Obstacle-Avoidance Trajectory Planning of Manipulator Based on BP-RRT*-FN Algorithm

To address the inefficiency of path computation for multi-Degree-Of-Freedom manipulator in cluttered environments, an improved BP-RRT*-FN obstacle avoidance trajectory planning algorithm is proposed. To achieve a more rapid and accurate convergence in the 3D space trajectory planning, the node sampling probability is calculated using the distance weighting function. Combined with the spherical envelope obstacle and the axial envelope method, an obstacle-free path is calculated. Through stage-local search, the BP network is trained to predict the number of nodes sampled in each stage of the local search, and then automatically move on to the next stage of search. To reduce the generation of redundant sampled nodes and improve the efficiency of path optimization, the FN algorithm is applied to random delete nodes. The simulation results show that the average sampling nodes have respectively optimized by 24.96% and 25.30%, and the path calculation time has been reduced by 6.47 seconds and 3.87 seconds respectively. In the six-degree-of-freedom mechanical arm experiment, the results indicate that when completing the obstacle avoidance movement, the average grasping time of the mechanical arm has been reduced by 3.58 seconds and the average search time has been reduced by 3.21 seconds. This algorithm can improve the path calculation efficiency of multi-degree-of-freedom mechanical arms in multi-obstacle spaces.