Thesis Type: Doctorate
Institution Of The Thesis: Gazi Üniversitesi, Fen Bilimleri Enstitüsü, Turkey
Approval Date: 2012
Student: ÖMÜR KAYGISIZ
Supervisor: METİN ŞENBİLAbstract:
The main objective this study is to contribute to traffic safety based urban planning (land use planning, road and junction design) strategies. The study defends the idea that there exists a meaningful relationship between land use decisions and frequency of traffic accidents. Primarily, development processes of transportation planning and the relationship between urban land use and traffic accidents were studied on a theoretical base. Later on, traffic accident prediction models were constructed by using traffic accident data and land use data of Eskişehir City. The models are intentionally constructed as simple and practicable that can predict the expected number of accidents as accurately as possible. These models are useful in land use and transportation planning, urban road network and junction safety analysis and determining the factors affecting the location of the hot spots. In the process of model construction, firstly two databases were created related to junctions and road segments by using GIS. Then, with the help of the established database, development of accident prediction models for junctions and road segments were carried out. The main concern of the established models is to identify the land use components related to traffic accidents by predicting traffic accidents for junctions and road segments. In both junction and road segment modeling, techniques to relate accident frequency to explanatory variables are used. For junctions, Panel estimation in Poisson regression was used whereas Generalized Linear Model (GLM) was used for road segments.