Estimating freeway accident-prone sections: research on vehicle dynamic simulation and accident prediction model

Abstract

Accident-prone sections are usually the most dangerous areas on a freeway, and identifying and studying them is of great significance. In this study, accident-prone sections were estimated using a 77-km long freeway section as an example located in Zhejiang province, China. This research obtained accident data and horizontal alignment data of the freeway and used two statistical methods, namely the improved cumulative frequency method considering equivalent accident numbers and the accident matrix method, to identify the accident-prone sections of the freeway. Additionally, a new method that combines vehicle dynamic simulation (using CarSIM) and speed consistency theory was introduced and used to estimate the accident-prone sections of this freeway. Furthermore, an accident prediction model was established by analyzing the relationships between the horizontal alignment geometric parameters and accident data to provide design advice. The results show that the new method has certain reliability, making it applicable for designers to verify the safety condition of newly-designed freeway and help enhance freeway security. The established model is logically reasonable and has certain precision, which can provide a valuable reference for freeway designers.