基于Q学习的供应链分销系统最优订货策略研究  

Optimum Order Strategies for Distribution System of Supply Chain Based on Q-learning

作　　者：李随成[1] 尹洪英[2] 

机构地区：[1]西北工业大学管理学院,西安710072 [2]西安理工大学管理学院,西安710048

出　　处：《控制与决策》2005年第12期1404-1407,共4页Control and Decision

基　　金：陕西省自然科学基金项目(04JK263)

摘　　要：研究由一个制造商和多个分销商组成的分销系统的最优订货策略问题.在外部顾客需求不断变化的情况下,以不断提高分销系统双方合作绩效为目标,基于Q学习算法来确定每个分销商的最优订货批量.实例结果表明,在外部需求不断变化的条件下,该算法能简便地解决供应链企业分销系统合作中的最优订货批量问题.The optimum order strategies for the distribution system consisting of a single manufacturer and multidistributors in the supply chain management is discussed. In the case of the varying demand of customers, and with an aim at improving the cooperative performance of the distribution system, the optimum order batch of each distributor is determined based on the Q -learning algorithm. An example shows that this algorithm can simply solve the problem of the optimum order batch in the distribution system in the case of external ongoing changes in demand.

关 键 词：供应链管理 分销系统 Q学习算法 最优订货批量 

分 类 号：F273[经济管理—企业管理]