Improved Genetic Algorithm In The Application Of The Distribution Center VRPTW

Logistics is "the third profit source" economic activity in the world,It has been the concern of the countries all over the world, Especially in developed countries, The logistics development has quite mature,Logistics management and logistics technology has been widely used.Logistics distribution vehicle scheduling is the main factors influencing for the logistics cost, Distribution scheme is reasonable or not on logistics cost, efficiency and customer satisfaction has an important influence. How to use scientific methods to improve the efficiency of distribution logistics management, to reduce logistics cost, to improve the quality of enterprise services has been the focus of the present study.First, This paper analyzes the logistics distribution vehicle routing problem research significance and research status at home and abroad,As well as the related concepts of logistics distribution and the vehicle routing problem encountered in some constraints. Then, The methods to solve the vehicle routing problem are introduced, Analyzes the advantages and disadvantages of them. Finally, Select the improved genetic algorithm to solve the vehicle routing problem of the optimization results.The current distribution tasks are generally large scale vehicle routing problem, And customer has a certain amount of time on the goods.So this article uses the improved genetic algorithm to solve moredistribution center with time Windows of the vehicle routing problem.Using the gravity method and problems of multiple distribution centers mid-perpendicular method partition method, It can be divided into a single distribution center to solve; Then the genetic algorithm major improvements:(1) Capacity of the vehicle loads and maximum mileage as a constraint condition of initial solution for the deletion of the in-feasible solution;(2) The roulette wheel and the best individual method as a selection operator;(3) Because in the whole process of the genetic probability of crossover and mutation have different requirements,So choose the adaptive crossover mutation probability. Finally using the MATLAB language to validation for first instance and second instance,The results show that the improved genetic algorithm for solving this optimization problem has greatly improved than before improvement,Not only has a better initial population, But also improve convergence speed than before improvement, First instance increased by 9.33%,Second instance increased by 7.91%.