Leader-following System Cluster Behavior And Control

In this paper, the new viewpoints and methods are used to the multi-agent system based on the Leader-Following control. It focus on the impact about the communication topology among the agents and the different terms of the model to the consensus of the multi-agent system.In the first place, a new model about the multi-agent system beasd on the Leader-Following control is proposed. According to Lyapunov method the sufficient conditions which make the system reach to consensus are given, and the topology among the agents is generalized to the balance topology. Secondly, the essence of the topology is discussed. In order to generalize the directed topology, simplification of the system is made. The Laplace transform is posed to analyse the problem, which can make the conditions simplify and close to the reality. It is get that if the Leader agent is globally reachable node in the topology of the system, all the agents of the system can reach consensus and make a cluster following the Leader. Moreover, the basic model called Vicsek model is discussed. According to Laplace transform, the topology can be extended to the directed topology, and the consensus center of the systemcan be solved. Finally, the time delay problem of the multi-agent system is considered. Not only the accurate bound of time delay that make the system reach to consensus is given, but also the expression of the consensus center of the system is obtained.During the discussion, the improvements and the extensions of the models are considered. A series of new models are proposed, and the theoretical analysis are given. Moreover, simulations are given to illustrate the theoretical results. The simulations provide verification for the multi-agent systems mentioned in this paper from different aspects. Finally, contribution and prospect of the multi-agent systems are summarized.