Study On Matching Between Traffic Requirement And Capacity Out Of The Distribution Center

After China entered the WTO, the domestic logistics industry is in a fast-growing trend. It provides the fundamental security for the development of the national economy, but also brings large cities increasing traffic problem which has affected the pace of development of the logistics industry. Logistics distribution center in the transportation and distribution operation has certain constraints of the ability with the external transport network. If the demand of distribution is more than external traffic capacity, it will give the entire transport network pressure, and result in the disorder and inefficiency of the outside traffic. Similarly, the logistics distribution plan can't fulfill on time in accordance with the customer's requirements if the distribution ability is exceed. So, the study of match between traffic requirement of the distribution center and transportation capacity of the road has a very important theoretical and practical significance.This thesis researches the traffic match level out of the distribution center, combining the project based on Nation Nature Science Fund-"the traffic link rationality identify and simulation of the urban rail passenger transport hub"(project number: 50778141). In this paper, the rough set theory and neural network theory are combined, to establish a traffic matching model used new BP neural network based on data pretreatment by rough set. This model gets the advantages of the rough set theory and neural network theory. Neural network has several characteristics, such as disposing datum in parallel, aiming for overall function, storing information distributed and so on, which can produce a non-linear map by training and learning, cluster data adaptively, with the abilities of restraining the noise's disturbance and good robustness. Its shortcomings are that when the space dimension of the input information is large, not only complex the network's structure is, but also the network need more time to train. Rough set can deduct feature and value of the data, eliminate the noise and redundant targets in sample. The merge not only reduced the scale of the network, lessened burden of training and learning by eliminating redundant targets, but also improved the accuracy of detection by eliminating noise. Therefore, the integration of rough set theory and neural network method is suited to deal with this kind of unstructured, non-linear complex system such as traffic matching because of their advantages.There are five parts in this paper. Firstly, it introduces the research background and significance of traffic Matching and the status quo of study at home and abroad. The second part describes the characteristics of the city main road traffic and the features of freight transport out of the distribution center. The third part describes the road traffic capacity and traffic demand, expounds the impact factors of the traffic capacity and demand and classifieds the level of traffic match. Part IV introduces the rough set theory and neural network technology, lists the methods of the model which builds by rough set theory and neural network technology, and creates a rough set -neural network traffic matching prediction model, then verifies this model's efficiency compared with traditional BP neural network models. Finally, it summarizes the main conclusions of this paper, and the future direction.