带外包受限批量模型的启发式遗传算法  

Heuristic Genetic Algorithm for Capacitated Lot Sizing Models with Outsourcing

作　　者：钟金宏[1,2] 黄玲[2] 

机构地区：[1]合肥工业大学管理学院,合肥230009 [2]过程优化与智能决策教育部重点实验室,合肥230009

出　　处：《系统仿真学报》2011年第12期2623-2628,共6页Journal of System Simulation

基　　金：国家自然科学基金(71171072);安徽省自然科学基金(090416249);教育部归国人员留学基金

摘　　要：研究了生产和外包数量受限的动态批量问题,引入虚拟生产量实现问题转换。设计了启发式遗传算法:针对01变量的编码方案;每周期虚拟生产量的最优分配方案;修正不可行解的局部启发式平移过程;修正遗传算法最好解的启发式前后向过程。进行了算子组合、交叉变异概率组合和精英策略影响试验;通过大量仿真试验,验证了所提算法的性能。A dynamic Lot Sizing Problem was addressed,in which outsourcing and production levels at each period were bounded.The original problem could be transformed into classical capacitated lot-sizing one by introducing virtual production level.A heuristic genetic algorithm was proposed,which included the encoding scheme only for the setup variables,an optimal assignment scheme of virtual production level at each period,a local heuristic shifting procedure to repair each infeasible individual,and a heuristic forward and backward procedure to modify the best solution by genetic algorithm.After the examinations of operator combination,crossover and mutation probabilities,and elite policy,the performance of the proposed algorithm was validated by a plenty of simulations.

关 键 词：经济批量问题 外包 启发式遗传算法 虚拟生产 

分 类 号：C931[经济管理—管理学] TP18[自动化与计算机技术—控制理论与控制工程]