Obstacle Avoidance Path Planning Based on Improved APF and RRT

Abstract

In recent years, collaborative robots have been widely used in the field of intelligent manufacturing. Obstacle avoidance path planning, as a key technology of collaborative robots in motion planning, is becoming more and more important. In order to solve the shortcomings of APF easily falling into local minimums and the shortcomings of RRT's strong randomness, this paper improves and combines these two algorithms. Firstly, artificial potential field algorithm is used in local obstacle avoidance path planning. When falling into local minimum, the improved RRT algorithm can adaptively select the temporary target point and make the search process jump out of the local minimum point. In addition, artificial potential field algorithm will be applied when the robot arm leaves the local minimum. Finally, we use a greedy algorithm to remove redundant nodes in the path. Experiments show that this algorithm can greatly improve the success rate of planning and reduce planning time.