多智能体有向网络的加权平均一致性  

Weighted Average-Consensus for Directed Networks of Multi-Agent

作　　者：俞辉[1,2] 蹇继贵[1] 王永骥[3] 

机构地区：[1]三峡大学非线性与复杂系统研究所 [2]430074湖北武汉华中科技大学控制科学与工程系 [3]华中科技大学控制科学与工程系

出　　处：《微计算机信息》2007年第02Z期239-241,共3页Control & Automation

基　　金：国家自然科学基金(60274020);国家自然科学基金国际合作项目(60340420431);湖北省教育厅科研项目(D200613002)

摘　　要：对多智能体有向网络的一致性问题进行了研究。提出了加权平均一致性的新概念,基于一类分散协调控制器使多智能体网络在固定的网络拓扑取得全局渐近加权平均一致性。从理论上证明了有向网络的强连通性及一类称为平衡图的有向图在解决加权平均一致性问题中扮演着关键角色。最后,给出了一个仿真示例,对理论结果进行了验证。The consensus problem in directed networks of multiagent is studied.A novel notion, called weighted average-consensus, is proposed and a class of distributed controller is applied to achieve global asymptotically weighted average-consensus in networks of mLLltiagent with fixed network topology.It is proved theoretically that the strongly connectedness of directed graph and a class of di- rected graphs,called balanced graphs, play a crucial role in solving weighted average-consensus problems. Finally, simulation example is given to validate the theoretical results.

关 键 词：多智能体 协调控制 加权平均一致性 代数图论 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]