Dual quaternion-based kinematic modeling for decoupling identification of geometric errors of rotary axes in five-axis platforms

Abstract

Identification of the geometric errors of rotary axes is important for deciding the machining accuracy of five-axis platforms. In this paper, a programmable identification method is proposed to decouple the position-dependent and position-independent geometric errors (PDGEs and PIGEs) of the rotation axis. First, the kinematics model of the five-axis platform is developed by dual quaternions theory, and then the physical significance of left multiplication and right multiplication of PDGEs error model is analyzed. The motion difference between the base coordinate system and its local coordinate system is compared by simulation analysis. The geometric relationship between rotation direction and motion elements is established based on the error model, which is used to design an optimization algorithm for identifying the orientation errors of PIGEs. Additionally, the effect of linear axis PIGEs is eliminated by equivalent replacement, and a sample regression model is developed to identify the offset errors of PIGEs. The PDGEs are separated considering the Motion First principle. Finally, the proposed model and identification method were verified on a five-axis dispensing platform.