The Model And Algorithm Research Of Logistics Distribution VRP Problem

As social and economic development, logistics industry giving more and more contribution to economic growth and known as"the third profit source"of economic growth, receiving more and more attentions from various sectors in society. Logistics distribution is one of the core aspects of logistics and a scientific and rational logistics distribution can save distribution distance and distribution time, thus saving distribution costs, improving customer's satisfaction at the same time. Therefore, the research on logistics distribution has important practical significance.In this paper, we studied a single logistic distribution center with free-recovery operation VRP problem model and algorithm and tried to explore more simple and efficient algorithm for VRP problem. In order to streamline the delivery mileage and improve distribution efficiency, we make an improvement to the traditional VRP problem model and propose a new model, in which the vehicles can load and unload the cargo in delivery process. In this paper we named this model as"VRP problem with carpool".Compared with traditional VRP solved algorithm, this paper has many innovations. In the third chapter, we use a binary tree programming convenience, the VRP problem is mapped to a particular binary tree and binary tree's simplicity traversal is used for solving VRP problems. In chapter IV, we get the enlightenment from the artificial neural network optimization, the VRP problems is mapped to Boltzmann machine energy function, through the Boltzmann machine learning mechanism to optimize its own terms, finally get the minimum of energy function, that is the optimal solution of the VRP problem. As complex formula reasoning was needed when VRP problems is mapped to Boltzmann machine, in order to optimize the calculation process, this paper proposes a cyberspace Boltzmann machine which using three-dimensional structure, the energy function of such a Boltzmann is more simple and calculation results is also better. In chapter V, as the vehicle's load is gradually decreasing during the process of the distribution, we construct a model that is when two vehicle's load are smaller than one of the vehicle's rated load, then the two vehicles can handling the goods in one vehicle in one cargo demand points and let this vehicle complete the remained distribution, the other vehicle go back to the distribution center. The results show that the introduction of carpool model calculated the best route than the traditional VRP model. At the same time we introduced the concept of carpooling in rigid time window VRP model, the results also show that the introduction of carpool VRP model can still achieve better results.Besides, in the first chapter of this paper, the research background, current research situation, the main research content and the structure of this paper were outlined.The main aim of this chapter is explain the purpose of writing this paper. Chapter II briefly introduces the concept of logistics, distribution center concept and the concept and role of distribution, mainly to introduce the logistics background of this paper. Finally, some of current used VRP problem solving algorithms been introduced, hoping to making a compartment to algorithm proposed in this paper. Chapter VI make a summary of this paper, primarily summarizes the work and innovation of this article, and also pointed out some shortcomings and inadequacies of the algorithm proposed in this paper. Future research direction and research work are put forward in the last.As traditional VRP problem solving algorithms, this paper presents various algorithms has its own unique advantages, but also have their own some of the defects. When solve specific problems in specific situation, we should select the appropriate method.