Multi-objective Vehicle Routing Optimization Research Under Stochastic Uncertainty

With the development of the economy and the popularity of the Internet,logistics and distribution play an important role in promoting social development and improving the quality of life.Using scientific management methods to improve the efficiency of logistics distribution is beneficial to enterprises,customers,and society.The most important part of logistics and distribution is the "last mile".It connects enterprises and customers through logistics services.Logistics enterprises need to use resources in a reasonable way to reduce enterprise costs and improve service efficiency.This paper mainly studies the multi-objective vehicle routing optimization problem in urban logistics distribution,and constructs a stochastic vehicle path model with the goal of minimizing economic cost and time cost,from four steps:problem analysis,model building,algorithm design and solution analysis.Conduct detailed research and in-depth analysis of this issue.Firstly,combined with the development situation and situation of the logistics industry in recent years,it is analyzed that the "last mile" problem affects the improvement of enterprise service efficiency.Therefore,it is very important to put forward appropriate research theories and methods.Reviewing and summarizing the key issues in the "last mile" research field at home and abroad in recent years,it is found that many models are simplified,and multi-objective optimization and uncertainty are ignored.Therefore,this paper proposes a problem of stochastic multi-target vehicle routing optimization based on the vehicle routing problem with time window.Secondly,in the process of model construction,using the characteristics of soft time window,from the perspective of logistics enterprise,taking economic benefit and time benefit as the goal of the model,analyzing the uncertainty information of the distribution link,introducing stochastic travel time and stochastic service time,Using the theory of Monte Carlo algorithm,the stochastic variables are transformed into deterministic variables,and then the stochastic multi-objective vehicle path optimization model is constructed.In terms of solving,this paper adopts a multi-objective optimization algorithm based on decomposition,and based on this,it uses Tabu Search rules to improve the ability of the algorithm in local search.Compared with the standard MOEA/D and NSGAII algorithms,the algorithm converges.Both sex and distribution are superior to the comparison algorithm.Finally,the experiment is designed and the vehicle distribution scheme under different scenarios is calculated.The solution results and distribution scheme of multi-target vehicle routing optimization model are analyzed under the scenarios of no stochastic factors,low stochastic factors and high stochastic factors.Through the analysis of several stochastic scenarios,it is concluded that the time benefit is greatly affected by stochastic factors,while the economic benefits are less affected.In the specific scenario,the different selection strategies of the enterprise are compared and analyzed,that is,the time benefit priority,the economic benefit priority and the balanced benefit priority,and the economic benefit priority distribution scheme is obtained in the vehicle transportation distance,the vehicle load utilization rate,the vehicle waiting time and the delay.It performs better in time and is more competitive than the other.