考虑装配关系复杂性的多目标装配线平衡优化方法  

Multi-objective assembly line balancing optimization method considering complexity of assembly relationship

作　　者：李爱平[1] 赵亚西 张家骅 刘雪梅[1] LI Aiping;ZHAO Yaxi;ZHANG Jiahua;LIU Xuemei(School of Mechanical Engineering, Tongji University, Shanghai 201804, China)

机构地区：[1]同济大学机械与能源工程学院

出　　处：《计算机集成制造系统》2019年第7期1665-1675,共11页Computer Integrated Manufacturing Systems

基　　金：国家“高档数控机床与基础制造装备”科技重大专项-04专项资助项目(2011ZX04015-022);上海市科委(上海市科技成果转化和产业化)资助项目(15111105500)~~

摘　　要：针对装配关系复杂性影响工人工作效率,进而产生瓶颈工位的问题,提出一种考虑装配关系复杂性的改进型多目标装配线平衡优化方法。建立了一种合理测度作业元素装配关系复杂性的方法,继而定义装配关系复杂性平滑系数这一指标。以生产节拍、平滑系数和装配关系复杂性平滑系数为优化目标,工序优先关系和工作站数为约束,建立了多目标装配线平衡优化方法模型。采用遗传算法作为优化算法,并对遗传算法的交叉环节利用模糊聚类算法进行改进,判断选作交叉的两个个体的近亲概率,在一定概率上防止了近亲交叉,增加了种群繁衍的多样性。通过算例验证了所提平衡优化方法的有效性,为制定考虑装配关系复杂性的平衡方案提供了一种可行方法。To deal with the problems that the complexity of assembly relations affected the work efficiency of workers and leaded to bottlenecks station, an improved multi-objective assembly line balancing optimization method considering the complexity of assembly relations was proposed. A reasonable method for measuring the complexity of assembly relations of job elements was established, and then the smoothness index of assembly relations was defined. By taking the production cycle time, smoothing index and smoothness index of the complexity of assembly relationship as the optimization objectives and the process precedence relation and workstations number as the constraint, the multi-objective assembly line balancing optimization model was established. Genetic algorithm was used as optimization algorithm and the multi objective parallel selection method was adopted to improve the fork crossover rule. Based on the fuzzy C -means algorithm, the probability of belonging to relatives between individuals selected to be crossed was calculated. In certain probability, inbreeding was avoided, and the diversity of population reproduction was increased. Through the numerical example, the effectiveness of the proposed optimization method was verified, which provided a feasible method to formulate the equilibrium scheme considering the complexity of assembly relations.

关 键 词：装配线平衡 装配关系复杂性 信息熵 遗传算法 模糊C-均值算法 多目标优化 

分 类 号：TP605[自动化与计算机技术—控制理论与控制工程] TH166[自动化与计算机技术—控制科学与工程]