基于仿真的车间作业计划优化设计  

Optimization of Job-Shop Scheduling Based on Simulation

作　　者：马玉敏[1] 樊留群[2] 张为民[2] 张浩[1] 

机构地区：[1]同济大学CIMS研究中心,上海200092 [2]同济大学中德学院,上海200092

出　　处：《系统仿真学报》2007年第19期4548-4552,共5页Journal of System Simulation

基　　金：中德政府项目(2002DFG00027)

摘　　要：车间作业排序是一个NP-Hard问题,针对其复杂性和难以求解的特点,提出了采用仿真与优化算法相结合的求解方法。首先,建立了车间作业计划仿真优化系统的框架,采用UML对系统进行建模分析,并在仿真软件eM-Plant中映射实现,应用遗传算法优化模型关键参数,实现最优排序,最后示例说明该系统。该方法简单可行,为解决这类复杂问题提供了一条捷径。A job-shop scheduling problem is proved to be a NP-hard problem. Considering its complexity and its difficulty in solving, a solution that combining simulation and optimization algorithms was provided. A framework of job-shop scheduling simulation optimization system was suggested. The system was analyzed using UML mapped in eM-Plant simulation software, and its model parameters were optimized through genetic algorithm （GA） to get the ideal scheduling. An example was given to explain the system, This solution is easy and feasible, and provides a shortcut for this kind of complicated problem.

关 键 词：作业计划排序 仿真 遗传算法 EM-PLANT 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术] F406.2[自动化与计算机技术—计算机科学与技术]