Research On Pickup And Delivery Problem With Time Windows And LIFO Loading

With the rapid development of economy and improvement of our living standard,the demand for logistics and distribution is increasingly diversified,more and more attention is paid to the quality of service.Therefore,logistics companies are constantly developing new distribution models to meet the needs of customers.In the classic mode of pickup and delivery,logistics companies take less consideration to the Last-in-First-out(LIFO)loading of the goods when they are dispatching vehicles,which will bring additional handling costs and increase the risk of damage to the goods.Therefore,considering the LIFO constraints of goods in the optimization of distribution routes can not only effectively reduce logistics costs and improve service quality,but also bring good social and environmental benefits.In view of this situation,this paper studies the Pickup and Delivery Problem with Time Windows and Last-in-First-out Loading(PDPTWL).The specific work and research results are as follows:1.This paper reviews the existing works on PDPTWL,analyzes the problems in the research results and classifies and summarizes the methods for solving PDPTWL.Through the analysis of the problem,two formal theorems about the LIFO constraint are firstly proposed and proved in this paper.Finally,a strict mathematical model of PDPTWL is constructed.On the basis of this,combined with the existing problems,the solution idea is given.2.This paper designs the overall solution process of the PDPTWL problem.The problem is divided into two stages: initial solution generation stage and optimization stage.We use the improved heuristic algorithm from Solomon to generate the initial solution,while in the optimization stage,we combine the idea of divide and conquer with the advantages of multicore modern CPUs,design a decomposition-recombination framework,and embed an improved tabu search algorithm to optimize the initial solution.3.Simulation experiments were performed on 119 of data instances of the public data set to verify the effectiveness of the solution.Compared with the existing best results,the solution of this paper achieves better results on 86.5% of the data instances,and the computational time of each instance is less than 10% of the similar algorithms on the data instances with more than 200 customers.In addition,the solution is applied to the self-developed logistics big data platform to solve real customer data.The results show that the proposed PDPTWL solution can effectively guide real logistics distribution.