Research on Trajectory Tracking Control for Autonomous Vehicles Based on Model Parameter Adaptive Correction Controller

Real-time performance and adaptability are critical factors influencing the safety and stability of autonomous vehicle trajectory tracking. Therefore, enhancing these aspects is essential for improving driving safety. This paper proposes a trajectory tracking control method for autonomous vehicles based on an adaptive model parameter correction controller (MPACC). First, by integrating the variable universe fuzzy control (VUFC) principle with a model predictive controller (MPC), a variable universe fuzzy model predictive controller (VUFMPC) is designed. This controller enables adaptive adjustment of MPC weighting coefficients, thereby effectively improving the real-time capability and adaptability of the MPC. Second, an adaptive square root cubature Kalman filter (ASRCKF) tire lateral force estimator with adaptive scaling factors is introduced to obtain real-time tire cornering stiffness values as MPC parameters, achieving adaptive correction of the MPC parameters and forming an adaptive model predictive controller (AMPC). Furthermore, an MPACC is designed by integrating VUFMPC and AMPC. This controller allows for real-time adaptive correction of control parameters according to the vehicle’s driving state. Finally, hardware in loop (HIL) tests are conducted for comparative analysis. The results demonstrate that the proposed MPACC exhibits excellent real-time performance and adaptability, while effectively balancing trajectory tracking accuracy and driving stability of autonomous vehicles.