Autonomous vehicle trajectory tracking control is one of the hot topics in the autonomous driving field. One of the most widely used control methods is MPC (Model Predictive Control). As the control system generally becomes more nonlinear and complex, more nonlinear system factors are added to the MPC method. However, tracking accuracy and the amount of calculation needed are both dependent on a lot of contradictions for NMPC (Nonlinear Model Predictive Control). This research proposes a control algorithm for MPC-fused PID (Proportional-Integral-Derivative) control that ensures tracking accuracy under different high-speed driving conditions on wet and slippery road surfaces. The objective of the algorithm is twofold: first, to enhance trajectory tracking accuracy, and second, to ensure real-time control and optimize the vehicle’s comfort, economy, and safety indexes. The results of the joint simulation in Carsim/MATLAB Simulink show that trajectory tracking accuracy is improved by at least 22.2% under high-speed driving conditions of a vehicle on a wet and slippery road. At the same time, the comfort, economy, and safety of the vehicle are improved by at least 9.4%, 19.8%, and 5.3%, respectively.

In autonomous driving, traditional model predictive control (MPC) faces a fundamental trade-off between control accuracy and real-time computational efficiency. This limitation frequently results in hazardous conditions during high-speed maneuvers. To address the tension between tracking accuracy and computational efficiency in high-speed emergency evasive maneuvers, a dual-horizon adaptive model predictive control (DHA-MPC) framework adopting an offline-online collaborative architecture is proposed. Offline, an adaptive weighted particle swarm optimization (AW-PSO) algorithm globally optimizes the prediction and control horizons (Np, Nc). This optimization rigorously incorporates coupled constraints, including vehicle speed, variable road adhesion, and path curvature. Online, a lightweight multilayer perceptron (MLP) model serves as a sub-millisecond real-time inference engine to dynamically adjust the horizon parameters without updating complex prediction matrices. Through experimental verification under the extreme condition of a high-speed double lane change (DLC) maneuver at 70 km/h with a sudden variation in the adhesion coefficient, the DHA-MPC framework effectively limits the peak lateral error to 0.78 m. Furthermore, it significantly suppresses severe transient oscillations in the yaw rate and lateral acceleration. Hardware-in-the-loop (HIL) testing confirms the high determinism of the MLP forward inference time, recording an average execution time of only 26.8 seconds. Finally, Gaussian noise injection experiments further verify the trajectory tracking robustness and the active safety baseline of the system under the degradation of state observation.