Target Grasping and Obstacle Avoidance Motion Planning of Humanoid Robot

Abstract

Aiming at the problem of robot motion planning under the environment of obstacles, this paper studies the obstacle avoidance planning method of robot arm under the ROS platform (Robot Operating System). Based on the service robot developed by the team, the robot URDF (Unified Robot Description Format) file is exported from the 3-D SolidWorks model using the SW2URDF plugin, and the configuration and launch files needed for robot arm motion planning are created by MoveIt! Setup Assistant. By using relevant APIs provided by MoveIt! and invoking random sampling based algorithm RRT we completed the obstacle avoidance planning, object grasping and placement tasks in 3-D real-time simulation environment Rviz (ROS visualization), and also derived the position and velocity curves of each joint of the robot arm. Finally, obstacle avoidance planning, grasping and placing experiments were carried out on the robot, the arm successfully grasped the object and placed it on the designated location without collision. The results show that the use of ROS platform for obstacle avoidance planning is convenient and intuitive, and it is instructive to the application research of robot.