一种基于拍卖与群体智能的作业分配方法  

A job assignment scheme based on auction and swarm intelligence

作　　者：王兴伟[1] 韩琳[1] 黄敏[1] 

机构地区：[1]东北大学信息科学与工程学院,辽宁沈阳110004

出　　处：《华中科技大学学报（自然科学版）》2007年第S2期124-127,共4页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家高技术研究发展计划资助项目(2006AA01Z214);国家自然科学基金资助项目(60673159);高等学校博士学科点专项科研基金资助项目;新世纪优秀人才支持计划资助项目

摘　　要：基于微观经济学与群体智能知识,设计了一种网格系统中的作业分配方法.该方法首先基于改进的英国式拍卖模型确定资源购买者和资源提供者之间的资源交易价格,然后使用混沌搜索混合粒子群优化算法寻找作业分配方案,通过引入适当的适宜值函数,达到了兼顾时限与成本的目的.仿真结果表明,与现有的一些方法相比,该方法是可行和有效的,作业对资源的分配较均衡,效用较高.In this paper,by introducing the knowledge of microeconomics and swarm intelligence,a job assignment method for grid computing is proposed.It determines the resource trading price between the resource buyer and the resource provider based on an improved auction model,and then finds the optimal job assignment solution based on a hybrid chaos search particle swarm optimization algorithm.By introducing an appropriate fitness function,both deadline and cost are considered simultaneously.Simulation results have shown that the proposed method is both feasible and effective with higher utility and much balanced job assignment to resource compared with some other existing schemes.

关 键 词：网格 作业分配 资源定价 拍卖 粒子群优化 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]