基于多目标遗传算法的混流加工/装配系统排序问题研究  

A MOGA-based Algorithm for Sequencing a Mixed-model Fabrication/Assembly System

作　　者：王炳刚[1] 饶运清[1] 邵新宇[1] 徐迟[1] 

机构地区：[1]华中科技大学数字化制造装备与技术国家重点实验室,武汉430074

出　　处：《中国机械工程》2009年第12期1434-1438,共5页China Mechanical Engineering

基　　金：国家863高技术研究发展计划资助项目(2007AA04Z186);国家自然科学基金资助项目(50875101);国家重点基础研究发展计划资助项目(2005CB724107)

摘　　要：为解决由一条混流装配线和一条柔性部件加工线组成的拉式生产系统的优化排序问题,以平顺化混流装配线的部件消耗和最小化加工线总的切换时间为优化目标,建立了优化数学模型;提出了一种多目标遗传算法(MOGA)用于求解该优化模型;在该算法中,提出了一种三阶段的实数编码方法用于可行解的表达,同时应用帕累托分级方法和共享函数方法对可行解适应度值进行评价,保证了解的分布性和均匀性。利用遗传算法对两个单目标分别进行优化,结果表明,该多目标遗传算法是可行的和有效的,应用该算法可以获得满意的非支配解集。This paper is concerned about how to optimize the input sequences in a pull production system which is composed of one mixed--model assembly line and one flexible parts fabrication line. Two objectives were considered simuhaneously：minimizing the total variation in parts consumption in the assembly line and minimizing the total setup time in the fabrication line. The mathematical models were presented. A multi-objective genetic algorithm （MOGA） was proposed for solving the model, in which a three-phase real number encoding method was put forward and the Pareto ranking method and the sharing function method were employed to evaluate the individuals＇ fitness, which guaranteed the dispersity and uniformity of the solutions. The feasibility and efficiency of the MOGA are shown by comparisons with a genetic algorithm （GA） over the two single objective respectively. The computational results show that satisfactory results can be obtained by the MOGA.

关 键 词：排序 混流加工/装配系统 多目标遗传算法(MOGA) 遗传算法 

分 类 号：TH16[机械工程—机械制造及自动化] TP39[自动化与计算机技术—计算机应用技术]