6자유도 로봇 디버링 시스템의 재현 성능 향상을 위한 경로 계획 및 접촉력 제어에 관한 연구

Abstract

As the automation systems for industrial robots are developed to replace the manual machining processes, which have been carried out by workers in small and medium companies, diverse repetitive processes can now be performed easily. In addition, studies on industrial robot automation are actively underway. In this study, the human-robot cooperation system was introduced to the robot automation system, wherein the worker directly performs teaching and the robot system playback what is taught. The deburring robot system was used to verify the direct teaching and playback method. To compensate the positional errors and uncertainty arising from direct teaching process, the Douglas-Peucker Algorithm (DPA) was used to recreate the teaching data for playback. The recreated DPA points were used to create the robot manipulator path. In addition, the repulsive direction vector was proposed to compensate the contact force between the robot's end-effector tool and the workpiece. In this method, if the contact force between the tool and the workpiece exceeds a specified value, the robot’s end-effector moves the tool in the direction that is normal to the contact to reduce the contact force. In the previous study, the repulsive direction vector was proposed considering only the component of the force measured from the F/T sensor. In this study, the correlation between force and moment was specified to make the repulsive direction vector react more intuitively, and the tool frame was also considered in addition to the end-effector to improve the estimation of the normal component of the proposed repulsive direction vector.