Research On Large-scale Logistics Distribution Method Based On Partition Combination Constraints

With the rapid development of the economy and society,the logistics industry has also flourished.The goal of logistics is to meet the needs of consumers with the least cost.The core of logistics distribution is vehicle scheduling,cargo matching and delivery line optimization.In the face of the VRP problem of multiple combined NP problems,constructing a reasonable algorithm can not only meet various constraints such as vehicles,time and types of goods,but also improve the operating efficiency of the enterprise,reduce the cost of vehicle distribution,and realize the scientific logistics.In the real world,the logistics problem is a huge system involving multiple regions,multiple categories and multiple transportation networks.Therefore,large-scale logistics assembly has gradually become a research trend.In the face of large-scale customer distribution and various customer requirements,the instability and long time spent of heuristic algorithms have been unable to solve the problem friendly.In order to meet the low cost of the enterprise and the high requirements of customers and break through the limitations of the original algorithm,this paper proposes a partition combination algorithm(PCA)to solve the problem of vehicle routing with multiple combinations.The partition combination algorithm reduces the difficulty of logistics calculation and cost by dividing the customer area,and can reasonably solve a series of problems such as complex vehicle distribution,long path,and long time-consuming in large-scale customer logistics distribution.In this paper,based on the actual application of PCA in VRP and software simulation,the validity of the model and the feasibility of the algorithm are verified.The main research contents are as follows:(1)According to the study of vehicle-vehicle-kilometer cost model,the impact of different vehicle types and distances on vehicle cost is analyzed.On this basis,it puts forward the combination of regional distribution customers and multi-vehicle vehicles for coordinated loading;at the same time,the administrative area is divided according to the geographical distribution of customers in large-scale logistics distribution.Construct the adjacency relationship of the region,and give the definition of relevant terms involved in the combined delivery of the partition.(2)Three regional combinations are designed based on vehicle cost constraints,covering all the problems of customer combination in actual transportation,which are independent regional combination,semi-independent regional combination and adjacent independent combination.Under the three combination constraints,according to the distance from the area to the warehouse and the amount of goods in the area,a vehicle allocation method of multiple models was designed to ensure that the vehicle was used least when the optimal loading rate was met;at the same time,under different combination constraints,the design Cross-regional vehicle distribution and cross-combined vehicle distribution.(3)Carry out multiple sets of experimental analysis based on the actual data of the "BBG" supermarket chain.The partition combination algorithm uses global distribution of multiple models to achieve cross-regional delivery of vehicles,increase the loading rate of vehicles and reduce delivery mileage.The experimental results are compared with the genetic algorithm.When the number of customers is small and the distribution is aggregated,the mileage of the genetic algorithm is saved by 1% to 6.5%,but the time consumption is 1200 to 2000 times that of the partition combination algorithm;when the number of customers is large or widely distributed,the partition combination algorithm has advantages in time consumption and mileage saving,and the partition combination algorithm saves mileage by about 10% to 35%.