Abstract: The accuracy of five-axis hybrid robots is one of the primary challenges they face in the field of precision machining and manufacturing. Due to the complexity and diversity of error propagation paths within its structure, constructing an accurate overall error model is particularly challenging. To address this issue, this study proposes a kinematic calibration method for five-axis hybrid robots. First, decompose the five-axis hybrid robot into a tool motion chain and a workpiece motion chain to reduce interference and error propagation between them. Next, a closed-loop vector equation is constructed for the tool motion chain, which includes a parallel mechanism, and the error model is derived through differential perturbation. By introducing an auxiliary measurement device to perform full pose measurements on the end-effector of the parallel robot, it is noted that the anisotropy of the end-effector's position/orientation errors leads to inaccurate identification results. To resolve this issue, a weighted matrix is introduced to establish an identification equation with isotropic residuals, thereby improving identification accuracy. Subsequently, error analysis and modeling are performed for the workpiece motion chain, and the error sources are decomposed into motion errors and zero-point errors, which are compensated through the calibration and tool alignment processes, respectively. Finally, the effectiveness of the proposed calibration algorithm is validated through both calibration and test workpiece machining experiments. The results show that by calibrating the tool and workpiece motion chains, the proposed method can significantly improve the robot’s motion accuracy and stability, thus meeting the high-precision machining requirements in the manufacturing field.

Key words: five-axis hybrid robot, error modeling, kinematic calibration, error identification