Trajectory Optimization for Truck-Trailer Systems Based on Predictive Path-Following Control

Autonomous maneuvering of truck-trailer systems is a challenging task, especially if collisions with obstacles must be avoided. Due to the complexity, global path planners typically focus on the efficient computation of feasible but sub-optimal solutions. In this context, this paper presents a model predictive path-following controller for local optimization of the globally planned paths. This allows to eliminate redundant path elements, to increase obstacle clearance and to compensate initial path deviations. Furthermore, it is proposed to constrain the terminal state to the reference path and to formulate the terminal costs in terms of a precomputed reference trajectory, both of which improve the numerical robustness of the optimization. Simulation results show the benefits of the optimized solution in a difficult planning scenario.