A Multi-Agent Coordination Framework Based on Markov Games

摘要：

Based on the analysis of the reinforcement learning andMarkov games, the paper proposes a layered multi-agent coordination framework. Based on the multi-agents interaction of competition and cooperation, thiscoordination framework adopts the zero-sum Markovgame in higher layer to compete with opponent andadopts the team Markov game in lower layer toaccomplish the teams cooperation. This coordinationframework is applied to Robot Soccer. The results of theexperiment illuminate that our proposed method isbetter than the traditional multi-agent learning.

作者: Bo Fan Quan Pan H.Cai Zhang

作者单位: Northwestern Polytechnical University,Xian,P.R.China

会议类型: 国际会议

会议名称: The 8th International Conference on Computer Supported Cooperative Work in Design(第八届计算机支持的协同工作设计国际会议)(CSCWD2004)

会议地点: 厦门

会议语种:英文

页码: 230-233

在线出版日期: 2004-05-26（万方平台首次上网日期，不代表论文的发表时间）