Driver injury severity in two-vehicle accidents considering collision role

Abstract

: In order to study the influencing factors of driver injury severity and the interaction effects of collision roles in two - vehicle accidents, based on the data of two - vehicle collision accidents in Shenzhen from 2018 to 2020, we calculate the value of importance degree of characteristic variables by using random forest algorithm to screen out the candidate independent variables, and establish a binary logit model of driver injury severity considering collision angle. The results show that collision role, accident responsibility, vehicle type, accident form, driving experience, visibility, collision form between vehicles, educational background, road grade, use of safety protection devices, roadside protection facilities and lighting conditions are significantly associated with the driver injury severity in two vehicle accidents. Meanwhile, there are significant interaction effects among collision role, vehicle type and accident form. The results are helpful for understanding the influence mechanism of driver injury severity in two - vehicle accidents and can help traffic management departments to formulate corresponding traffic safety countermeasures.