Abstract: Under the background of high-quality urban development and territorial planning in the new era, we need to accurately grasp the characteristics and structural mechanism of urban network. It is of great importance  for urban planning and transportation planning. This paper constructs the urban data matrix from the perspective of railway travels in different periods. The matrix depicts the pattern of urban networks on weekdays, weekends and holidays, and shows the structural characteristics and organizational patterns of different spatial scales. The results show that: cities that are closely related to railway travels are mainly concentrated in the east of the Hu Line, which is coupled with China’s high economically developed and heavily populated areas; The migration intensity of railway travels in holidays is significantly higher than that in other periods; The average travel distance on weekdays is the longest, and the whole network tends to be flat and balanced; The centrality shows a decreasing trend from weekdays, weekends to holidays; The level of railway travel population flow network is apparent. Working days, weekends and holidays form cross-regional urban agglomerations, high-speed rail corridors and urban agglomerations. Through the analysis of path dependence association, advantageous node linkage, and spatial structure interaction at different scales, this paper explores the formation mechanism of the spatial structure of urban network, supporting suggestions on global orbit fusion, key node lifting, and nested space connection.