Font Size: a A A

Hybrid Neural Network Applications In The Logistics System

Posted on:2007-12-21Degree:MasterType:Thesis
Country:ChinaCandidate:X S JiaFull Text:PDF
GTID:2208360182493804Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the development of the society, the importance of logistics gradually emerges. Distribution is an operation linking with consumer directly. As the core of logistics system, distribute routing problem has become the focus of many scholars in order to slash costs down and increase benefits.The background of this article is an intelligent distribution system project which is based on GIS .As an essential question of the intelligent distribution system, route optimization has many problem-solving models, the most typical one is Traveling Salesman Problem (short for TSP). The solution of TSP has lots of algorithms, this paper mainly studies Simulated Annealing Algorithm and Neural Network Algorithm. Although SA and Hopfield can be used to solve route optimization problem, but many problems emerge when they are used in practice. Combining the superiority of Hopfield network with the superiority of Simulated Annealing Algorithm, we can use the mechanism of SA which can receive non-prepreerence solutions to overcome the disadvantage of Hopfield which may get into the local minima, At the same time we improved the traditional SA and Hopfield network algorithm. Finally we can obtain an effective, convergent and heuristic algorithm—Hybrid Neural Network algorithm(short for HNN), which is an expeditious and effective algorithm.Using the algorithm into practice can reduce costs and increase benefits, which isproved to be very valuable in practice.
Keywords/Search Tags:Simulate Annealing, Hopfield Neural Network, Logistics
PDF Full Text Request
Related items