基于可重生PSO的双层产品组合决策问题  

Two-level product mix decision problem based on regenerated particle swarm optimization

作　　者：郑永前[1] 喻赛君[1] 

机构地区：[1]同济大学机械工程学院,上海200092

出　　处：《计算机集成制造系统》2013年第4期782-789,共8页Computer Integrated Manufacturing Systems

基　　金：上海市自然科学基金资助项目(10ZR1431700)~~

摘　　要：针对实际生产中有中间品生产的产品组合所遇到的机器超载严重、原材料限制等问题,建立了一个集成的双层的产品组合决策模型并进行优化。模型的目标是企业所获利润最大化,综合考虑了订单需求、机器能力及产品与中间品组合关系等约束,运用双层嵌套粒子群算法求解模型,并通过改进的可重生粒子群算法提高解的质量。以企业实际生产数据为例,通过与遗传算法、基本粒子群算法的对比,证明了算法的可行性和优越性。An integrated two-level Product Mix Decision （PMD） model was established for the practical manufacturing process with regard to the machine load problem and material limitations. The objective of the model was to maximize enterprise＇s profit by considering constraints such as order requirement, machine capacity and the relationships between in-process goods and finished products. To solve the PMD model, a revised multi-layer embedded particle swarm optimization approach was presented and a regenerated Particle Swarm Optimization （PSO） was pro- posed to improve the quality of the algorithm in real operation. The feasibility and superiority were proved by comparing with genetic algorithm and PSO algorithm through practical production data.

关 键 词：产品组合决策 双层 可重生 粒子群算法 

分 类 号：TP18[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]