A Novel Heading Predictive Control Model for Autonomous Ground Vehicles

A novel heading predictive control model for autonomous ground vehicles is proposed in the present paper. The developed approach includes a nonlinear vehicle dynamic model which takes the actual tire side - slip characteristics into account and a new predictive function obtained from the stationary solution of the 2DOF vehicle dynamic model. The effectiveness of the new heading predictive control model is identified in different vehicle speed and desired heading angle inputs, also showing the increasing tracking error as the vehicle is running faster or turning more severely. Comparisons to the previous model are conducted for both step response and sinusoidal response, and it is revealed that the new model is more realistic and appropriate for the handling of future autonomous vehicles. Although the old model shows good tracking accuracy, it should be used more carefully in practice due to the imprecision of the control system. The present model has a potential to test whether the planning of path and heading angle can be correctly tracked in the real condition, and it is valuable to be used to investigate the correlation between the constraints of path planning and vehicle dynamics.