Research On Vehicle Routing Optimization Problem Of E-commerce Logistics Terminal Distribution

With the rapid development of Internet technology and the rapid popularization of computer equipment and mobile Internet equipment,the influence and transaction scale of e-commerce continue to expand in China.As an important supporting industry of the e-commerce,the logistics industry has also ushered in unprecedented development opportunities.At present,there are mainly two types of parcels circulating in the terminal distribution link of the logistics industry in China.One is express parcels formed by traditional e-commerce orders,and the other is intra-city O2O parcels formed by emerging intra-city O2O orders.The research of this thesis is focused on the optimization of the terminal distribution routes of these two kinds of parcels.It is hoped that through model optimization and data analysis,a more time-saving and efficient distribution scheme can be provided for the "last mile" distribution of these two kinds of parcels,in order to maximize the economic benefits of the logistics industry in the terminal distribution.The research of this thesis is mainly divided into three parts:In the first part,this thesis takes the terminal delivery route planning of express parcels as the research background,constructs a vehicle route optimization model with minimizing delivery time as the objective function,and proposes an improved tabu search algorithm to solve the model,and then gives the delivery route of all vehicles and verifies the superiority of the improved TS algorithm through the experiments.In the second part,based on the research background of the terminal distribution route planning of O2O parcels in the city,this thesis constructs a vehicle route optimization model with minimizing delivery time as the objective function.And this model has both service time window constraints and celebrations and destinations that appear in pairs.With the idea of "decompose first,solve later",a two-stage solution algorithm is designed: first,the distribution area is divided into seven parts by the K-Means++ algorithm,and solved by an adaptive large neighborhood search algorithm in every divided area.Finally,this thesis gives the distribution routes of all vehicles,and verifies the superiority of the ALNS algorithm through comparative experiments.In the third part,this thesis constructs a terminal distribution route optimization model based on the collaborative delivery mode of express parcels and O2O parcels,and proposes a hybrid meta-heuristic algorithm that combines tabu search and adaptive large neighborhood search algorithms to solve the problem.And this thesis also designed multiple sets of data experiments,the results show that collaborative delivery is more time-saving and faster than decentralized delivery,and the calculation results of the TS algorithm based on the ALNS framework are also better than those of other heuristic algorithms.