Trajectory tracking control for autonomous vehicles

Abstract

Trajectory tracking control is an essential technology for achieving intelligent driving in autonomous vehicles, involving both longitudinal speed tracking and lateral path tracking. This study proposes an integrated lateral and longitudinal control strategy to address vehicle trajectory tracking challenges. First, a two-degree-of-freedom single track lateral dynamics model and a longitudinal dynamics model are established to describe vehicle behavior. Based on these models, lateral controllers and longitudinal controllers are designed, along with driving/braking system. A joint simulation platform combining MATLAB/Simulink and CarSim is used to validate the feasibility and effectiveness of the proposed controllers. Comparative analyses are conducted on four controller combinations: PID+LQR, SMC+LQR, PID+MPC, and SMC+MPC. Simulation results indicate that all combinations achieve satisfactory trajectory tracking performance. Notably, the PID+MPC and SMC+MPC strategies exhibit superior tracking precision and stability, which can effectively improve the passengers’ comfort.