基于多智能体技术的车间设备调度模型

Model of shop-equipment scheduling system based on multi-agent

车间调度是实现生产高效率、高柔性的关键.根据车间调度的动态复杂性,提出了一种将多智能体(Multi-Agent)技术应用于车间设备调度的方法.整个调度系统建立于任务Agent,资源Agent和协调Agent基础上,通过招标—投标机制及奖罚制度完成车间调度.在招标-投标过程中,定义了一项重要的综合指标,即招标系数.通过招标系数的评估实现系统的多目标优化调度.仿真实验结果表明,基于多智能体的车间调度模型具有良好的动态调度功能,能够改善调度系统的性能,提高工作效率和机床利用率.

According to the complexity of Shop-Scheduling, this paper gives a scheduling model for shop-scheduling based on Multi-Agent architecture. The scheduling system consists of three kinds of agents, task agent, resource agent and harmonizing agent. The architecture can distribute the task to the best resource agent by the mechanism of inviting public bidding and the mechanism of rewarding and punishing. In public bidding, a most important parameter is introduced. It is called synthesis index, which is made up of several parameters. It can guarantee the system optimization. The emulator result shows that the model can improve performances of scheduling system and advance work efficiency of machining equipments.