A linear time-varying model predictive control algorithm for autonomous vehicles

Autonomous vehicle will be widely used in airport flight zones, industrial parks, ports and other scenarios in the future. A linear time-varying model predictive control (LTVMPC) algorithm is proposed for the trajectory tracking accuracy and stability requirements of autonomous vehicle. Firstly, a vehicle dynamics model based on an empirical tire model and small angle assumption is developed. Then, the linear time-varying model predictive control algorithm is designed. Relaxation factors are introduced to solve non-feasible solutions of algorithms due to hard constraints in optimization solving processes. Finally, the Carsim/Simulink co-simulation experiment is carried out. The simulation results of tracking double-shifted trajectories show that the trajectory tracking controller designed in this paper can accurately track the reference trajectory, and has smaller tracking error, smaller tire sideslip angle, vehicle sideslip angle, and lateral acceleration compared with the LQR algorithm, and has better performance.