摘要
提出了一种基于机器人几何参数误差与基坐标系位姿误差的六轴串联型机器人误差标定方法。首先基于MD-H方法建立了IRB6700机器人几何参数误差模型,得到机器人连杆几何参数误差到机器人末端位姿误差的映射关系;然后进一步考虑了基坐标系的位姿误差,并建立了考虑基坐标系误差的机器人误差模型;此外,针对传统标定方法操作繁琐、偶然误差大等问题,提出了一种改进的参数辨识方法,进一步提高了标定过程的可操作性与标定精度。最后,开展了本体标定实验与基坐标系误差扰动实验,结果表明所提方法将机器人位置误差的平均值由3.1928 mm减小至0.1756 mm,位置误差的标准差由0.5494 mm下降到0.0830 mm。基坐标系位姿误差扰动前后参数辨识的结果一致性高于99%,标定精度及稳定性得到显著提高。
A six-axis serial robot error calibration method was proposed based on the robot geometric parameter errors and the base frame position and attitude errors.Firstly,an geometric parameter error model of the IRB6700 robot was established based on the MD-H method,and the mapping relationship between the geometric parameter errors of the robot connecting rods and the end pose errors of the robot was obtained.Then,the position and attitude errors of the base frame were further considered,and the robot error model was established considering the errors of the base frame.In addition,to address the cumbersome operation and large accidental errors of traditional calibration methods,an improved parameter identification method was proposed to further improve the operability and calibration accuracy of the calibration processes.Finally,the body calibration experiments and the base frame error disturbance experiments were carried out.The results show that the proposed method decreases the average position error of the robot from 3.1928 mm to 0.1756 mm,and the standard deviation of the position errors decreases from 0.5494 mm to 0.0830 mm.Moreover,the consistency of the parameter identification results under base frame error disturbance is higher than 99%,and the calibration accuracy and stability are improved.
作者
倪华康
杨泽源
杨一帆
陈新渡
严思杰
丁汉
NI Huakang;YANG Zeyuan;YANG Yifan;CHEN Xindu;YAN Sijie;DING Han(Jiangsu Jitri-HUST Intelligent Equipment Technology Co.,Ltd.,Wuxi,Jiangsu,214100;School of Mechanical Science and Engineering,Huazhong University of Science and Technology,Wuhan,430074)
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2022年第6期647-655,共9页
China Mechanical Engineering
基金
国家重点研发计划重点专项(2019YFA0706703)
国家自然科学基金(52075204)。
关键词
工业机器人
运动学标定
误差模型
基坐标系位姿误差
industrial robot
kinematics calibration
error model
base frame position and attitude error
