集成化物流中的定位配给问题的启发式算法

以Baumol Wolfe提出的传统的0 1混合整数规划模型为基础,建立一类更接近物流配送实际的定位 配给模型·在已确定每个配送中心的服务范围内,根据客户群的总需求量接近或等于单车容量的整数倍为原则,提出将不同客户需求量引入最小包络法进行混合法选址的启发式算法·最小包络法即以一个配送中心为基点,分别与其他配送中心用直线连接,作这些直线的垂直平分线,以每个配送中心的内点的垂直平分线的最小包络,作为此配送中心的服务范围·该方法既考虑到不同客户之间的相对距离,又考虑到不同客户需求量,此模型有助于解决实际的定位 配给问题·计算机仿真实例证明了此方法的有效性·

Heuristic Algorithm for Location-Allocation Problem in Integrated Distribution

A model of LA (location-allocation) was set up to approach further to the actuality of goods/service distribution, based on the conventional 0-1 mixed programming model by Baumol-Wolfe. A heuristic algorithm (HA) was thus proposed to combine different customers' demands into a minimum wrap-net for the location of distribution centers in accordance to the assumption that in every defined service area of a distribution center the aggregate demand of a group of customers is equal or near to the integral multiple of unit vehicle capacity. The minimum wrap is formed by taking a distribution center as a basis point and connecting it to other centers with straight lines as distances then making bisectors vertical to these lines and forming a minimum wrap in which all the tangent points of the circles taking relevant half of such distances as radii are enveloped, i.e., a shortest distance is available to a defined group of customers to go to a nearest center as they are within its service area. The algorithm has the advantage of taking account of not only the different distances among customers but also their different distances demands. The model will benefit the solution to actual LA problem. The simulation result by computer as a case study proved its validity.