A Research Of Logistics Distribution Vehicle Routing Problem Algorithm

This paper studies vehicle routing problem with stochastic travel timewindows. The problem can be widely applied to vehicle networking, intelligenttransportation, smart city and so on. The study has important practicalsignificance.For vehicle routing problem with stochastic travel time including soft timewindows, a description of the problem and mathematical model are given. Wefirstly propose the distinction between the cost of transportation and service costs,and a hybrid three-phase algorithm to solve this problem is designed. The firstphase uses the improved Solomn insertion algorithm as the initial-solutionalgorithm, the second stage uses tabu search algorithm which is improved by avariety of neighborhood structures and improving operators, the subsequent thirdstage uses optimization algorithms. Testing the algorithms on the internationalstandard tests set and analyzing the results. The results show that: the algorithmcan get a good result within a reasonable time at the instance of the differentnetwork structures and different travel time per unit will directly affect the costof the service. In addition, the model has been successfully created varioussolutions to adpat the company’s preference.For vehicle routing problem with stochastic travel time including hard timewindows, a description of the problem and a corresponding mathematical modelare given. At the same time, we propose a general method for calculating themixing random variables. The method can be used with any random andindependent variable calculations, including addition operation and maximizingoperation, etc. At the same time, the method has great versatility.A simulation system of logistics distribution vehicle routing algorithm issuccessfully designed, which uses MVC application architecture, JAVA languageand realizes the visualization of the vehicle path. This system can be applied tofuture large-scale logistics enterprise systems. It has great application prospects.