| Road network, which is regarded as the backbone of integrated urban transport system, is not only a carrier of social and economic activity, but also the main supporting condition of the city’s sustainable development. However, the modern transportation situation is becoming more and more severe, which not only affects the economic construction and development of the society but also causes much inconvenience in people’s life and work. Many traffic problems are partly due to the problems of the road network such as unreasonable layout, unclear hierarchies, and the lack comprehension of road network planning. Therefore, it is necessary to identify the methods and indicators to describe network structure, and analyses its impacts on the distribution of vehicle flow, which will provide theoretical basis and technical support for the future’s urban road network planning and easing urban traffic congestion.In this paper, the dual approach are respectively used for road network topology analysis and abstracting stroke to point, which as a result figures out the three indicators of road network structure:Degree centrality, Closeness centrality, betweenness centrality. The author analyzed the impact of the structural characteristics of road network to path selection from the perspective of spatial cognition, thus affecting the distribution of vehicle flow. The statistical analysis of five cities’ instance data were done respectively, and through one dimension linear regression analysis, btw shows a best correlation with vehicle kilometers travelled than other two indicators of road network structure. At the same time, the betweenness variation is acquired by analyzing the research fields of different road networks. And the effects caused by different research fields on predicting the vehicle flow are analyzing based on it. The suggestion is adopting the bigger research field of the road network with the consideration of the effect on the distribution of the traffic flow caused by road network structure. |