多智能体系统中一致性和可控性问题综述.docVIP

多智能体系统中一致性和可控性问题综述 摘要：本文对多智能体系统中一致性和可控性问题的研究现状进行综述。介绍无向、有向、切 换网络通信拓扑和复杂网络通信拓扑下的渐近一致性问题和有限时间一致性问题；给出目前可 控性和结构可控性方向的理论结果；并指出今后的研究方向。 关键词：多智能体系统；复杂网络；一致性问题；可控性 中图分类号： TP13 A Survey of Consensus and Controllability Problems in Multi-Agent Systems CAO Mengtao1, WANG Long1, XIAO Feng2 1 Center for Systems and Control, College of Engineering, Peking University, Beijing 100871 2 School of Automation, Beijing Institute of Technology, Beijing 100081 Abstract: In this paper, a survey of multi-agent systems on consensus problems and controllability problems is provided. For consensus problems, both asymptotical consensus problems and ?nite-time consensus problems with the topologies of undirected, directed and switching networks, and complex networks are investigated; for controllability, theoretical results regarding both controllability and structural controllability are summarized; future research directions are also proposed. Key words: Multi-Agent Systems; Complex Networks; Consensus Problems; Controllability 0 Introduction Multi-agent systems have a long research history, and in recently years, more and more researchers have put their interest in them and lots of results have been published. One can see multi-agent systems everywhere in nature and social networks, for instance, birds migrating in a ?xed formation, lots of ?sh moving together, people clapping consistently [1] and so on. In these phenomena, agents have to communicate with other agents to ?nish a complicated task, -1- which is very hard to be accomplished by a single agent, and also agents are simple and make their decisions by themselves, which means what they are going to do are only determined by the information of themselves and the information they received. Many researchers make their e?ort to understand this kind of collective phenomena. Reynolds explained it by three principles: collision avoidance, velocity matching and ?ock centering[2], which are the basic movement protocols in natur