Vehicle motion planning and control for autonomous driving intelligence system based on risk potential optimization framework

Motion planning is a key aspect of autonomous vehicle motion control system. Current driver assistance systems for collision avoidance reach their limit in the case that the obstacle appears suddenly due to physical limits of vehicle dynamics and sensor latency time. The autonomous driving control is designed based on the key concept of experienced driver behaviour modeling by assuming the artificial potential fields on driving environments. Considering the situational risk assessment and hazard anticipation characteristics of experienced drivers is a key point of this study to enhance active safety performance of the current driver assistance systems. This paper combines the idea of the optimal control and the potential field to optimize the safe trajectory and safe velocity for autonomous vehicles and its effectiveness is verified by simulation compared with the measurement data of experienced drivers.