Investigacion De Operaciones

2008 International Symposiums on Information Processing

Grey Relation Degree Analysis for the Facility’s Location of Logistics Distribution Network

Bao Zhen-qiang, Wang Peng, Yang Fang, Zhu Cong-wei, Guo Lei College of information Engine, Yangzhou University, Yangzhou, 225009, China yzbzq@163.com

Abstract
For the facility’s location of Logistics Distribution Network (LDN), we analyze acase that Wal-Mart chooses the logistics distribution centers in China. According to the candidate combination schemes and theirs criterions of the case, we retain the experts using Delphi method to get the weight distribution vectors of these schemes. Because analytic hierarchy process (AHP) method only uses the single criterion to compare but not takes the criterion relation into account. Thecriterions in facility’s location of Logistics Distribution network are not independence, the relation of the criterions are not clear but indeed. So, when we consider the location, we use the grey relation degree (GRD) method to compensate for the lack of AHP. And then we combine the GRD analysis to get the judgment matrix. Lastly, according to the weight distribution vector, we can get theintegrative judgment result vector of the schemes and choose one with the maximum value as the optimal scheme.

1. Introduction
The design of Logistics Distribution Network (LDN) is the stratagem decision-making in Logistics Distribution System’s optimizing, which includes the facilities location and the circulation between products in process and finished products. This field has been the hotspot inthe world. Because the location of the distribution centers determines the pattern, the frame and the form of the distribution system, and constrains the usable methods and the correlation cost in the overall distribution system, we can consider the location of the distribution centers as the design of LDN through some papers. The field on facility’s location is getting more attention,

and ithas some traditional methods as follows: Integer Programming, Graphical Method, Baumol-Wolfe, simulation calculation method [1] and so on. And now mainly has two ways: (1)According to different angles, we can improve the existing model by different factors’ impact: In the frame about the location of multi-varieties goods and multi-facilities, Syam Siddhartha and others [2] import the logisticscost, such as inventory costs, indention costs and transport costs, to develop the traditional model of location; On the basis of fully considering consumers’ choice, Sun Hui-jun[3] puts forwards the model to get maximum profit and market’s share in the circumstance of existing logistics distribution centers competing with each other. According to consider the inventory costs of first leveldistribution center and the transport costs of second level distribution, Tao Yin[4] of Shanghai Jiao Tong University adopt math modeling to settle location of first and second level distribution centers in the circumstance. Studying on the relationship of inventory and orientation distribution, Zhang Chang-Xing [5] put the order costs, inventory costs, out-of-stock costs of distribution centers into themodel, at the same time of solving the problem on orientation distribution, he confirms the order place and order quantity of distribution centers in the order strategy with Genetic Arithmetic. Tan Ling and others study [6] the problem on location of distribution centers in chain retail enterprises when the demand is not unsure, he also particularly analyzes the inventory system. And according tosolving the total annual cost model of distribution system, he gets the optimal inventory level and minimum costs in each scheme, and analyzes the impact of the total system costs when the demand and transport distance are changed. (2)Importing new analysis method, we can constitute new model:

This paper is supported by the National Natural Science Foundation of China No.70471073, and Natural...