Study On Location Issues Of Logistics Distribution Centers Bases On RBF Neural Network

At the end of the 20th century, with the development of science and enhancement of the trend of economic globalization, logistics modernization is developing rapidly. Modern logistics is generally paid close attention to as "the third profit source", and becomes the new point of growth of modern economy. During the process of moving towards the economic globalization and information, our country make developing logistics economy as the important insurance factor of accelerating economic development, adjusting the industrial structure, improving quality and benefit of economic running.Distribution which contacts with the customers is an important part of the logistics, which is developing along with the logistics development. In terms of the logistics, distribution almost includes function elements of the logistics. It is the reduction of the logistics function. Therefore, the coming and development of the distribution is the necessary outcome of the development of the logistics. Distribution center is a crucial part of the distribution. Whether the distribution center location is good or not, it directly connects with the efficiency of the distribution, the logistic cost and the contentment of the customers and affects the whole business acting.This paper illustrates the basic problem: the logistics distribution centers, principles and influence factors of distribution center, and analyses the existed method of distribution center location selection. Distribution centers' site selection, involving numerous elements, is crucial to logistics planning. Traditional algorithm neglects some important qualitative considerations. In this paper, scholars in front of the application of AHP-Fuzzy comprehensive evaluation method and BP neural network site distribution center, based on the method to do the optimization and improvement. This paper proposes a new site selection method, using RBF artificial neural network, where the data derived from a fuzzy appraisal matrix is processed. The new method can make the site selection in an objective manner and avoid the bias brought up by traditional algorithm. And practical examples are cited in the paper to illustrate the process.