针对MES的生产管理的特征，利用Agent自主性、智能性等特点，提出基于多Agent的MES 任务调度模型，介绍该模型的结构以及各部分Agent 的功能和特点。并使用遗传算法对参与调度的多Agent的协商机制进行优化以提高任务调度中多Agent 的协商效率。关键词: MES；多Agent；任务调度；遗传算法Abstract: In view of the production management characteristic in MES, this paper resents a task scheduling system model based on multi-Agent by means of utilizing ndependent and intelligent of Agent. The paper introduces the structure of model nd every Agent’s function and trait. Moreover, the model also uses the genetic lgorithm to optimize the negotiation mechanism of multi-Agent attending the task cheduling and enhance the efficiency of the task scheduling.Key Words: MES；multi-Agent；task scheduling；genetic algorithm