A distance calibration method for kinematic parameters of serial industrial robots considering the accuracy of relative position. (30th November 2022)

• Record Type: Journal Article
• Authors: Guo, Yixuan; Jiang, Zhouxiang; Song, Bao; Tang, Xiaoqi; Min, Huasong; Fu, Chenglong
• Is Part Of: Measurement. Volume 204(2023)
• Publication Date: 2022-11-30
• Subjects: Industrial robot -- Kinematic calibration -- Relative position error -- Three-dimensional measurement system; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2022.111842
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: A non-redundant distance optimization function is established based on the redundancy analysis of the error parameters in the distance error model. The strategy for improving the positioning accuracy of relative position is proposed, which further improves the compensation effect of the distance error model on positioning errors. In the experiment, a prototype of the measurement system based on the R-test was developed and first successfully applied to the robot calibration. Comparison simulations and experiments prove the performance and practicality of the proposed method. Abstract: This paper proposes a kinematic calibration method based on the optimized distance error function to improve the accuracy of the relative positions of industrial robots. For implementation, a three-dimensional measurement system with R-test is designed to measure the robot end-effector's distance, which is the first successful use of this system for robot calibration. Then, an optimization function of the distance errors is proposed without an approximation condition to identify the controller's modifiable parameter errors. In addition, the influence of the redundant parameter on the error of relative position is analyzed, which indicates it can further improve the relative position's accuracy. Simulations prove the performance of the proposed solution by comparing it with the current distance and positioning error methods. The validation experiment shows the proposed method reduced the mean of distance error/distance and relative positioning error/relative positioning from 14.30 µm/mm to 4.4 µm/mm and from 20.15 µm/mm to 10.85 µm/mm.