Research on lane change trajectory and tracking control of unmanned vehicle
The unmanned vehicle lane change is a key technology in autonomous driving.To improve the planning and tracking control of unmanned vehicle lane change trajectory, an electric vehicle is selected as the research object and a trajectory tracking controller is designed considering transverse-longitudinal cooperative control.Firstly, the vehicle path change is simplified into lateral and longitudinal movements by decoupling.A linear quadratic regulator lateral path tracking controller is designed according to the path tracking error dynamics model and a feed-forward control is introduced to design a feed-forward-feedback LQR lateral controller as a way to reduce the system steady-state error.Subsequently, a dual PID longitudinal controller is designed to track the vehicle speed considering position-velocity errors.And a fifth polynomial curve is chosen to plan the change of path trajectory.Finally, a joint simulation platform is built by Matlab and CarSim to test the designed transverse-longitudinal cooperative controller for lane change trajectory tracking.Our results show the controller is able to track the lane change trajectory with high accuracy under dual lane change conditions, with a maximum deviation of 0.01 m in distance and 0.007 rad in heading angle, demonstrating the controller' ssuperior performances in achieving accuracy and stability.
unmanned vehiclestrajectory planningtrajectory trackinglinear quadratic regulatorlateral and longitudinal control
国家科技期刊平台
NSTL
万方数据
