The Research On Forecasting Reasonable Scale Of Regional Railway Network

In January2004, the "Medium and Long-term Railway Network Planning" delib-erated and considerated by the State Council has opened a new page for China’s railway construction. The reasonable scale of the railway network is considered as a basic measure to instruct the construction, layout and planning of the railway network. However, for the complexity of the influence factors of the regional railway network, which not only includes the level of the regional socio-economic development, railway transport demand, population, industrial structure, natural conditions and land resources, but also the regional transport policy and development strategy. Various factors are interacted with each other. Meanwhile, the particularity of the history development and the uncertainty of the future development in different regions enhance the complexity of the influence factors. Current calculation methods about the regional railway network scale, such as the Land coefficient method, the Analogy method, the Node connectivity method and the Transport strength method, are all lack of comprehensive consideration of the influencing factors. For this reason, aiming to select a more feasible and effective method, the thesis can calculate their railway network scale more accurately.Based on the research background above, the work of the thesis is summarized as follows:1、The thesis focuses on two keywords of the region and the scale of the railway network. Firstly, it elaborates the definition and spatial development theory of the region. Secondly, the thesis analyzes the elements and characteristics of the railway network. In addition, it indicates the adaptation and indexes of the railway network scale, then discussing the general laws of the railway network scale, which sets up the theoretical basis for the thesis.2、On the basis of analyzing the influence factors of the regional railway network scale, the thesis establishes a hierarchical structure model to reflect their hierarchical interaction. Based on a reference to the research of scholars at home and abroad, the thesis builds a influence factors index system of the regional railway network scale. After that, the gray correlation analysis method is used to filter out the main influence factors.3、Based on the analysis of the traditional calculation methods of the railway network scale, the thesis selects the RBF neural network model. Considering the influence factors indexes of the regional railway network scale, filtered out quantitatively in Chapter3, the RBF neural network model to calculate the scale of the regional railway network are established.4、Taking Guangdong province for an example, the thesis calculates and analyzes its railway network scale by the RBF neural network model. Finally, making a conclusion that it is more reasonable to calculate the regional railway network scale through the RBF neural network method.