基于自适应PSO和类别分解的多任务串行联盟生成  

Multi-task coalition formation serially based on adaptive PSO and category division

作　　者：许金友[1] 李文立[1] 

机构地区：[1]大连理工大学系统工程研究所,辽宁大连116023

出　　处：《计算机应用研究》2009年第4期1338-1341,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(70572099);辽宁省自然科学基金资助项目(1050349)

摘　　要：现有的联盟生成方案多针对一个agent只能加入一个联盟,不利于联盟总效用的最大化以及联盟中agent能力的充分利用。提出了基于能力类别的agent分解策略,通过定义子agent使得agent可以同时加入多个联盟,在此基础上设计了基于二维离散粒子群的多任务串行联盟生成算法,并对粒子的惯性权重进行动态自适应调整;最后通过算例验证了该方法的有效性。Existing researches concentrated mainly on each agent could only participate in one coalition, which goes against optimization of coalition utility and making full use of agents＇ ability. Proposed an agent division strategy according to the abili- ty category and defined subagent to ensure one agent had the possibility of joining in several different coalitions simultaneously. Then designed a two-dimension particle swarm optimization algorithm （PSO） with adaptive dynamic weight value to solve the serial multi-task coalition formation. Finally, proved the effectiveness of the proposed algorithm by the experiment.

关 键 词：多代理系统 多任务联盟 交叉联盟 子代理 自适应粒子群优化算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]