Optimization for railway freight transport network based on BP Neural Network

According to the current situation of China railway freight transport, in order to optimize the railway freight transport network and overcome the existing problems, this study proposes a three-layer network system, analyses the function locating of the three-layer railway stations and the hierarchical structure of the network system, then describes the basic transport organization mode of the network system. The classification of three-layer railway stations plays an important role in the construction of the network system, so a method for the classification based on BP Neural Network is proposed. In the method, an index system containing various influencing factors is established, and an improved BP Neural Network model is designed to divide the railway stations into three layers. In the design, Adaptive Learning Rate Algorithm and Momentum BP Algorithm are used to improve the BP Neural Network´s performance of solving the classification problem. Finally, an empirical study is given to verify the feasibility and accuracy of the model.