Analysis Of China's Urban Network Structure From The Perspective Of "Streaming"

The complex network composed of people flow,logistics and information flow between cities provides a new perspective for studying the city scale and urban agglomeration system,and it is a research hotspot in the current academic circles.With the help of network science theory and the spatial analysis method of traditional geographic information science,the urban network constructed by “flow” can be used to understand the level,scale and role of different cities,as well as the urban architecture and group situation.The analysis of the urban network structure can grasp the spatial pattern characteristics of China's urban development from a global perspective,and provide reference for the coordination of industrial division of labor,the complementarity of urban functions,and the preparation of national spatial planning.Based on the migration data of Tencent population,this paper analyzes the importance degree of 363 cities in China by abstracting the city as a node and inter-city migration as the side,and analyzes the cohesive subgroup and hierarchical structure in the network.In order to accurately obtain the ranking of urban influence,this paper adopts four methods: degree centrality,closeness centrality,eigenvector centrality and Page Rank centrality to complete node centrality analysis.The extraction of cohesive subgroups in the network is realized by three community division kernel algorithms: Louvain,Infomap and Label.Propagation,and finally the module structure of China's urban network.Based on the community division,the K-core algorithm is used to further analyze the multi-level structure of the urban network and the topological structure of the cohesive subgroup.Correlation path analysis,node centrality analysis,community detection and K-core analytic results are highly consistent,which can effectively identify the spatial structure,topological structure,geographical proximity characteristics,and gradient spillover effects of urban networks in China.China's urban network has the following characteristics: 1)The core cities and backbone networks with high influence areconcentrated in the southeast wall of the“Hu Huanyong Line”.The southeast coastal cities have close networks,and the central and western cities have relatively loose and low-level networks.2)The urban network topology is mainly divided into four major regions,among which the diamond structure with the apex of “Beijing-Tianjin-Hebei--Yangtze River--Pearl River Delta--Chengdu-Chongqing City Group” is obvious.It can be seen that the national-level urban agglomeration has a significant tie action.3)The network presents an obvious "core-edge structure" with hierarchical distribution.4)Condensed subgroups are highly correlated with administrative divisions,but community division and K-core analysis can effectively identify micro-communities that transcend administrative boundaries and relatively independent provincial-level urban agglomerations(such as Fujian Province),reflecting the need to make up for Network structure vulnerabilities and short board areas.5)K-core algorithm can accurately identify regional portal cities,provide a reference for each region to improve the external communication path and reduce path dependence.6)The emerging urban clusters represented by the Lanxi urban agglomeration and the northern slopes of the Tianshan Mountains show that the urban network has a great trend of breaking through the “Hu Huanyong line”,which is of great significance for the balanced development of China's cities.