Longitudinal Collision Avoidance of Driverless Vehicles based on Improved Risk Assessment

Aiming at the problem that longitudinal collision may occur when the vehicle is driving, this paper designs a collision avoidance control system suitable for high / low speed. Firstly, an improved risk assessment algorithm is proposed. Based on the calibration of longitudinal collision avoidance threshold, a layered collision avoidance controller is established. The upper controller outputs the expected acceleration based on fuzzy rules, and the lower controller controls the output of the actuator based on inverse dynamics model and PID algorithm. To achieve the purpose of automatic collision avoidance by outputting appropriate acceleration and deceleration according to different situations. Finally, the longitudinal collision avoidance system is simulated and verified through three working conditions and six scenarios. Results show that the vehicle can stop or follow the vehicle within a certain safe distance from the vehicle in front, and meet the requirements of longitudinal collision avoidance.