Nonlinear model predictive lateral control for automated parking system with position and attitude coordination control

Abstract

Automated parking has been estimated to be one of the first commercially available functions for automated driving. For most automated parking controllers, however, the vehicle attitude and the desired yaw angle are not considered, which means that the vehicle can park but stop in the wrong direction. To overcome these issues, a nonlinear model predictive control (MPC) algorithm is proposed to improve the overall following accuracy of the desired vehicle position and attitude. The proposed MPC controller not only leads to a small lateral distance error but also guarantees that the vehicle follows the desired yaw angle of the reference path, which can make automated parking more accurate. The proposed nonlinear MPC controller is firstly verified by simulations on PreScan. Furthermore, the controller is verified the effectiveness and practicability by a Lincoln MKZ platform.