Vehicle Routing Problem Of Multiple Pallet Service Centers In Condition Of Pallet Reposition

In view of the problems existing in the pallets pooling practice in China,we find out some developing matters,such as the recycling difficulties,unbalanced demand distribution of multiple service centers,roundabout,the coexistence of empty pallet and repeated transportation,multiple loading and unloading etc.It is necessary to establish advanced pallet pooling system for solving those matters,which is also one of the effective development paths for green logistics engineering.Based on the literature review of the operation practice of pallet pooling system and multi depot vehicle routing problem,the operation decision of pallet pooling system and the decision of vehicle routing problem are integrated and abstractly described as vehicle routing problem of multiple pallet service centers in condition of pallet reposition.The pallet pooling system in this thesis is defined as a network composed of multiple pallet service centers and multiple customer points.Based on the models of mufti depot vehicle routing problem,the pallet breakage rate and pallet redistribution are considered to improve the initial load of the delivery vehicle and the fuzzy time window constraints.In order to make the model closer to the characteristics of the pallet distribution activities in the real-world pallet sharing system,reduce the operating cost of the pallet pooling system and improve customer satisfaction,the pallet operators prefer to minimize operating costs and maximize customer satisfaction as the decision-making goals to establish bi-objective mixed integer programming model of vehicle routing problem of multiple pallet service centers in condition of pallet reposition.Just because the ant colony algorithm has the strong global search ability,positive feedback mechanism,strong robustness and does not depend on the initial solution,the clustering ant colony algorithm and multi-objective ant colony algorithm are integrated to solve the model.Firstly,the clustering ant colony algorithm is designed so that each pallet demand node is assigned to each pallet service center with the Euclidean distance function between the node and the depot as the evaluation index.Then,a multi-objective ant colony algorithm is designed to achieve the goal of minimizing the total cost C and maximizing customer satisfaction u of multi-pallet service center path planning according to customer needs and customer time window requirements.Through the sensitivity analysis of the three scenarios of pallet service center quantity,resource allocation scheme and pallet reposition,outcomes are shown that the established model is effective.Tradeoff curve analysis and performance evaluation index analysis of multi-objective algorithm show that ant colony algorithm has obvious advantages in solving this problem offered in this thesis.