Manipulator Path Planning based on RRT-Connect with fused Artificial Potential Field

In this paper, a fast planning algorithm based on the Rapidly-exploring Random Trees (RRT) algorithm is proposed to plan the efficient and obstruction-free path for manipulators. Firstly, the algorithm utilises the target bias method to reduce blind search. An adaptive probability method is also employed to adjust the algorithm's preferences according to the environment. Secondly, the efficiency is improved through the use of a dual-tree search algorithm. Additionally, the algorithm incorporates the Artificial Potential Field (APF) method, which analyses the forces acting on the path tree's points closest to the target point. The identified points are incorporated as a random point generation option in the RRT algorithm, enhancing the map's perception. Finally, it is demonstrated through simulations and experiments that the proposed algorithm reduces the search time and the total number of sampling points by 80 % compared to the traditional RRT algorithm. Compared with the existing algorithm (Fast-RRT), the search time is reduced by 20 % and the sampling points are reduced by 30 %.