Development of Traffic Accident Frequency Model for Evaluating Safety at Rural Signalized Intersections

Even though accident frequencies in roadway segments have been decreasing since 2000, there has been increasing the number of vehicle crashes at intersections. Due to this increase, safety problems at intersection recently started to be regarded as significant issues. The purpose of this study is to analyze the effects of road conditions, traffic operational conditions, and other influencing condition on intersection safety. Then a traffic accident frequency prediction model to evaluate the safety at intersections was developed based on the correlations between influencing factors and vehicle crashes. In this research, critically significant factors affecting vehicle crashes at rural four-legs signalized intersections were investigated. It was found that Poisson regression was the best fit method to developing a accident frequency modeling using the collected data in this study. Through this study, it was concluded that exclusive left turn lane, crosswalk, posted speed, lighting, angle, and ADT are significant influencing factors on the intersection safety.