SPATIAL CONFIGUARION AND VEHICLE FLOW : TOPOLOGICALLY ANALYZING AND MODELING THE HONG KONG STREET NETWORK

Space syntax has been considered to be an important theory and analytical tool to study the correlation between spatial configuration and human social activities. But its traditional Axial Model has limitations in representing street. The conclusion got form Axial Model，that spatial configuration of street network can well predict the traffic flow, has been widely doubled. In order to testify the conclusion, the thesis sets out to use Axial, Stroke and Named Street Models to model and analyze Hong Kong street network. Our research methodology is first to create and study different models of street network in pilot study area- Kowloon peninsula of Hong Kong, from the perspectives of space syntax theory and properties of complicated network. Through the pilot study, tentative correlations and conclusions could be derived, which are verified through the case study of whole street network of Hong Kong by taking samples from three different sampling criteria. Through analysis, we find out that local integration best correlates with vehicle flow, and this correlation is called predictability of street network. Through comparisons of different models in terms of predictability, we conclude that stroke model has the best ability to predict vehicle flow. By analyzing the axial model of Hong Kong street network and comparing its result to early study, we prove that axial model does have limitations to represent street network. Also we find out all models of street network have properties of small world network and scale free, from the topological studies of these models. In the research of this thesis, we develop an extension of ArcGIS, named Axwoman 4 in order to calculate and extract space syntax parameters from different models. And important implementation algorithms are introduced in this thesis. The thesis is summed up at the end, and future research directions are given