Research On Vehicle Routing Problem Of Pure Electric Vehicles Based On Multiple Distribution Center

With the rapid development of China’s economy and logistics industry,modern logistics has become an important pillar to support the national economy.At the same time,global environmental problems have seriously affected people’s life,and all countries have deeply realized the importance of green development.For the logistics industry,the government began to promote electric vehicles for logistics distribution,and issued preferential policies to encourage and support the development of electric vehicle related industries.For logistics distribution enterprises,with the formation of regional economy and the expansion of enterprise scale,single distribution center is difficult to meet customer needs.Enterprises must consider the problem of multiple distribution centers.At present,with the continuous improvement of electric vehicle related technologies,there are more and more studies on electric vehicle routing problem(EVRP),but there are few studies on EVRP for multiple distribution centers,which need to be further improved and improved.Therefore,the research on multiple depot electric vehicle routing problem with time windows(MDEVRPTW)is of great significance in academic theory and practical application.Based on the existing research,this paper first expounds the research status of multi-distribution center and electric vehicle routing problem at home and abroad,introduces the concept,classification and solution methods of vehicle routing problem,analyzes the characteristics and applications of EV,and then makes a detailed analysis of EVRPTW and MDEVRPTW.Based on the research and analysis of vehicle routing problems and electric vehicles,a single distribution center EVRPTW model is constructed,and AVNS＿SA algorithm is designed to solve the model.AVNS algorithm has strong local search ability,and SA algorithm uses Metropolis criterion to jump out of local optimum.The AVNS＿SA algorithm combining the advantages of both has stronger global optimization performance.Based on the research of EVRPTW of single distribution center,combined with the multi-distribution center problem,the MDEVRPTW model with time window constraints and power constraints was constructed,and a two-stage algorithm was designed to solve it.In the first stage,k-means algorithm was used to assign customers to each distribution center,and the multi-distribution center problem was decomposed into multiple single-distribution center problems.In stage two,AVNS＿SA hybrid algorithm was used to solve EVRPTW of single distribution center.Based on the above research problems,the feasibility of the relevant model and algorithm was verified by horizontal experiments with EVRPTW examples.Then,the longitudinal experiments were conducted to compare the solution results of other algorithms,which proved that AVNS＿SA algorithm has stronger optimization ability.Finally,an example of MDEVRPTW is designed to verify the effectiveness of solving the MDEVRPTW model using the two-stage algorithm.