基于混合差分进化算法的联合补货-配送优化模型  

Integrated Joint Replenishment and Distribution Model Using Hybrid Differential Evolution Algorithm

作　　者：王林[1] 顿彩霞[1] 张金隆[1] 

机构地区：[1]华中科技大学管理学院,湖北武汉430074

出　　处：《中国管理科学》2012年第3期86-93,共8页Chinese Journal of Management Science

基　　金：国家自然科学基金资助项目(70801030;71131004);教育部人文社会科学研究青年基金项目(11YJC630275);湖北省教育厅科研重点资助项目(D20112201);中央高校基本科研业务费资助项目(HUST:2012TS065)

摘　　要：本文综合考虑联合补货与配送决策,研究了随机需求、允许缺货环境下多企业多产品联合补货与配送集成优化模型,设计了混合差分进化算法(Hybrid Differential Evolution,HDE)对该模型进行求解,同时通过算例与遗传算法、标准的DE算法进行了比较,证实HDE算法高效且稳定;另外,设计了一个先补货再配送的两阶段优化模型,对比优化结果发现采用供应链协同时补货成本较高,配送成本较低,且总成本较低。最后,对相关参数进行了敏感性分析,发现需求率和库存维持成本的变动对总成本的影响远远大过次要订货成本对总成本的影响。Considering joint replenishment and distribution decision when stock out is allowed, a model in- tegrating multi-buyer joint replenishment and distribution with stochastic demand is analyzed. A hybrid differential evolution （HDE） algorithm is designed to solve the proposed model. An example shows the performance of the HDE is more stable and efficient compared with genetic algorithm and typical differenti- al evolution algorithm. In addition, a two-stage model with replenishment and delivery uncoordinated is designed. The numerical experiments illustrate the fact that the replenishment cost is higher, distribution cost is lower and total cost is lower under supply chain collaboration. Finally, the sensitivity analysis on relevant parameters is addressed and the results show the influence of variations in demand rate and inven- tory holding cost on the total cost is far greater than that of variations in minor ordering cost on the total COSt.

关 键 词：联合补货-配送 混合差分进化算法 协同 敏感性分析 

分 类 号：F273[经济管理—企业管理]