Path Planning for Obstacle Avoidance of Manipulators Based on Improved Artificial Potential Field

In this paper, artificial potential field is applied to real-time path planning for obstacle avoidance of manipulators. In this approach, a collision-free path from the initial state to target is planned in three-dimensional space. Artificial potential field is widely used in obstacle avoidance for manipulators, with the advantages of simple structure and real-time performance, but there is a local minima problem in this method. Firstly, the artificial potential field method is used to determine the joint-space path of manipulator, and the path is smoothed to ensure the flexibility of joint-motion. Secondly, the local minima problem is decomposed into two categories: unreachable target and local minima. Virtual obstacles and targets are set to avoid the problem. Finally, regarding time as the optimization target, simulated annealing algorithm is used to search the optimal parameters of artificial potential field. The simulation results verify the effectiveness and adaptability of the algorithm.