基于个性的MAS默契合作模型

MAS(MultiAgentSystem)技术是解决复杂、动态、分布式智能应用问题的重要技术。个性表现为Agent的行为能力,是MAS系统实现一个基础。合作方式把各个Agent联系在一起,是MAS形成的关键技术。RoboCup仿真比赛提供了一个测试各种MAS理论的平台。本文基于Agent个性提出了一种MAS合作机制——默契,并在RoboCup中进行了仿真实验,实验结果表明该方法比能力对等的合作策略更能体现MAS的自适应性。

Tacit Cooperate Model Based Personality in MAS

The MAS(Multi Agent System)is a key technique to deal with distributed and intelligent applications that are always dynam- ic and complex. Personality which represent the ability of the Agent is the base of the MAS. Cooperation which link each Agent to- gether is the key of the MAS. RoboCup simulation match provide a platform on which every theory of MAS could be tested and vali- dated. In this paper, we present a personal MAS cooperation way named privity, and use a probability way to realize the Agent per- sonality, and apply it in the RoboCup simulation match.