,sufficient conditions for ensuring multi-agent system consensus under deception attacks are studied and obtained.Chapter 5 summarizes the research content of this paper,and provides a brief overview of some of the shortcomings and deficiencies in the research process.Several possible directions for improvement are proposed,which will be implemented in future research work.