Motion Planning and Control for Improved Ride Comfort in Urban Autonomous Driving

Self-driving shuttle services frequently provide a less comfortable experience for passengers than human drivers. This observation emphasizes the importance of motion planning and control methods. This paper proposes a method to improve ride comfort in urban autonomous driving. The path information is utilized for velocity planning to minimize lateral acceleration and limiting the acceleration when controlling the PID to avoid rapid deceleration. Additionally, the hybrid lateral controller increases path following performance. The proposed method proved excellent performance compared to the existing geometric lateral control method. A decision tree was employed to evaluate ride comfort to meet ISO standards. The acceleration and jerk were assessed to determine their impact on ride comfort. Furthermore, the lateral error and the heading error were analyzed to evaluate the stability of the route during autonomous driving. An experiment was conducted and verified in the actual urban environment of Songdo New Town, Incheon, to ascertain the efficacy of the proposed method in enhancing ride comfort.