Efficient Path Planning Using Harmonic Functions Computed On A Non-Regular Grid

Robot path planning is devoted to the planning of a trajectory for the robot to perform a collision-free motion between to configurations. Potential fields methods have been used to guide the robot through the obstacles in the robotic cell. The use of harmonic functions allows the definition of a potential field without local minima and with the ability to be incrementally updated. In order to efficiently use this approach to motion planning in higher dimensional C-spaces, the paper presents a hierarchical and dynamic method to compute harmonic functions on a non-regular grid.