Robust Predictive Control for Intelligent Vehicle Path Tracking Considering Error Feedback Correction

To further enhance the robustness and real-time performance of path tracking predictive control for high-speed autonomous vehicle, a predictive control method for path tracking considering error feedback correction mechanism is proposed. A three-degree-of-freedom vehicle model with weak nonlinearity is established considering the longitudinal, lateral and yaw motions of the vehicle. The linear time-varying first-order approximation of the control model is realized by Taylor formula expansion. The predictive state variable is introduced to construct a path tracking optimization scheme. Aiming at the problems of mismatch caused by uncertainties such as vehicle parameter deviation, system time-varying and external disturbance, the error between the actual state value and the predicted one is defined, and the prediction output model with feedback correction term is obtained. The robustness of the prediction model of the path tracking control system is enhanced by the error feedback correction mechanism. The optimization of path tracking prediction is transformed into a constrained quadratic programming problem, and the receding optimization of robust predictive control law for path tracking is numerically achieved. The results show that the error feedback correction mechanism can effectively improve the robustness of the predictive control algorithm for path tracking due to the change of vehicle mass and yaw moment of inertia, can realize the desired path tracking function under high-speed driving conditions, and has good real-time performance. © 2020, Editorial Office of Journal of Xi'an Jiaotong University. All right reserved.