An improved kinematic model for serial robot calibration based on local POE formula using position measurement

Purpose This paper aims to propose a calibration model for kinematic parameters identification of serial robot to improve its positioning accuracy, which only requires position measurement of the end-effector. Design/methodology/approach The proposed model is established based on local frame representation of the product of exponentials (local POE) formula, which integrates all kinematic errors into the twist coordinates errors; then they are identified with the tool frame' position deviations simultaneously by an iterative least squares algorithm. Findings To verify the effectiveness of the proposed method, extensive simulations and calibration experiments have been conducted on a 4DOF SCARA robot and a 5DOF drilling machine, respectively. The results indicate that the proposed model outperforms the existing model in convergence, accuracy, robustness and efficiency; fewer measurements are needed to gain an acceptable identification result. Practical implications This calibration method has been applied to a variable-radius circumferential drilling machine. The machine's positioning accuracy can be significantly improved from 11.153 initially to 0.301 mm, which is well in the tolerance (0.5 mm) for fastener hole drilling in aircraft assembly. Originality/value An accurate and efficient kinematic calibration model has been proposed, which satisfies the completeness, continuity and minimality requirements. Due to generality, this model can be widely used for serial robot kinematic calibration with any combination of revolute and prismatic joints.