基于对策论的Multi-Agent多边协商机制研究  

Research on Multi-Agent Negotiating Mechanism Based on Game-theory

作　　者：张孟资[1] 徐国明[1] 

机构地区：[1]陆军军官学院计算中心,安徽合肥230031

出　　处：《计算机技术与发展》2012年第10期51-54,58,共5页Computer Technology and Development

基　　金：军内基金资助项目(pbxy-2011-042)

摘　　要：针对决策支持系统中智能化程度低和协作困难的问题,特别是对多层次、多目标的决策困难问题,提出了基于对策论的Multi-Agent多边协商机制完成决策支持任务;探讨了基于对策论的多边协商机制,给出了Agent协商联盟的构成条件,提出了一个基于对策论的Multi-Agent多边协商机制,设计了基于Multi-Agent多边协商联盟的构成算法。该算法利用Multi-Agent的多边协商解决了多层次、多目标、多属性的竞争决策问题。最后通过算法验证实例证明了协商机制能更好地帮助用户进行决策,提高了决策系统的灵活性和准确性。To resolve the issues of the low level of artificial intelligence and the difficulty of collaboration in the decision support system, especially in multi-level, multi-objective decision-making, put forward the game-theory based Multi-Agent negotiation mechanism to accomplish the decision support task;Investigate multilateral consultative mechanism based on game theory,given Agent-Espace consti- tutes a game-theory based,design construction algorithm based on Multi-Agent multilateral consultative union. The algorithm uses the Multi-Agent multilateral negotiated settlement of the multi-level, multi-objective, multi -attribute competitive decision- making problem, and eventually testifies that by simulation experiment ,the negotiation mechanism can better help users to make decisions.

关 键 词：对策论 多AGENT 多边协商机制 决策支持系统 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]