Abstract: According to the nonlinear and time-varying characteristics of vehicle dynamics and driver’s behavior characteristic, a novel algorithm of vehicle direction preview adaptive proportional-derivative (PD) control is proposed based on preview-follower theory, and this approach’s feasibility and validity are verified with the simulation and field experiment. In the algorithm, an first-order linear reference model is established to approximatively describe vehicle direction dynamics characteristic, and the parameters of this reference model’s transfer function were identified on-line; then the parameters of PD controller are self-tuning based on the structure of optimal preview-follower. This algorithm can control vehicle to follow the anticipate trajectory well, the robustness and adaptivity are also improved because of the adaptive control algorithm, and provide an applicable way to the research of intelligent vehicle.

Key words: Adaptive PD control, Direction, Intelligent vehicle, Preview follower