政策导向型多agent协同系统研究  

Study on policy-oriented multi-agent cooperation system

作　　者：胡军[1] 李正通[1] 

机构地区：[1]湖南大学计算机与通信学院,长沙410082

出　　处：《计算机应用研究》2010年第8期3009-3012,3015,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(60773208);高等学校博士学科点专项科研基金(20070532075);湖南省普通高校青年骨干教师培养对象项目

摘　　要：为了使多agent协同过程可控、有序的进行,克服传统多agent系统灵活有余而可控制性不足的缺点,提出以动态约束和调整系统行为的政策为管控手段的政策导向型多agent协同系统。首先阐述了政策导向型多a-gent协同系统原理,给出该系统的体系结构以及各个子结构的描述;然后介绍该多agent的原型系统,通过对运行实例的描述,验证了系统的可控制和易管理特性。For the multi-agent cooperation process was controllable,to resolve the problem of flexibility sufficiency but controllability insufficiency in traditional multi-agent system,this paper proposed a policy-oriented multi-agent cooperation system,which took the dynamic restraint and regulate system behavior policy as the management and control method. Firstly, presented the policy-oriented multi-agent cooperation theory; then,introduced the system architecture and the system substructure; finally,presented an example analysis which verifies the controllability and management features of the system.

关 键 词：多AGENT系统 协同 政策 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]