Motion Planning With Differential Constraints as Guided Search Over Continuous and Discrete Spaces

To compute a motion trajectory that avoids collisions, reaches a goalregion, and satisfies differential constraints imposed by robotdynamics, this paper proposes an approach that conducts a guidedsearch over the continuous space of motions and over a discrete spaceobtained by a workspace decomposition. A tree of feasible motions anda frontier of workspace regions are expanded simultaneously by firstdetermining the next region along which to expand the search and thenusing sampling-based motion planning to add trajectories to the treeto reach the selected region. When motion planning is not able toreach the selected region, its cost is increased sothat the approach has the flexibility to expand the search along newregions. Comparisons to related work show significant computationalspeedups