Grasping Trajectory Generation of a 7-Dof Robotic Arm Based on Cartesian Direct Teaching Technology

Human-Computer Interaction (HRI) is crucial in robotics. This article introduces a new method for human-machine interaction on robotic arms using a drag teaching approach. By analyzing the kinematics and dynamics of the Xarm7 robotic arm, the method reduces positioning difficulties. The F/T sensor on the arm detects force and torque. Direct human guidance is matched with robot response in real-time to control the arm's position and direction. The research also records and reproduces motion trajectories using cubic interpolation for precision. A mechanical claw on the arm detects and compares sensor data to position and grasp objects. Experimental results show similar shapes between the dragging and reproduced trajectories, with reduced motion errors.