Kinematic Calibration for Serial Robots Based on a Vector Inner Product Error Model

Abstract

The positioning accuracy of articulated serial robots in the workpiece coordinate system (WCS) is vital for practical applications, as command positions or planned paths are typically defined in WCS. However, conventional error models for kinematic calibration primarily focus on positioning accuracy in the base coordinate system (BCS), without adequately fulfilling the accuracy requirements of WCS. To enhance the positioning accuracy in WCS, we propose a novel error model based on the vector inner product, the calibration accuracy of which is independent of coordinate systems. The vector inner product error is constructed from positioning error, and a mapping model correlating with kinematic parameter errors is derived. Following the establishment of the model, the kinematic parameters are identified. The results of experiments using a CS612 robot reveal a 36% and 15% improvement in the positioning accuracy within WCS after calibration, over the existing distance error model and position error model. Finally, a method for compensating the positioning error in WCS is presented, and a deburring application case study demonstrates that the proposed model reduces the maximum positioning error of the deburring path from 1.270 to 0.207 mm after compensation, outperforming the distance error model and position error model by 61% and 31%, respectively.