Research On Several Problems Of4PL Based On Graph Clustering

With the rapidly change of Internet, the demand of the production and life changes of peopleconstantly. At the same time, the demand of the logistics services has the deep-seated changes.From the early days of the company’s own logistics and distribution system to more professionalthird party logistics services, logistics service system is undergoing changes in different levels.Today, with the rapid development of e-commerce platform, online shopping becomesfashionable and also becomes an indispensable part for the new generation. Online shopping notonly has brought a tremendous profit on e-commerce platform, but also has brought benefits tobusinesses that provide logistics services. However, the number of logistics enterprises is hugetoday, and electronic business platform is a new way of shopping. Supervision and managementof distribution process is not perfect, and the service levels of the distribution system needs to beimproved.Due to the above content, the fourth party logistics (4PL) comes into our sight. Developed inthe third party logistics, the fourth party logistics is a platform or organization for unifiedmanagement, unified distribution unified dispatching of the logistics demand and logisticsresources. It can effectively organize and direct the logistics distribution, transportation andscheduling. As a platform for unified management and scheduling, it can provide strong supportfor the entire process of the logistics.The current study for the fourth party logistics focuses on the theoretical level, its practicalapplicability need verification. As a core issue in the fourth party logistics research, logisticsalliance-building and logistics center location problem is urgently needed to be resolved. Fromthese two angles, the fourth party logistics still need to solve the following problems:(1) Detailed mathematical model of logistics alliance established needs to be establishedfrom the point of view of practice and application, then design the optimize scheme by means ofthe corresponding algorithm. Logistics alliance building is the key problem of fourth partylogistics, how to formulate efficient optimization scheme is an important aspect of success orfailure of the fourth party logistics construction.(2) Problem of location of logistics distribution center needs to be solved by means ofestablishing a mathematical model to optimize the distribution structures and processes. Problemof location of logistics distribution center is the core issue to improve logistics distribution systemand guarantee the supply of logistics resources. Research and modeling of the problem can lay the foundation for the fourth party logistics.Based on the above analysis, this paper takes realistic applicability and effectiveness as goalof the model, and studies the key issues in the course of the research, the main contributions areas follows:(1) Establish mathematical model of the logistics alliance and proposes an improved graphclustering algorithms for problem solving.Logistics alliance is on the basis of the third party logistics, doing resource integration andharmonization of planning to logistics enterprises or distribution center, in order to achieveeffective management of logistics enterprises, provide a basis support for the fourth party logisticsconstruction and lays the foundation for the problem of the location of logistics distribution center.This paper presents an improved algorithm based on graph clustering algorithm, combined withparticle swarm algorithm to solve the problem. The experiments show that the graph clusteringalgorithm based on particle swarm algorithm can effectively solve the problem of logisticsalliance established.(2) Three synchronous evolutionary models are proposed based on genetic algorithm andhierarchical clustering algorithm to solve the problem of the location of logistics distributioncenter.Logistics distribution center problem is conducted a detailed study by multi-level logisticsdistribution center selection, and is taken an effective integration and distribution of logisticsresources and logistics needs within the region, thus provide support for the construction of thefourth party logistics platform system. Experimental results show that the location of logisticsdistribution center problem based on genetic algorithm and hierarchical graph clusteringalgorithm can be solved with lower cost.In this paper executes the mathematical modeling process and proposes improved graphclustering algorithm combined with particle swarm algorithm and genetic algorithm, providingideas and support for the fourth party logistics issues.