改进遗传算法求解面向订单多目标排产问题  

Multi-Objective Scheduling Problem Solution in Make-to-Order Production Via Improved Genetic Algorithm

作　　者：武韶敏 胡晓兵[1] 王江武[1] 徐兴伟[1] 

机构地区：[1]四川大学制造科学与工程学院,四川成都610065

出　　处：《机械设计与制造》2018年第3期102-104,108,共4页Machinery Design & Manufacture

基　　金：国家科技支撑计划(2015BAF02B02)

摘　　要：针对面向订单的多目标排产优化问题,综合考虑订单的延期成本、库存成本和生产单元的负载均衡,构建了面向订单的多目标排产优化的数学模型;针对上述模型设计了一种基于矩阵编码的改进遗传算法,并详细介绍了算法初始种群产生策略、行交叉算子、列交叉算子、行变异算子、列变异算子、适应度函数和选择算子的设计。通过实验对比,验证了提出的算法可以有效地解决面向订单生产的多目标排产优化问题,降低企业生产成本,充分利用生产资源。Aiming at the multi-objective scheduling optimization problem in Make-to-Order(MTO)production,a mathematical optimization model is proposed considering the delivery delaying cost,inventory cost and loading balance of the production units. A matrix coded improved genetic algorithm is presented based on the above mathematical model,where the design of initial population generating strategy,row-based crossover/mutation operator,column-based crossover/mutation operator,fitness function and selection operator is introduced in detail. Finally,experiment comparisons are conducted,and validity of the proposed algorithm to cope with multi-objective scheduling optimization problem in MTO productionis demonstrated,which can help the enterprise to lower the production cost and make full use of production resources.

关 键 词：面向订单生产 订单排产 多目标优化 改进遗传算法 

分 类 号：TH16[机械工程—机械制造及自动化]