A lateral and longitudinal control method based on linear quadratic regulator for intelligent vehicles considering future path changes

Abstract To improve the tracking performance of intelligent vehicles, a lateral controller based on linear quadratic regulator (LQR) theory, and a longitudinal controller based on backstepping sliding mode control (SMC) theory are proposed in this paper. Firstly, a feedforward LQR controller was established based on a two-DOF (degree of freedom) vehicle dynamics model. To solve the stability reduce problem of feedforward LQR controller caused by model linearization, the controller was improved based on the constant turn rate and velocity (CTRV) model. To further improve the predictive controller, an adaptive prediction time mechanism based on the Particle Swarm Optimization algorithm (PSO) was established. Finally, a longitudinal tracking algorithm based on backstepping SMC was proposed. To verify the performances of the proposed controllers, co-simulation and hardware in loop (HIL) experiments were conducted. The results show that the proposed controllers have both stability and accuracy, which can significantly improve tracking performance.