Abstract: According to the low lateral control accuracy and low robustness of lane keeping system based on typically position error control , a lateral control algorithm based on tracking desired yaw rate is proposed. A virtual path is planned between vehicle and preview target point, yaw rate and velocity are analyzed while a vehicle tracking the virtual path with zero tracking error, a desired yaw rate generator is designed based on the body status and correspond positions of vehicle and preview target point. A yaw rate tracking controller is designed to track the desired yaw rate based on vehicle seven-degree-of-freedom nonlinear dynamic model, and vehicle can run according to the virtual path accurately in the end. Lane edge detection algorithm based on lane width and statistical edge points can detect vague lane and eliminate noise,vanishing point is detected to filter the straight line which are not lane line. Simulation and experiment studies shows that the proposed lateral control algorithm not only can effectively control vehicle to run according to the pre-given path perfectly under different velocities and road curvature but also have good robustness and adaptability.

Key words: Desired yaw rate, Lane detection, Lane keeping system, Lateral control, Path planning