Optimal Design of MPC Autonomous Vehicle Trajectory Tracking Controller Considering Variable Time Domain

In recent years, with the indepth research on driverless technology, model predictive control theory was extensively applied in the field of vehicle control. In order to improve the accurate tracking of reference trajectories by driverless vehicles, a model predictive control trajectory tracking controller for driverless vehicles optimized by an improved sparrow search algorithm is proposed. Firstly, an objective function with constraints is added to the model predictive control trajectory tracking controller by establishing the vehicle dynamics model; Secondly, the improved sparrow search algorithm is enhanced to speed up convergence and expand the program's search capabilities; Then, in order to discover the best value, the model predictive control trajectory tracking controller's prediction time domain and control time domain are optimized using the improved sparrow search algorithm; Finally, to confirm the method's viability, collaborative simulations in Simulink/Carsim were completed. The simulation results show that the lateral errors generated by the improved sparrow search algorithm-based optimized model predictive control trajectory tracking controller are reduced by 53.53% and 65.44%, respectively, when the vehicle speed is 36 km/h, compared with the traditional model predictive control trajectory tracking controller. When the vehicle speed is 54 km/h, the lateral deviations are reduced by 81.08% and 86.76%, respectively. In addition, the optimized model predictive control trajectory tracking controller improves the accuracy and at the same time, the driving stability of the control vehicle is significantly improved.