Rollover-Free Trajectory Planning for Reconfigurable Vehicles Under Terrain Uncertainty

This paper presents a trajectory planning formulation that enables efficient collision and rollover-free motion planning for reconfigurable off-road vehicles, under terrain uncertainty. The corresponding motion planning problem is stated as a nonlinear optimal control problem (NOCP) subject to vehicle kinematics, obstacle avoidance and rollover stability constraints. The latter takes the form of chance constraint expressed for probability of the machine maintaining rollover stability in the presence of uncertainty in the terrain map. To achieve efficient solution to the NOCP, a novel linear model for uncertainty in rollover based on the properties of rotational transformations is derived. The novel model is then utilized to modify the rollover-stability constraint. To showcase its benefit, comparisons between the proposed model of probabilistic rollover stability measure and the naive noise model are provided through a Monte Carlo simulation. The formulation's practicality is demonstrated in simulation for a feller-buncher machine, traversing undulating 3-D terrain with uncertain map. Results demonstrate stable relocation of the machine and computational efficiency of the trajectory generation.