基于多Agent的虚拟企业任务调度模型及优化  

The multi-agent-based model for task scheduling in virtual enterprise and its optimization

作　　者：赵强[1,2] 肖人彬[1] 

机构地区：[1]华中科技大学系统工程研究所,湖北武汉430074 [2]武汉科技大学机械学院,湖北武汉430081

出　　处：《控制理论与应用》2009年第4期459-462,共4页Control Theory & Applications

基　　金：国家自然科学基金资助项目(60474077);教育部新世纪人才支持计划项目(05653)

摘　　要：采用多智能体技术构建虚拟企业任务调度模型,并对基于该模型的任务调度运作过程进行说明.针对调度优化问题,以资源智能体承担的生产任务为研究对象,综合考虑生产任务之间的时序逻辑关系、作业时间及资源自身已确定的生产任务等影响因素,建立以生产延续时间最小为目标的优化模型,给出粒子群优化求解算法.应用实例及数字仿真验证了模型及优化算法有效性。By means of multi-agent technology, the task-scheduling model in a virtual enterprise （VE） is presented, and the operation process of task-scheduling in VE based on the above model is explained. In the optimization of task-scheduling, the production tasks of each resource agent are taken as the research objective; and the model for taskscheduling in VE for minimizing the production duration is established by comprehensively considering the logical relation among subtasks, the operation time of each subtask, the assigned production activities of each resource agent, etc. The particle swarm optimization （PSO） algorithm is employed for the problem-solving. Finally, a real example and the numerical simulation show that this task-scheduling model in VE and the optimization algorithm are effective.

关 键 词：虚拟企业 任务调度 多智能体系统 粒子群优化 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]