Obstacle avoidance path planning for the UR10 robotic manipulator: an improved RRT* algorithm

Abstract

This dissertation focuses on the path planning of the UR10 robotic manipu- lator. Using the D-H method, the link coordinate frames for the UR10 are established, and both the forward and inverse kinematics are verified. Then this dissertation develops an improved RRT* algorithm. These improvements include a goal-oriented strategy, a dynamic step strategy, K-d tree optimization, prun- ing, and path smoothing, aimed at increasing the efficiency and quality of path planning. Based on the foundation above, simulations are conducted in both 2D and 3D obstacle environments in MATLAB. The results demonstrate the superi- ority of the new algorithm compared to traditional RRT and RRT* algorithms. In addition, a UR10 robotic manipulator simulation system is also built in ROS environment, to verify the improved algorithm’s capability in practical applica- tions.