Abstract:

This paper proposes a method for motion planning of tractor-trailer combinations having active trailer steering. Autonomous driving (AD) in structured environments involves a set of predefined requirements that the vehicle should satisfy, such as lane following, safety distances to surrounding obstacles, speed preferences, or ride smoothness. We have previously shown that by interpreting the motion-planning problem as a nonlinear, non-Gaussian estimation problem, we can leverage particle filtering to efficiecntly determine suitable vehicle trajec- tories satisfying such requirements. In this paper, we extend the motion planner to determine safe and drivable trajectories for semi-trailer articulated vehicles in scenarios requiring complex maneuvers. In a closed-loop simulation study, the trajectories are tracked with a few centimeter accuracy, validating dynamic feasibility of the proposed method.