Akademik Veri Yönetim Sistemi
Tezin Türü: Yüksek Lisans
Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, İnşaat Mühendisliği Bölümü, Türkiye
Tezin Onay Tarihi: 2014
Öğrenci: EZGİ KUNDAKCI
Danışman: HEDİYE TÜYDEŞ YAMAN
Özet:A major problem in the world and in Turkey is traffic accidents and consequent loss of life or property. In addition to the literature on highway black spots, it is necessary to identify urban locations with high traffic accident occurrence rates in order to shed light to underlying problems and develop preventive measures.Despite the more strict black spot definition for highways, accident prone locations in urban areas are regarded as hot spots, which can be simply defined as clusters representing dense accident occurrence areas. This thesis presents “Nearest Neighborhood Hierarchical (NNH) Clustering” method to find hot spots in urban region. Hot spot distribution is evaluated for different urban zones and based on accident type (pedestrian versus non-pedestrian) and occurrence time (morning peak, off-peak, evening peak and nighttime). Moreover, distances between hot spots and major intersections are examined to understand the relationship between them. Study also performs “Kernel Density Estimation(KDE) method”, which is widely used in traffic safety analysis. Different levels of distance threshold are used to evaluate its impact on hot spot definition.Numerical results were obtained for Ankara using traffic accidents which occurred between 2008 and 2010.Since fatality accidents are rare events, it was possible to detect hot spots only when a threshold distance of 1km was assumed. Injury accidents create a significant number of hot spots with even 100 m threshold, which are located at or near major intersections. Pedestrian accident hot spots were condensed at or closer to Central Business District in Ankara, as expected from the literature.