FORECASTING SEVERITY OF TRAFFIC ACCIDENTS USING ROAD GEOMETRY EXTRACTED FROM MOBILE LASER SCANNING DATA

摘要

This study analyzed the effects of road geometric elements such as horizontal curvature, superelevation, grade change, and design speed on accident severity. A semi-automatic method is proposed to extract aforementioned geometric elements from mobile laser scanning point clouds. Then a logistic regression model is used to establish relationships between road geometric elements and accident severity. Based on the coefficients calculated by the logistic regression model, the effects of each road geometric element on accident severity are discussed. Results showed that the average superelevation was the most contributed factor to serious injuries, grade change factor was the most critical factor for minor injuries and the damage only of accident severity levels. The analysis revealed that improving superelevations on horizontal curves should be considered by the safety agencies in Malaysia. In addition, design of vertical curves of expressways is also should be imp-roved.