Kinematic parameter estimation technique for calibration and repeatability improvement of articulated arm coordinate measuring machines

The kinematic modeling of articulated arm coordinate measuring machines (AACMM) has inherited both the previous developments in the field of robot arms and manipulators, and their calibration and parameter identification techniques, given the similarity of their mechanical characteristics. The different accuracy and repeatability requirements of both systems make it necessary to consider different identification techniques covering the characteristic operational parameters in each case. This paper presents a new data capture technique for subsequent identification of an AACMM kinematic model parameters, using nominal data reached by a ball bar gauge, along with the algorithm and objective functions used, based on a new approach including terms regarding measurement accuracy and repeatability. Thus, an estimation and error measurement correction method using a repeatability error model based on Fourier polynomials is derived from the identification scheme. A Sterling series FARO arm with 1.5 m long and 6 degrees of freedom (dof) was used to carry out experimental tests in order to evaluate the efficiency of the techniques presented, showing improved accuracy and repeatability performance.