Obstacle Avoidance Trajectory Planning of Manipulator Based on Improved RRT Algorithm

An improved RRT algorithm based on centripetal expansion, target bias and local path optimization is proposed to solve the problems of weak goal orientation, low planning efficiency and long planning path of trajectory planning algorithm for obstacle avoidance and grasping task of manipulator. The algorithm uses the bi-directional centripetal expansion strategy to change the extension decision condition and speed up the convergence of the expansion tree. Then the super-ellipsoid bias method is introduced to increase the target orientation of the sampling points and greatly accelerate the convergence speed. Finally, the expansion tree nodes are optimized by path cost to locally reduce path length. Simulation results show that the improved RRT algorithm has higher operation efficiency, can reduce the path cost to a certain extent, and can perform obstacle avoidance and crawling task more efficiently.