Linear Time-Variant Vehicle Trajectory Guidance with Nonlinear Inversion-Based Feedforward Control

Abstract

Abstract This work presents a two-degree-of-freedom (2DOF) control concept for vehicle trajectory guidance for automated driving on highways consisting of a nonlinear feedforward control and a linear time-variant feedback controller. To obtain a linear time-variant system of the error dynamics the nonlinear vehicle model is linearized along a given trajectory. To relieve the feedback controller future knowledge of the vehicle movement contained in the trajectory is used to create a nonlinear inversion-based feedforward control. Finally, the control concept is applied in a sophisticated simulation environment and compared to a nonlinear benchmark controller. In addition, model states are chosen in a way that all required signals can be obtained by state-of-the-art hardware of production vehicles.