Modified Driving Safety Field Theory and its Application to Collision Warning Algorithm in Complex Traffic Environments

Vehicle driving safety is influenced by many factors including drivers, vehicles and roads. It is difficult to adapt vehicle kinematics- and dynamics-based driving safety assistance systems to complex traffic environments. In this study, a vehicle collision warning algorithm is proposed based on a driving safety field, which consists of potential, kinetic and behavior fields. The algorithm uses field theory to represent driving risk because of various traffic factors. Field experiment results show that the driving safety field can reveal driver–vehicle–road interactions and their influences on driving safety, and the vehicle collision warning algorithm can adapt to driving decision-making and provide relevant warnings.