基于个性的MAS默契合作模型

Tacit Cooperate Model Based Personality in MAS

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。

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