基于任务能力排序的虚拟企业联盟生成方法  

Task ability-sorting based virtual enterprise coalition generation method

作　　者：杨怡[1] 王洪国[1] 邵增珍[2] 成照乾[2] 

机构地区：[1]山东师范大学管理与经济学院,济南250014 [2]山东师范大学信息科学与工程学院,济南250014

出　　处：《计算机工程与应用》2011年第7期229-232,共4页Computer Engineering and Applications

基　　金：山东省科技攻关计划No.2009GG10001008;济南市高校院所自主创新项目(No.200906001);山东省软科学研究计划项目(No.2009RKA285)~~

摘　　要：将虚拟企业联盟生成问题转化为多Agent问题,针对传统求解多Agent问题过程中仅考虑成本和联盟值最优的局限性,引入"一企多标"和资源配置等实际问题并给出解决方案。重点论述了基于改进的蚁群算法及任务能力排序的虚拟企业联盟生成方法,在求解过程中首先根据任务所需的核心能力对Agent集进行划分,然后利用改进的蚁群算法来求解该问题,以使满足该任务能力要求的Agent优先被选择,从而避免优势资源得不到利用的现象,起到合理配置资源的作用。通过模拟算例分析表明,该方法能有效、快速求解多虚拟企业联盟生成问题,表现了该方法的正确性。The problem of virtual enterprise coalition generation is converted into multi-agent problem in the paper.In view of the limitation of the only considering the cost and the optimal value of the union in solving the multiple agent problems in traditional way,this paper introduces allocation of resources and then puts forward the solution.Focusing on the method of virtual enterprise coalition generation which is based on improved ant colony algorithm and the task ability-sorting,the agents are divided in the process of solving according to the core abilities which the task needs,then this paper adopts improved ant colony algorithm to solve the problem,so that the agent meeting the mission requirements has the priority to be selected,that means the waste of advantage resources can be avoided and the allocation of resources is reasonable.The example through simulation analysis shows that the algorithm can solve multiple agent problems quickly and validly,showing the correctness of the algorithm.

关 键 词：虚拟企业 多AGENT系统 蚁群算法 信息素 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]