A technique for the independent-axis calibration of robot manipulators with experimental verification

Abstract

Accurate use of robots in an off-line programming mode is only possible through a proper calibration procedure. In this procedure, the end-effector is made to move along a set of known spatial poses where the positional errors are to be measured and employed in mathematical models. The models are subsequently solved for the manipulator dimensions (geometric parameters) using suitable regression techniques. Calibration is usually performed using either aggregate or independent-axis models. While the aggregate models result in all the system parameters being worked out simultaneously, the independent-axis models are meant to work out the geometric particulars of each joint-axis individually. In the present work, the independent-axis technique is used for the analysis with new mathematical models proposed to overcome the drawbacks of the existing methods. Moreover, the techniques employed here result in the prediction of transmission error functions and the modelling of the joint motion dependencies. This is a new concept in the field of robot calibration. Finally, the models proposed have been used to calibrate an ASEA IRB/L6 robot and the results are reported at the end of the paper.