Multi-depot Vehicle Routing Problem Of Electric Vehicles

With the upgrade of automobile industry and the launch of restrictions on urban traffic roads,fuel vehicles are no longer suitable for urban distribution,prompting distribution enterprises to introduce new energy electric vehicles and related facilities to respond to the government's urban traffic management policies,ensure the efficiency of distribution,and enhance the competitiveness of enterprises' core business.In recent years,electric vehicle routing problem(EVRP)has attracted much attention.As electric vehicles can extend the limited driving mileage by charging,EVRP focuses on analyzing the charging characteristics of EVRP to consider the impact of mileage constraints on the planning results.Only one distribution center is considered to distribute to customers in EVRP research.In actual urban distribution,there are often multiple distribution centers,which need to determine the corresponding distribution centers of customer distribution services.One distribution center is is considered in recent EVRP research.However,there are multiple distribution centers in actual urban distribution.As more and more research of EVRP is being done,the EVRP research for multi-depot needs to be further improved.To enrich existing research and provide theoretical basis for enterprise path optimization,based on the analysis of the characteristics of electric vehicles,a model and the two-stage hybrid algorithm has been proposed to make a profound study on the multiple depot electric vehicle routing problem with time windows(MDEVRPTW).The market scale and policy environment of electric vehicles at home and abroad has been described at first.After the charging mode of electric vehicles,the construction status of charging station and the characteristics of urban distribution has been studied,the application scenario of electric vehicles in commercial express delivery and urban distribution terminal distribution is depicted.In the framework of classical multi-depot vehicle routing model,with a lowest total distribution cost including fast charging costs during distribution,night charging costs in distribution center,fixed vehicle cost,driving cost and penalty cost as a goal of optimization,a MDEVRPTW model with time window and energy constraints is constructed.In the first stage,K-means algorithm is used to assign customers to each distribution center.In the second stage,constrained by load and residual electricity and considering charging characteristics,ant searching strategies in three types of nodes are formulated respectively,and the judgement conditions of ant completing all customer paths are improved.The validity of the model and the algorithm is verified by a design example and compared with the results of other algorithms.The distribution data of Jingdong Logistics has been introduced as a case study.Vehicle route planning scheme for introducing electric vehicles has been obtained,using the model and algorithm of this paper,which verified the practical feasibility of the model and algorithm.