Identification of novel six-axis robot kinematic model with angular positioning deviations by R-Test   Tianhao Cui, Soichi Ibaraki       Abstract   Positioning accuracy is critically important for extending an industrial robot to offline programming applications such as robotic machining. Numerous previous works have presented the positioning error compensation for a six-axis robot based on its kinematic model with position and orientation errors of the rotary axis average lines, known as Denavit–Hartenberg (D-H) parameters. The paper presents its extension to rotary axes’ angular positioning deviations, and a scheme to indirectly identify them using the R-Test, a three-dimensional displacement probing system. The R-Test is performed at multiple locations, and a novel algorithm is presented considering that its exact position and orientation are unknown. Identified results and prediction accuracy are experimentally evaluated on a rectangular path covering a large working space and compared with the model identified by a tracking interferometer. The model identified by the RTest shows a satisfactory performance although its prediction error was larger than the model identified by using a tracking interferometer. Significantly lower instrumental cost is a strong advantage. Furthermore, uncertainty analysis is conducted, identification uncertainty is demonstrated by a distribution plot.   Keywords: Industrial robot, Error calibration, R-Test         >> Back to Home