Research On The Arithmetic Improvement Of Locating Logistics Centers

At present, modern logistics has become a new economic hot spot in foreign countries. Logistics is the third source of profits to the enterprise. Along with the development of China's economy, the logistics industry in China is in its period of great development.The development of the logistics industry will inevitably lead to the expansion of the logistics system in space. As logistics network nodes, logistics centers will be in a great substantial increase, which will bring a lot of problems of their locating. Logistics centers'location determines the overall efficiency of the logistics system to a large extent. Under the situation, much research wok has been done to the methods of locating logistics center, but many models, including the most popular LA model, lack of effective algorithms to solve, resulting in significantly lower utility.The work I have done in the paper is as follows:First, various elements of the theory of locating the logistics centers are narrated in this paper at the beginning. Based on that, a practical problem of locating the logistics centers is attributed to a theoretical LA model. Then the paper sums up the AlA algorithm's disadvantages-the result heavily depends on the initial solution and it can not automatically look for optimal solution, and proposes an improved AlA algorithm. According to SOFM (self-organizing feature map) neural network features, this paper puts forwards a wholly new algorithm to the LA model- SOFM neural network combined with bar centric method.Second, the practical problem of in this paper is solved by the two new algorithms proposed by myself, which is also a test of the effectiveness of the new algorithms. Through comparison and analysis of the calculation results, SOFM neural network combined with bar centric method may not be able to find optimal solution, but demand centers clusters gotten by it are nearly optimal classification; the result of improved AlA algorithm still partly depend on the initial solution. According to the findings above, this paper presented a wholly new algorithm -- improved AlA algorithm taking the result of SOFM neural...