计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (3): 7-10.DOI: 10.3778/j.issn.1002-8331.2009.03.002

• 博士论坛 • 上一篇    下一篇

基于领域知识的多智能体分层协作规划方法

张晓勇,吴 敏,彭 军   

  1. 中南大学 信息科学与工程学院,长沙 410083
  • 收稿日期:2008-08-21 修回日期:2008-10-31 出版日期:2009-01-21 发布日期:2009-01-21
  • 通讯作者: 张晓勇

Multi-agent layered cooperation planning method with domain knowledge

ZHANG Xiao-yong,WU Min,PENG Jun   

  1. School of Information Science and Engineering,Central South University,Changsha 410083,China
  • Received:2008-08-21 Revised:2008-10-31 Online:2009-01-21 Published:2009-01-21
  • Contact: ZHANG Xiao-yong

摘要: 针对复杂动态RoboCup仿真环境下的多智能体协作问题,提出一种使用领域知识的多智能体分层协作规划方法。在根据系统提供的环境状态信息建立智能体初始行为规划的基础上,通过逐步填入智能体的状态和行为特征属性知识,分析行动实现的效果,对初始规划行动进行调整,为智能体制定协调一致的行为规划序列,并获得相应的协作策略信息,使其在实际对抗比赛中能针对当前环境快速做出决策,实现通信受限的情况下多智能体的协作。

关键词: 多智能体系统, 协作规划, 领域知识, 机器人世界杯

Abstract: To solve the cooperation problem in RoboCup,a multi-agent layered cooperation planning method with domain knowledge is proposed.Agent’s initial behavior plan can be established according to the information about environment.The plan can be adjusted by considering the agent’s domain knowledge and analysis of effect of behavior to obtain a consistent behavior plan sequence that enable agent to make a high-speed decision and realize a MAS cooperation for limited communication.

Key words: Multi-Agent System(MAS), cooperation planning, domain knowledge, RoboCup