Research On Vehicle Routing Problems Models And Algorithms Of Logistics Distribution

With the increasing focus on environmental protection and cost decrease, many enterprises realize that logistics is an important measure to improve the ability of market competition, and advanced logistical theory and logistical technic should be used necessarily in manufacturation and management. As an important approach to realize logistic rationalization, research on vehicle routing problems will help enterprises to reduce logistical cost, improve operation efficiency, and enhance customer satisfaction comprehensively. Since vehicle routing problems tightly connect theory of Operation Research with practice of management, which was named as one of the most successful areas in Operation Research in the past half century. In the classical vehicle routing problems, the relations of vehicle routing problems between forward and reverese logistics is separate, but in many distribution/ redistribution systems, operating the forward and reverse channel separately may result in an unnecessary vehicle utilization, and enterprises or customers hope to operate simulanteously the forward and reverse logistics for reducing cost and protecting environment. We consider the models and algorithms of some vehicle routing problems in distribution-logistics. The vehicle routiong problem with simultaneous delivery and pick-up (VRP-SDP) that effectively integrate the vehicle routing problem between forward logistics and reverse logistics, we present its mathematic models by analyzing its theory and practical context, and construct several metaheuristics algorithms to solve the relative problems. And we consider three special problems about VRP-SDP in uncertainty situation, we also propose corresponding uncertainty models and present relative algorithms for solving these problems, at last the models and algorithms are checked by the numerical experimentations.The main contents and innovations of this thesis are outlined as follows:(1)Research on the vehicle routing problem with simultaneous divery and pick-up based on improved differential evolution algorithmThe existing models about vehicle routing problem can not describle the characteristic of VRP-SDP, we propose the integer programming model about VRP-SDP, and adopt an improved differential evolution(IDE) to solve this problem, the numerical experimentations show the validity about the model and algoriothm.(2) Research on the vehicle routing problem with simultaneous divery and pick-up and time windows based on improved genetic algorithmWe construct a mixed integer programming mathematic model of VRP-SDPTW in detail, it can be transformed into other classical vehicle routing problems by setting different parameters. An improved genetic algorithm (IGA) is proposed for solving this problem, and the numerical experimentations show the validity about model and algoriothm.(3) Research on variable fleet vehicle routing problem with time windows based on compound optimum model particle swarm optimizationThe vehicle routing problem with time windows under uncertain vehicle number is an especial case about VRP-SDPTW. We propose its mixed integer programming mathematical model. We design a compound optimum model particle swarm optimization (COMPSO) algorithm to solve this problem. Numerical experiments are used to compare the performance of the proposed method with genetic algorithm (GA) and assignment algorithm. Experimental results indicate that the COMPSO is better than the other algorithms.(4) Research on the heterogenous fleet vehicle routing problem with stochastic demandWe study the heterogenous fleet vehicle routing problem (HFVRP) with stochastic customer demand. We suppose that the customer's demand can not be divided.We present a prior optimization strategy different from reoptimization, and analyze upper bound,lower bound and asymptotic property of the policies. And based on total expectation cost of prior touer as objective function, we design improved genetic algorithm and differential evolution algorithm to solve the problem, and the numerical experimentations show the validity about model and algoriothm.(5)Research on vehicle routing problem with fuzzy demandsThe classic vehicle routing problem is expanded to the situation that customer's demand is fuzzy.We develop a fuzzy chance constrained programming model based on fuzzy possibility theorem by introducing decisionmaker's preference and the latest theorem of fuzzy mathematics. We propse a mixed genetic algorithm to solve this problem based on fuzzy simulation. And we propose a mixed differential evolution algorithm to solve the fuzzy chance constrained programming model based on fuzzy credibility. Finally, the influence of the decisionmaker's preference on the final objective of the problem is discussed using the method of stochastic simulation, and the rational range of the prefenence number is obtained.