Unmanned vehicle trajectory tracking control based on weighted variable time domain MPC

In order to improve the accuracy and stability of unmanned vehicle trajectory tracking, a vehicle motion control method based on weighted variable time domain model predictive control (MPC) was proposed. The grey relational analysis method was used to determine the optimal time domain under different speed conditions, and the Fourier approximation method was used to fit the time domain parameters. Taking the controller calculation time as the weight coefficient, a weighted variable time domain MPC semi-empirical model with time domain parameters varying with vehicle speeds was obtained. The model can select the relative optimal time domain according to the change of vehicle trajectory tracking speed, and improve the tracking accuracy and stability of the controller under different vehicle speeds. Selecting 50 and 100 km/h as representative vehicle speeds, the control effects of the traditional time-domain MPC controller and the weighted time-domain MPC controller were compared by simulation experiments. The results show that the weighted variable time domain controller can effectively improve the tracking performance of the controller, the maximum lateral deviation is reduced by 1.98%, the maximum yaw angle deviation is reduced by 60%, and the controller solution time is reduced by 7.2%. And it also has strong adaptability to different target speed conditions.