An improved kinematic calibration method of industrial robots based on double ball bar measurements

A novel measurement method based on the double ball bar (DBB) is proposed to improve the positioning accuracy of industrial robots. The kinematic model is established based on the Modified Denavit-Hartenberg (MD-H) method. A measurement posture of the tool center point (TCP) tilted 30° is proposed to ensure that all joints can fully move, and a spherical spiral trajectory is adopted to increase the stroke in the Z direction. This paper combines the error sensitivity analysis with the identification to reduce the dimensionality of parameter identification. A sub-space identification based on the Levenberg-Marquardt algorithm is proposed, and the corresponding error model is established. The experiments are completed on the ABB IRB120 industrial robot. The experiment demonstrated that the sub-space identification method has better identification accuracy, and the full movement of all joints is conducive to improving the calibration effect. Compared with the conventional calibration method based on the DBB, the calibration effect of the proposed method is improved by 22.82%.