航空学报 > 2019, Vol. 40 Issue (10): 422871-422871   doi: 10.7527/S1000-6893.2019.22871

一种适用于曲面结构的机器人制孔误差在线补偿技术

王龙飞, 张丽艳, 叶南   

  1. 南京航空航天大学 机电学院, 南京 210016
  • 收稿日期:2018-12-24 修回日期:2019-01-21 出版日期:2019-10-15 发布日期:2019-03-08
  • 通讯作者: 张丽艳 E-mail:zhangly@nuaa.edu.cn
  • 基金资助:
    民用飞机专项科研(MJ-2015-G-084);国家自然科学基金(51605222);江苏省基础研究计划(自然科学基金)(BK20160799)

An on-line compensation technology for robotic drilling error suitable for curved structure

WANG Longfei, ZHANG Liyan, YE Nan   

  1. College of Mechanical and Electrical Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China
  • Received:2018-12-24 Revised:2019-01-21 Online:2019-10-15 Published:2019-03-08
  • Supported by:
    Special Research Program on Civil Aircraft (MJ-2015-G-084); National Natural Science Foundation of China (51605222); Jiangsu Foundation Research Program (Natural Science Foundation of Jiangsu Province) (BK20160799)

摘要: 针对工业机器人应用于飞机零部件自动钻孔时各项误差累积造成制孔精度差的问题,提出一种利用单应关系计算机器人驱动坐标三维偏差,以在线补偿机器人制孔精度的方法。首先利用外部测量设备建立机器人制孔系统中各坐标系关系;在标定阶段,通过以一定倾斜角度固联于机器人末端的相机拍摄一幅安装于制孔工作平面上与刀轴正对的平面标定板图像,并据此完成基于单应变换的手-眼关系标定;在实际制孔过程中,机器人在测距传感器及相机的辅助下,从基准孔理论坐标对应的姿态,不断调整至基准孔正上方理想位置,通过手-眼关系计算基准孔实际位置对应的机器人驱动坐标,然后根据一组基准孔的机器人三维驱动误差,计算三维驱动误差变换矩阵,据此获得这组基准孔邻域范围内各待钻孔的机器人驱动坐标补偿量,从而实现待钻孔定位误差补偿。以飞机结构实验件为对象进行了模拟制孔验证,实验结果表明,补偿前待钻孔三维综合定位误差和法向误差测量值范围分别为2.28~2.85 mm和2.09°~3.93°,平均为2.55 mm和3.30°,补偿后制孔最大误差分别不超过0.30 mm和0.21°,满足自动制孔位置精度要求。

关键词: 工业机器人, 视觉辅助制造, 手-眼标定, 基准孔定位, 自动制孔, 在线误差补偿

Abstract: When industrial robots are used for automatic drilling of aircraft components, their accumulative errors inevitably result in poor positioning accuracy. An on-machine positioning error compensation method is proposed, utilizing the homographic hand-eye relation to measure 3D error of robot driving coordinates. Firstly, the geometric relations among the coordinate frames involved in the robot drilling system are established by using an external measurement device. In the calibration stage, a planar calibration target is specifically installed to coincide with the working plane, and the robot-carried camera, which mounted at an angle, takes one image of the calibration target that is in front of the drilling bit. The homographic relation between the vision image plane and the working plane is then calculated according to the image of the calibration target. In the drilling process, with the aid of range sensor and monocular, the robot adjusts its pose, which corresponding to the theoretical coordinates of the datum hole, to the ideal pose above the datum hole. The practical driving coordinates corresponding to datum holes are calculated by utilizing the calibrated hand-eye homography. And then the 3D transformational matrix is calculated based on a group of 3D driving errors of datum holes. The position errors of the holes to be drilled/riveted are then compensated based on the 3D transformational matrix. The proposed error compensation method is verified by stimulated drillings on an experimental aircraft structural part. The results show that the ranges of measured value of positioning and normal error of the drilling holes before compensation are 2.28-2.85 mm and 2.09°-3.93° respectively, and the average errors are 2.55 mm and 3.30° respectively. After compensation, the maximum error of the drilling errors is within 0.30 mm and 0.21°, respectively, meeting the requirement of aircraft structure drilling.

Key words: industrial robot, vision assisted manufacturing, hand-eye calibration, datum holes positioning, automatic drilling, on-line error compensation

中图分类号: