Active Peg-in-hole of Chamferless Parts using Force/Moment Sensor

Abstract

This paper discuss the peg-in-hole task of chamferless parts using force/mement sensor. In this paper, when the directional error occurs during the task, it is supposed that there are two cases with small and large errors, and quasi-static analysis has been accomplished for each case. In most other algorithms, however, when there is large directional error during the task, the robot has to return to the initial position and start the task from the beginning, which is considered to be inefficient. This paper proposes the algorithm which enables a robot to continue the task even with large directional error by adjusting the positional and directional error. This algorithms has been implemented for the 5-axis robot and is shown to be effective.