Research On Distribution Problem Of An E-commerce Platform Based On Multi-vehicle

Nowadays,electronic commerce,as a new business operation mode in our country,is rapidly penetrating into every field of industry by virtue of electronic equipment and network technology.It not only contains the main connotation of shopping,but also includes logistics distribution,aftersales service and other additional services.Its commercial advantages are increasingly prominent,the market share is gradually expanding,showing the strong development potential and unrivalled competitiveness of e-commerce.To achieve the optimal logistics path,low cost,fast logistics distribution service,for enterprises,not only makes customers faster to receive their own purchases,improve customer satisfaction and loyalty,but also makes enterprises more competitive.Therefore,reducing the cost of socialized logistics has become the main goal of e-commerce enterprises to provide efficient,accurate and agile logistics services."Internet +" promotes the rapid development of the car rental industry and expands the market scale.When the enterprise’s own capacity is insufficient,choosing the mode of "rent for purchase" to meet the daily needs of car use can save the funds for car purchase and invest the funds for car purchase into the core business.The resulting income belongs to the indirect income of "rent for purchase",which can improve the return on assets of the enterprise.In addition,the car rental company can provide a variety of alternative models,vehicle maintenance and insurance for the enterprise.Claims and other products and services to save human costs,time costs and so on.With the change of information technology and consumer’s consumption concept,the transaction scale of e-commerce industry is increasing day by day,especially during the promotion festival,the scale of e-commerce logistics service is increasing sharply,which brings enormous distribution pressure to e-commerce enterprises.For enterprise A,how to rationally arrange the route of joint distribution of various types of vehicles,reduce the total cost of distribution,and improve the economic efficiency and service level of enterprises has been a matter of great concern.It is known that enterprise A has a large number of customers and a complex distribution scale.It is relatively difficult to provide direct distribution services to customers.This paper simplifies the problem of multi-vehicle routing with soft time windows by analyzing the current situation of logistics distribution in enterprise A.According to the idea of "clustering first,scheduling later",firstly,the geographical location of customer distribution is mapped to subcells for partition,then,the customers are preliminarily clustered according to the improved maximum and minimum distance algorithm,and finally,the partitioned customers are categorized.The TSP model of single vehicle with soft time window is established for group distribution problem.The Cplex software is used to solve the TSP model by using Yalmip tool of MATLAB,and the distribution roadmap and cost of each category are solved.Aiming at the applicability of Cplex software,in the process of invoking Cplex to solve integer non-linear programming problems such as soft time window and rental cost,this paper proposes a method of linear transformation of soft time window constraints to solve the difficult problem of solving soft time window in Cplex.In the process of solving,it finds out the starting time of distribution service and the total cost of distribution.There is a close relationship between them,and the sensitivity analysis of the two,and then find a suitable distribution plan for company A.In this paper,a maximum and minimum distance clustering algorithm based on grid partitioning is designed by using the idea of "clustering first,scheduling later".The results of calculation and analysis show that the improved algorithm can not only effectively degrade the complexity of solving the problem,but also show the effectiveness of Yalmip toolbox in solving the integer programming model.It also provides a way to solve the multi-vehicle VRP problem with time windows.A reference solution.