Research On Optimization Strategy Of Intra-city Instant Delivery Based On Cluster Analysis

In recent years,the continuous updating and development of Internet methods and the strong rise of the new retail industry have made more and more large-scale instant delivery requirements,and the time to solve large-scale problems using only heuristic algorithms will increase exponentially.This paper adopts a two-stage method of clustering and ant colony algorithm.Firstly,the problem size is reduced and then an efficient optimization algorithm is used,which not only reduces the total distribution cost,but also improves customer satisfaction with time.Therefore,this research has practical significance and necessity.This thesis first introduces the research status of vehicle routing problems at home and abroad,analyzes the mode of instant delivery in the same city and the main problems at the current stage;secondly,based on the characteristics of instant delivery and the operating models of different enterprises,two cluster-based Real-time distribution area division method and improvement strategy of basic ant colony algorithm;Third,the mathematical model of vehicle routing problem with time window is established with the sum of costs incurred during the distribution process as the objective function;Finally,customers,vehicles and Store data,respectively verify the effectiveness of the solution strategy in the two cases with different numbers of stores,and compare and analyze the results of the solution.The experimental results show that(1)using the two-stage method of clustering-improved ant colony algorithm to solve the vehicle routing problem can greatly reduce the total distribution cost;(2)in the case of a single store,compared with the basic ant colony algorithm,an improved ant colony is used The timeout time obtained by the algorithm is 27.5% less and the total distribution cost is 22.9% lower.(3)In the case of multiple stores,compared with the basic ant colony algorithm,the improved ant colony algorithm is 35.7% less and the total distribution cost is lower.17.4% lower.The two-stage method based on clustering-improved ant colony algorithm can effectively reduce the total delivery cost and improve customer satisfaction in the intra-city instant delivery.