Research On VRP Problem Based On GIS

In the development of mobile internet era,3S technology with GIS as the core is applied more and more widely in logistics distribution, especially in the choice of vehicle routing problem. Practical studies have shown that it can significantly and effectively promotethe time efficiency and space in the logistics through the further study of vehicle routing choice and optimization problem (VRP).This paper, through comparison and analysis of the studies on application of GIS at home and abroad about the VRP problems and related solving VRP algorithm, deeply study the VRP problem space, time and the characteristic of network under the environment of GIS space. Based on GIS-T technology and under the guidance of new technology, new environment, new ideas, It explains a road network model with obvious characteristics of GIS, which is established by the road network structure and the traffic impedance analysis, further based on network of VRP problems in GIS environment constraints, the analysis of the cost of transportation. Finally, it has improved the reasonableness of VRP model, combined with flow distribution of the thought and method.The scale of VRP is effectively defined through the related theory and methods of GIS, further optimized by experimental methods. With obvious spatial characteristics of VRP algorithm is improved and optimized by the use of combination algorithm ideas.In this paper, distribution areas are classified to multilevel under the guidance of the thinking of combination, dynamic optimization and according to different regions, it uses appropriate methods to solve the problem of VRP. First of all, it extracts the secondary distribution center through the cluster analysis of all the clusters of customers and combined with GIS spatial analysis method. The experimental data show that it can promote the basic optimization rate over 30% and meet the requirements of the time complexityat the time of the secondary distribution center distribution by using Hopfield neural network for vehicle routing selection and optimization. Small range to each secondary distribution center's distribution problem can aggregate and bind by the improved Floyd algorithm for client path in the region to meet certain conditions, and combining Taigu County, Jinzhong City, Shanxi Province, the empirical analysis. The results show that alone this index of route length optimization rate reached 19%, but also to meet certain constraints.