关键词: 机械臂, 动态避障, 快速搜索随机树, 人工势场法, 跟随机制

Abstract: Aiming at the issues of weak path searching capabilities,long durations,numerous redundant nodes,and poor safety in traditional robotic arm dynamic obstacle avoidance algorithms within human-machine collaborative environments,a Bi-GB-APF-RRT^* algorithm was proposed for global path planning. Meanwhile,an improved artificial velocity potential field algorithm was employed for real-time local path planning,combined with a following mechanism to achieve real-time and continuity in dynamic obstacle avoidance,thereby enhancing the utilization rate of the global path. Initially,a random tree was generated from both the starting point and the target point for bidirectional path searching. The goal probability bias strategy and the concept of the artificial potential field guide the growth of tree nodes towards the target,accelerating the path search process. During the growth process,adjacent space search and tree reconnection of tree nodes are conducted to reduce the path′s cost value,thereby optimizing the path and improving the smoothness of the robotic arm′s motion. During the robotic arm′s motion,if the distance between the joint point of the arm and the point cloud of the obstacle is less than the safe distance,the robotic arm will perform obstacle avoidance maneuvers under the combined effects of the attractive velocities and repulsive velocities of the improved artificial velocity potential field. Experimental results indicate that the global path planned by the Bi-GB-APF-RRT^* algorithm is significantly enhanced compared to the traditional RRT algorithm and validate the effectiveness of dynamic obstacle avoidance planning.

Key words: robotic arm, dynamic obstacle avoidance, rapidly-exploring random tree, artificial potential field method, follow-ing mechanism