Adaptive rapidly-exploring random tree for efficient path planning of high-degree-of-freedom articulated robots

This article proposes a method for the path planning of high-degree-of-freedom articulated robots with adaptive dimensionality. For efficient path planning in a high-dimensional configuration space, we first describe an adaptive body selection that selects the robot bodies depending on the complexity of the path planning. Then, the involved joints of the selected body are included in the planning process. That is, it builds the C-space (configuration space) with adaptive dimensionality for sampling-based path planner. Next, by using adaptive body selection, the adaptive rapidly-exploring random tree (RRT) algorithm is introduced, which incrementally grows RRTs in the adaptive dimensional C-space. We show through several simulation results that the proposed method is more efficient than the basic RRT-based path planner, which requires full-dimensional planning.