The Modification Of Betweenness Centrality With Combination Of Traffic Demand Information

As the skeleton and basis of urban road traffic, urban road network plays an important role in the operation and development of urban. Space syntax and its derived theories for space analysis argue layout and configuration of urban space effect the human movement,and that topological characteristics of urban road network have a close relation with traffic flow. In previous studies, researchers have been trying to find an appropriate topological measure with high correlation with traffic flow, which could be used to model traffic flow applying in transport planning. Betweenness centrality used to be regarded as the ideal one to build that model, while recent researches have demonstrated that it's not able to logically describe the characteristic of traffic flow because of the limitations of its own definition.This thesis focuses on the modification of betweenness centrality, expecting to propose a suitable measure. The main contents are as follows:(1) We summary mainly existing topological representations of urban road network,and compare road segment approach and stroke approach by a case. According to the result,road segment approach is selected and applied in this thesis.(2) We analyze the relation between structure characteristics of urban road network and the inner roads, and explore the mechanism from the point of space syntax. Based on recent studies, limitations of betweenness centrality in describing traffic flow are pointed out.Traffic demand information (i.e., OD distribution), then, is introduced to modify betweenness centrality, and corresponding method and algorithm and its implementation are proposed.(3) We extract the flow and traffic demand of taxi travels from the taxi GPS data of Nanjing (China) and San Francisco (USA), and calculate the value of the modified measure of each road segment in both cities. Then correlation analysis between the modified measure and taxi flow is conducted, and results show that the proposed measure has a significant improvement in correlation with flow. In the end, based on the data of a certain day of Nanjing, we analyze and judge the distributions of demand and flow of taxi trips according to the distribution of the proposed modified betweenness centrality (MBC),which is compared with the distribution of observed road-segment taxi volume.