首页|大型工业机器人运动学标定及精度补偿方法

大型工业机器人运动学标定及精度补偿方法

扫码查看
将大型工业机器人引入航天制造领域,能够实现航天装备的高精度和可靠性.为了提高机器人绝对定位精度,更好地适应航天领域产品装配需求,基于改进的D-H(MD-H)模型,提出了大型工业机器人运动学标定及精度补偿方法.首先在误差建模中加入了坐标系转换误差和靶球安装误差,建立了更为完善的标定误差模型.然后通过激光跟踪仪对末端靶球进行测量来获得实际位姿数据,并采用改进的迭代最小二乘法辨识出误差参数.最后根据计算误差对机器人进行修正与补偿,完成运动学标定过程.为验证辨识方法的有效性,以KUKA KR1000 Titan大型工业机器人为研究对象进行仿真和实验.结果表明,经过标定补偿后的机器人末端平均定位误差由1.122mm降低到0.340mm,平均定位精度提升了约70%,补偿后机器人的绝对定位精度得到明显改善.
Kinematics Calibration of Large Industrial Robot and Precision Compensation Method
Introducing large industrial robots into aerospace manufacturing can achieve high precision and reliability of aero-space equipment.The kinematics calibration of large industrial robot and precision compensation method was suggested based on the modified D-H(MD-H)model to improve the absolute positioning accuracy,which can better adapt to the product assembly re-quirements of the aerospace field.There was a more complete error parameter identification model by adding the coordinate sys-tem transformation error and target ball installation error.Then the method using a laser tracker got the position of robot terminal by the target ball attached to the flange plate.The robot terminal error model was calculated by improved iterative least square method.Finally the compensation can be made to the robot terminal to complete the calibration.In order to further verify the va-lidity of the identification algorithm,KUKA KR1000 Titan industrial robot was used for simulation and experiment.The results illustrate that the average error decreased from 1.122mm to 0.340mm and the average positioning accuracy has increased by about 70%.The absolute positioning accuracy rate significantly improved with the process of calibration compensation.

Kinematics CalibrationMD-H ModelLarge Industrial RobotAbsolute Positioning AccuracyImp-roved L-S MethodError Compensation

李博文、张晓辉、何煦、罗敬

展开 >

中国科学院长春光学精密机械与物理研究所,吉林 长春 130033

中国科学院大学,北京 100049

运动学标定 MD-H模型 大型工业机器人 绝对定位精度 改进最小二乘法 位置误差补偿

国家自然科学基金国家自然科学基金

6187519012003033

2023

机械设计与制造
辽宁省机械研究院

机械设计与制造

CSTPCD北大核心
影响因子:0.511
ISSN:1001-3997
年,卷(期):2023.394(12)
  • 8