The Research And Application Of Multiobjective Evolutionary Algorithm In The Location Of Logistics Distribution Center

With the rapid development of the social economy and the advent of the ecommerce era,the logistics industry as an important part of the tertiary industry,the improvement of its logistics system and service capabilities has play an important role in developing modern logistics and promoting the upgrading of the industrial structure.The distribution center,as an inevitable product of the systematic and large-scale logistics activities,has logistics system functions such as procurement,storage,allocation,and processing,is a key link in the construction of modern logistics systems.However,there are still many problems in its development,such as weak infrastructure,low level of modernization,imperfect functions,and unreasonable site selection.Among them,the location of the distribution center is not only related to the normal operation of the distribution center and play of the system functions,but also affects the logistics cost and service level of the distribution center.Therefore,optimizing the location of logistics distribution centers is of great theoretical and practical significance to improve the efficiency of logistics distribution and improve the construction of modern logistics service systems.In this paper,by collating relevant literature on the location of distribution centers at home and abroad,it is found that most of the current researches focus on the pursuit of the lowest logistics cost of distribution centers as the main goal of modeling and optimization.The results of location selection often come at the expense of customer service experience.And with the increasing diversification of customer demand and the overall improvement of the industry service level,the contradiction between the distribution center's distribution efficiency and service quality and logistics costs has become increasingly prominent in the issue of distribution center location.In order to solve this problem,this article considers both corporate goals and customer satisfaction,optimizes logistics costs and customer time satisfaction as the optimization goals,and considers practical constraints such as distribution center capacity constraints and customer delivery time constraints to build a multi-objective location of logistics distribution central problem model with constraint penalties.In order to solve this model,this paper further studied the solution algorithm and finds that the decomposition-based multi-objective evolutionary algorithm(MOEA/D)assigns the same neighborhood size to each sub-problem with different search capabilities.Based on it,this paper proposes a decomposition multi-objective evolutionary algorithm based on Individual evolution to dynamically allocation neighborhoods(MOEA/D-SD),first it evaluates the evolution state of the subproblems by the relative improvement between old and new solutions,then dynamically allocate appropriate neighborhood size to subproblems according to the evolution state,and by comparing several popular multiobjective evolutionary algorithms on standard test functions for validation the convergence performance of MOEA/D-SD algorithm is significantly improved,the algorithm resource allocation is more reasonable,and the overall quality of the solution is improved,finally,the algorithm is applied to solve the problem of logistics distribution center location in this paper.The optimization results achieve the double optimization objectives of the distribution center location with the lowest logistics cost and the highest customer time satisfaction,and provide decision makers with multiple pareto-optimal location options.