Cooperative Behavior Analysis Of Multi Agent System Based On Cooperation-competition Network

With the advent of the era of artificial intelligence and big data,the idea of distributed control has developed rapidly in the fields of control science and computer,which has caused extensive discussions among experts and scholars.Compared with centralized control,distributed control has the advantages of strong dispersion and high stability.In the framework of cooperation-competition network,this paper discusses the distributed stabilization of the systems with switching topologies and the cooperative behavior of heterogeneous uncertain multi-agent systems with structural balance and unbalance.The sufficient conditions for realizing the corresponding cluster behavior are obtained.The work is as follows:In Chapter 1,from the perspective of system science and control,the research background and current research status of cooperative behavior of multi-agent systems are introduced.In Chapter 2,we introduce the knowledge of algebraic graph theory and the corresponding tools for judging stability,such as the theories of the laplacian matrix,the barbart's lemma,etc.In Chapter 3,under the premise of structural balance,we discuss the distributed stabilization of multi-agent systems in cooperation-competition networks with switching topologies.The overall network is switched in the non-communication and cooperation-competition subsystems.It is proved that the non-communication subsystem can be regarded as the stable subsystem in the switching system,meanwhile,the cooperation-competition subsystem is the unstable subsystem.The closed-loop systems are obtained sufficient conditions of distributed stabilization by designing the switching law.On this basis,in order to simplify the design process of the switching law,a periodic switching law is proposed.Finally,the validity of the theoretical results is verified by numerical simulations.In Chapter 4,under the premise that the network is undirected,for a class of heterogeneous uncertain agents,the corresponding swarming behavior for structural balanced and unbalanced situations is discussed.Through the distributed Lyapunov function redesign approach,the corresponding distributed controller and the estimation laws of unknown parameters are obtained.Under their joint action,the heterogeneous uncertain multi-agent system can achieve distributed stabilization for structurally unbalanced network,and output bipartite consensus for structurally balanced networks.Finally,the validity of the theoretical results is verified by numerical simulations.In Chapter 5,summarize the full text,and point out the directions can be research in the future,namely the cluster behavior under the high-order individual model and the corresponding cluster behavior in the network with multiple competitive relationships.