Operational kinematic parameter identification of industrial robots based on a motion capture system through the recurrence way

Abstract

This article introduces a new kinematic identification approach for serial robots using a motion capture system in the recurrence way. This method is based on the motion capture system as a measuring instrument, which can measure and identify the robot’s movements in any feasible domain. Different from the traditional method of obtaining measurement information from the end effector of a robot, multiple markers pasted on a robot each joint are identified and tracked by a motion capture system to obtain the 3-D coordinates of markers of each joint in every frame when the robot is being moved continuously. After optimally fitting the rotation matrix and the translation vector of each joint according to the singular value decomposition, the joint axis vectors and DH parameters can be reconstructed in the recurrence way. Both simulation and experiment of a six revolute joint robot are conducted to verify the effectiveness and efficiency of the method. The identification results including 3 invariants are close to the accuracy of the laser tracker. Interestingly, the identified joint angle is in close proximity to that obtained by the control system of the calibration robot.