Research On Real-time Delivery Routing Optimization Based On Customer Classification

Vehicle routing problem(VRP)is one of the important issues in the process of enterprise logistics activities.The distribution service is not only related to the profits of enterprise,it is also related to the number of customers that the enterprise can attract and maintain.With the increase of customer time awareness,the role of distribution in logistics economic activities is also increasing day by day.The research on fast-response and time-efficient vehicle routing problems is even more important.In recent years,instant delivery enterprise,which are mainly characterized by fast punctuality,have sprung up like mushrooms,which has brought convenience to people's lives while satisfying Consumers' demand for timeliness of intra-city distribution.The rapid development of the e-commerce industry has led to an increase of real-time delivery problems.In addition,the distribution industry giants such as Meituan Distribution and Hummingbird Distribution have announced their independence,and the market competition is becoming increasingly fierce.Nowadays,with the accelerated pace of social production and life,consumers' requirements for punctuality of delivery become more and more strict,and punctuality of delivery has become one of the key factors affecting consumers' online shopping decisions.How to improve the on-time and economical efficiency of urban real-time delivery has become a problem faced by real-time delivery enterprises that is urgently needed to solve.In order to meet customer demand and improve their competitiveness,real-time delivery enterprises invest more costs to improve delivery punctuality to retain customers.However,blindly improving the punctuality is likely to result in the loss of the enterprise.Owing to the value of customers is different,that is,the loyalty and profit contribution to the merchants are different.At the same time,it is possible for the real-time delivery enterprises to perform refined operations on customers under the background of big data.Enterprises should use the technology of big data to identify high-quality customers,and invest limited resources on more valuable customers,so that the enterprises can obtain greater benefits with the least cost.In response to this problem,this paper intends to find a distribution strategy for real-time delivery enterprises,so that enterprises can improve delivery punctuality and obtain the maintenance and development of high-quality customers with minimum cost to win more potential benefits and achieve sustainable development.Therefore,the customer classification is incorporated into the problem of real-time distribution routing optimization.Specific research contents include the following aspects:(1)This paper introduces the relevant theories of customer segmentation,data mining,and vehicle routing problems.These theories are used to select customer segmentation variables,determine customer classification research ideas and methods,and construct mathematical models for real-time delivery routing optimization.(2)This paper use data mining methods to construct a classification and prediction model for customers based on their consumption behavior data.First,use the DBSCAN algorithm to perform cluster analysis on customers according to the RFM model,and divide customers into different categories.Based on the results,use Naive Bayes algorithm to build a classification prediction model.Using this model,customers of unknown categories can be classification and discrimination.(3)According to the characteristics of customers at different levels,a delivery timeout penalty function is set for each type of customer.Based on this,a mathematical model for selecting and optimizing the real-time delivery routing is established,and a genetic algorithm that the number of vehicles corresponding and the path of the vehicle encode correspondingly is designed to solve the model.To achieve the goal of improving delivery punctuality with minimum cost,so that the enterprises can obtain the maintenance and development of high-quality customers.(4)The examples demonstrate the scientificalness and rationality of the ideas presented in this paper,in addition,the effectiveness of the models and algorithms constructed in this paper are verified.The experimental results prove that the real-time delivery scheme based on customer classification improves the punctuality and timeliness of delivery,especially in improving the punctuality and effectiveness of high-quality customers.