Path Planning Algorithm for Manipulators in Complex Scenes Based on Improved RRT

Aiming at the problems of a six-degree-of-freedom robotic arm in a three-dimensional multi-obstacle space, such as low sampling efficiency and path search failure, an improved fast extended random tree (RRT*) algorithm for robotic arm path planning method (abbreviated as HP-APF-RRT*) is proposed. The algorithm generates multiple candidate points per iteration, selecting a sampling point probabilistically based on heuristic values, thereby optimizing sampling efficiency and reducing unnecessary nodes. To mitigate increased search times in obstacle-dense areas, an artificial potential field (APF) approach is integrated, establishing gravitational and repulsive fields to guide sampling points around obstacles toward the target. This method enhances path search in complex environments, yielding near-optimal paths. Furthermore, the path is simplified using the triangle inequality, and redundant intermediate nodes are utilized to further refine the path. Finally, the simulation experiment of the improved HP-APF-RRT* is executed on Matlab R2022b and ROS, and the physical experiment is performed on the NZ500-500 robotic arm. The effectiveness and superiority of the improved algorithm are determined by comparing it with the existing algorithms.