Abstract: China has the largest high-speed railway (HSR) system in the world, and it has gradually reshaped the urban network. The HSR system can be represented as different types of networks in terms of the nodes and various relationships (i.e., linkages) between them. In this paper, we first introduce a general dual network model, including a physical network (PN) and a logical network (LN) to provide a comparative analysis for China's high-speed rail network via complex network theory. The PN represents a layout of stations and rail tracks, and forms the basis for operating all trains. The LN is a network composed of the origin and destination stations of each high-speed train and the train flows between them. China's high-speed railway (CHSR) has different topological structures and link strengths for PN in comparison with the LN. In the study, the community detection is used to analyze China's high-speed rail networks and several communities are found to be similar to the layout of planned urban agglomerations in China. Furthermore, the hierarchies of urban agglomerations are different from each other according to the strength of inter-regional interaction and intra-regional interaction, which are respectively related to location and spatial development strategies. Moreover, a case study of the Yangtze River Delta shows that the hub stations have different resource divisions and are major contributors to the gap between train departure and arrival flows.

Key words: China's high-speed rail, dual network, network analysis, urban agglomeration