Solving The Vehicle Routing Problem Based On The Customers' Satisfaction

With the globalization of the economy, modern logistics has expanded from providing the society traditional transportation service to the integrated logistic system with the support of modern science, information technology and management. Over the past four decades, the problem concerning the distribution of goods between depots and final users have received extensive attention and researches have made enormous progress in developing theory, models and solution methods for these problems.With the development of logistics, customers' satisfaction has more important in the process of physical distribution and it has become one of the important ways to get the competitive advantages among logistics business enterprise. Some recent literatures set about research on customers' satisfaction objective from the standpoint of customers but they only regard meeting customers' expectation as constraints of the models in the literatures, which belong to the category of a single-objective decision-making. And many dissatisfactory items for the models await amelioration and modification.Above all, the paper studies the vehicle routing problem based on the satisfaction of customer. The main contents of this dissertation are as follow.In chapter 1: The paper reviews the background and the basic concept including logistics, routing. Subsequently based on summarizing relevant references, the paper reviews domestic and foreign researches on the vehicle routing problem, points out shortcomings of research on this problem and finds some potential areas of research. At last, the paper explains the necessity and realistic significance of solving VRP by introducing the importance of distribution in the whole physical logistic and the status quo in the home distribution industry.In chapter 2: The traveling salesman problem (TSP) is a problem in which a traveler must traverse a set of completely connected cities using the minimum length cycle. This problem is at the root of the vehicle routing problem (VRP). In this chapter, formal and mathematical definitions, together with a brief overview about genetic algorithm and detailed description of solution approaches. An experiment is also performed in order to improve the existing heuristic.In chapter 3, the paper analyses the theoretic part of the whole dissertation. At first, the paper makes formal definitions with the realistic vehicle routing problem, togetherwith detail descriptions about customers' satisfaction. Then the essential meanings of the CS in the VRP are given. According to it, the solving strategy for the problem is given.In chapter 4, the paper studies the problem from the point of view of models and algorithms. The paper studies three sub-parts according to the realistic VRP: For the deterministic demands, this research attempts to formulate a mathematical model for the Vehicle Routing Problem with time window and propose a two-phase algorithm and an improving Genetic Algorithm (GA) to solve the problem. And by analyzing the running-time, the solving quality, the stability, the expansibility and the truly experiments, the paper proves the feasibility of the approach;For the stochastic demands, the dynamic programming algorithm is given by taking the algorithms in two parts according to the different scales of the customers. Then the mathematic approach is given to approve the feasibility of the dynamic programming. Because some of the stochastic demands can be translated into the deterministic demands, the paper builds up the improving GM (1, 1) based on the traditional grey theory and seasonal index and regression model which are approved by data analysis.In chapter 5: The paper makes the new approach into truth by analyzing the three parts including the routing parts and the forecasting demands and the real-time controlling, which are integrated by the human-together interaction mode.Finally, Chapter 6 presents summary, conclusions and recommendation for future research prospective.