Research on trajectory tracking control based on adaptive prediction time-domain MPC

To enhance trajectory tracking performance in autonomous vehicles, this study develops a fuzzy-based adaptive model predictive control strategy that dynamically adjusts the prediction horizon by incorporating road curvature and target speed information. The proposed approach establishes comprehensive vehicle dynamics and control models, employing fuzzy logic to optimize prediction horizon parameters in real-time based on varying operational conditions. Simulation experiments conducted through Carsim/Simulink co-simulation demonstrate substantial performance improvements: in double lane change scenarios, the maximum lateral tracking errors decrease by 72.37% (low speed) and 72.14% (high speed), while serpentine scenarios show reductions of 96.51% and 71.90% respectively. These results confirm the controller's effectiveness in significantly improving both tracking precision and stability across different speed ranges.