Vehicle rear-end collision warning algorithm based on driver's braking behavior

A large number of naturalistic driving data are collected in Beijing by using video drive recorder. Based on driver's braking behavior in rear-end collision, the relationships between safe/dangerous vehicle following states and vehicle state parameters are analyzed, and a rear-end collision warning algorithm in accordance with the characteristics of driver's risk perception is established by Fisher discriminant analysis. The results of study show that the discrimination accuracy of the algorithm reaches as high as 95%.