Research On Location-Routing Problem Of Logistics Distribution In The Electronic Commerce

E-commerce is the product of the rapid development of information and network in the twentieth century.It has changed people's consumption pattern and attracted wide attention from all walks of life because of its rapid iterative evolution.With the integration of the world economy and the rapid development of information technology,e-commerce has become a new mode of business activities.More and more enterprises are transforming into e-commerce,and gradually realize that logistics has become a key factor for the smooth and effective development of e-commerce.If there is no scientific,efficient,reasonable and smooth logistics system,the advantages of e-commerce will be difficult to effectively play,so how to establish an efficient and low-cost logistics system to ensure the efficient operation of e-commerce until landing has become an important issue we need to consider.Distribution network is the basis of logistics system.Vehicle allocation and route selection are the core of logistics distribution.How to balance the contradiction between the speed of logistics distribution and the cost of distribution has become the core issue for E-commerce logistics enterprises to consider.Location-Routing Problem(LRP)solves two interrelated complex decision-making problems of facility location and vehicle dispatch in the strategic layer of an enterprise.Multi-level distribution in the actual logistics distribution network should consider the choice between multiple distribution centers and transfer stations.After we deliberately weaken the construction cost of facilities in the model design,multi-level distribution is actually a disguised multi-level LRP problem.This paper focuses on the study of single-level,two-level and multi-level LRP,which has important practical significance for the planning,design and operation of logistics distribution network in the era of e-commerce and the environment of green supply chain.The emergence of big data solves the problem of determining complex parameters in many models.By using big data,the sales volume of products can be predicted relatively accurately,and the traffic condition in distribution network can be evaluated.Driven by large data,it improves the operational efficiency of all walks of life and the operation efficiency of the whole value chain,and greatly improves the feasibility of the model.In practical application,the demand forecasting of E-commerce for customers usually superimposes the fuzzy demand or elastic demand with the actual customer demand to form a new fuzzy demand,which becomes an important factor to optimize the loading of distribution vehicles,also known as marginal load,but the utility value is very obvious.Large-scale neighborhood search algorithm can be combined with many other intelligent algorithms.By adjusting the destruction factor,recombination factor and local search factor,the algorithm has a good performance of jumping out of local optimization.Meanwhile,local search can also take advantage of other intelligent algorithms.LNS algorithm has great flexibility.The algorithm can obtain excellent performance by dynamically adjusting the combination of optimization factors and the starting time according to the information such as solution characteristics and solution state.The essence of intelligent algorithms is the process of randomly moving and jumping in the neighborhood of solutions according to certain rules,thus generating new solutions.Therefore,there is a certain probability mutation when the optimal solution is found.Parallel algorithms make full use of this feature.In the process of solving each thread,the optimal value of the solution is memorized and the combination characteristics of the solutions are analyzed.The obtained optimal value and solution characteristics are shared and exchanged through mailbox mechanism.The threads cooperate with each other and benefit from each other,thus further optimizing the overall optimization performance.The research and optimization of distribution network system in E-commerce environment is helpful to improve the service quality of E-commerce logistics enterprises,enhance the trust and satisfaction of customers in E-commerce,on the other hand,it is helpful to alleviate the bottleneck of basic logistics distribution in E-commerce environment,and is Conducive to efficient implementation of logistics and distribution.Therefore,it has high research value and practical value.