战区装备保障点动态选址仿真问题研究  

Research on Dynamic Locating Decision-making Model and Arithmetic of War Zone Equipments Support System

作　　者：董鹏[1,2] 杨超[1] 刘刚[2] 

机构地区：[1]华中科技大学管理学院,湖北武汉430074 [2]海军工程大学管理工程系,湖北武汉430033

出　　处：《计算机仿真》2012年第9期32-36,共5页Computer Simulation

基　　金：国家自然科学基金资助项目(70471042;70601011);教育部新世纪优秀人才支持计划项目(NCET-06-0653);海军工程大学自然科学基金项目(HGDQNJJ038)

摘　　要：传统的选址决策多为静态选址,很少分析动态因素影响。综合考虑需求变化和物价上涨等因素,针对解决战区装备保障点的动态选址优化决策问题,首次提出建立了动态的广义最大覆盖选址模型,采用分析遗传算法、BP神经网络和模拟退火算法的优缺点基础上,设计并提出了一种遗传模拟退火算法与BP的混合算法,并根据算法步骤设计了动态选址实验流程。用实例进行仿真研究,结果表明了改进算法的有效性,同时较好地解决了战区装备保障点的动态选址优化问题。This article studied dynamic locating problem of war zone equipments support system. Considering such factors as demand variety and price rising, we firstly designed a dynamic reciprocal location model. Basing on the an-alyses of virtues and flaws of genetic algorithm, back propagation neural network and annealing algorithm, we provid-ed BP algorithm based on genetic algorithm and simulated annealing algorithm. At last, we designed the dynamic lo- cating problem＇ s simulation flow, and studied an actual example. By comparing with traditional algorithms, we vali-dated its validity. This new algorithm can solve the dynamic locating problem of war zone equipments support system very well.

关 键 词：战区装备保障点 遗传算法 反向传播神经网络 模拟退火算法 动态选址仿真 

分 类 号：F272[经济管理—企业管理]