Research on Trajectory Tracking Control of Driverless Vehicles

The As a key technology for the realization of self-driving cars, the merit of trajectory tracking control directly determines the stability, safety and comfort of vehicle driving. In order to explore the performance of existing mainstream trajectory tracking controllers under different vehicle speeds, this paper designs three types of methods, PI- pure tracking, PI-Stanley and PI-MPC, and obtains the trajectory tracking effect by changing the key parameters of controllers, so as to grasp the parameter design laws of different methods. Firstly, a vehicle system dynamics model is established based on the assumption of planar motion, to guide the design of the trajectory tracking controller. Secondly, the trajectory tracking controller is divided into longitudinal controller and transverse controller, the longitudinal controller is designed based on PI control theory and the transverse controller is designed based on pure tracking theory, Stanley control theory and model predictive control theory respectively. Finally, a double-shift test condition is built based on Carsim/Simulink to verify the performance of the designed method at different vehicle speeds.