Kinetostatic calibration of a SCARA robot

In this paper we propose a new algorithm for the kinetostatic calibration of a robot based on contact force measurement. The proposed calibration methodology, called kinetostatic calibration, does not require the use of a sensor to measure the robot pose, avoiding all problems related to the use of external devices for this measurement. On the contrary, the robot gripper must be equipped with a force sensor. The robot gripper is requested to touch, with a predefined force, several points of a body of known shape, called calibration object. At each contact point, the contact force and the joint rotations are recorded, whereas the end-effector pose can be estimated using a kinetostatic manipulator model. The calibration can be performed on the base of the difference between the measured contact force and that predicted by a model of the manipulator and of the calibration object. In this case the set of the parameters L describing the robot includes the geometrical dimensions, the robot compliance and the backlash. After a theoretical introduction, this paper describes the implemented algorithm, the simulations and some experimental results obtained with a SCARA robot.