基于遗传算法的集成产销系统参数优化  

Prameter Optimization of Production-Distribution Integrated System Based on Genetic Algorithm

作　　者：王瑛[1] 孙林岩[1] 李天飞[1] 

机构地区：[1]西安交通大学管理学院

出　　处：《中国管理科学》2004年第1期51-56,共6页Chinese Journal of Management Science

基　　金：国家自然科学基金重大项目(59990470-4);国家自然科学基金海外杰出青年基金(B类70028102)

摘　　要：为了解决供应链中生产和分销的目标不同导致的产销系统的矛盾,本文将基于需求拉动的供应链中分销节点、成品库、加工中心、原料库集成,形成一个集成的产销系统。另外通过设计合适的染色体结构和遗传算子,确定了各个子系统的最优参数,这些参数使得集成系统在满足外生需求最低服务水平的前提下整体成本达到极小化。为了保证最优参数的可行性,提出了有效染色体生成算法。对算例的分析表明,服务水平的增加并不是以总成本的增加为代价的,集成产销系统可以在提高服务水平的同时降低总成本。To solve the conflicts resulted from the inconsistent targets of production and distribution,production system and distribution system should be integrated.In this paper,an integrated system is proposed which covers the distribution nodes,the final product inventories,the producing center and the material inventories in a demand-pulled supply chain.Furthermore,the optimal parameters of each subsystem,which make the system total cost be minimum while a given service level of markets demands is met,are determined by a genetic algorithm together with a generation algorithm of valid chromosomes.Analysis of an example concludes that the improvement of service level needs no increase of the total cost,in other words,the service level could be improved while the total cost could be reduced if the production system and the distribution system are integrated.

关 键 词：供应链 产销集成系统 递归方程 遗传算法 参数优化 

分 类 号：C931[经济管理—管理学]