Hotspot Location Identification Using Accident Data, Traffic and Geometric Characteristics

Abstract

Determining the criterion for critical limits is always one of the essential challenges for traffic safety authorities. The purpose of identifying accident hotspots is to achieve high-priority locations in order to effectively allocate the safety budgets as well as to promote more efficient and faster safety at the road network level. In recent years, human, vehicle, road and environment have been recognized as the three main effective elements of the road transportation in the occurrence of accidents. In the present study, with combining the parameters related to accidents, geometric parameters of the accident location and traffic parameters, hotspots were identified by using the superior methods of Poisson regression and negative binomial distribution and based on the combined criteria of frequency and severity of accidents and equivalent damage factors. Then using Time Series Models in ANN, result were compared and validated. The results of ANN models demonstrate that the frequency method of accidents tends toward places with high traffic volume. MATLAB and STATA software were used. Non-native plumbing, curvature, slope, section length and residential area had more significance, and their coefficients indicated the significant effect of these parameters on the occurrence of the frequency and severity of accidents in hotspot locations.