掠入射结构光测量机械研磨面平面度技术研究

发布时间：2018-05-15 02:03

本文选题：结构光 + 平面度　；参考：《中国科学院大学(中国科学院光电技术研究所)》2017年硕士论文

【摘要】：随着工业技术不断发展,工件精度要求也在不断提高。无论是依赖于手工研磨的传统加工技术,还是现代机器人研磨技术,在研磨工件之前,都需要充分获得被测面的整体和局部面形信息。传统平面度单点采样检测难以满足需求,伴随机器自动化水平的不断提高,自动测量工件表面面形信息并研磨已成为未来发展的方向,其主要思想为:通过快捷、可靠的检测方法获取被检测平面的三维信息,并对被检平面的平面度客观合理评估;在自动加工时其检测数据直接作为研磨的指导。本课题所选方案主要应用在机械零件高精度自动研磨在位检测领域。本课题针对450mm×450mm机械平面机器人自动研磨进行在位高精度检测,分辨力达到微米量级。通过掠入射线激光扫描、高精度气浮导轨位移、图像采集及处理实现整个面平面度高精度测量,从而指导机器人完成自动研磨。在未来,该技术结合拼接算法,可以实现大型机械面平面度检测,从而指导该类零件修研加工。本文主要包含以下关键技术:(1)掠入射放大技术;(2)结构光系统标定技术;(3)系统误差修正技术。在验证实验中,通过一字激光器-精密导轨位移实现对450mm×450mm的机械面的扫描,其效果等同于线阵结构光。数据处理过程基于小孔成像原理,以标定板投影条纹中心作为基准,对比被测区域条纹中心与基准中心的差异,通过标定后的放大倍率关系得到物空间相对高度值,依次映射得到三维信息,从而获取被测面坐标信息。在对检测面三维重建后,通过最小二乘拟合算法得到理想平面,并进一步得到被测面的面形信息,从而完成对被测面的平面度评估。作为对比实验,激光跟踪仪检测样品平面度0.177mm,本文方法结果为0.189mm,差值为0.012mm;通过数据插值,两种方法测量得到的高度点对点相减,均方根为0.014mm。从而验证了掠入射结构光检测方法的可行性。
[Abstract]:With the development of industrial technology, the requirement of workpiece precision is improving. Whether the traditional machining technology depends on manual grinding or the modern robot grinding technology, it is necessary to obtain the whole and local surface information of the measured surface before grinding the workpiece. The traditional single point sampling of flatness is difficult to meet the demand. With the continuous improvement of machine automation level, automatic measurement of workpiece surface profile information and grinding has become the direction of development in the future. The 3D information of the detected plane is obtained by a reliable detection method, and the flatness of the detected plane is evaluated objectively and reasonably, and the testing data is directly used as the guide of grinding in automatic machining. This project is mainly used in the field of high precision automatic grinding of mechanical parts. In this paper, the automatic grinding of 450mm 脳 450mm mechanical plane robot is tested with high accuracy in situ, and the resolution is in micron order. The high precision measurement of plane flatness is realized by sweep ray laser scanning, high precision displacement of air floatation guide rail and image acquisition and processing, so as to guide robot to finish automatic grinding. In the future, this technology combined with splicing algorithm can be used to detect the planeness of large mechanical surface and guide the machining of this kind of parts. This paper mainly includes the following key techniques: 1) grazing incident magnification technique 2) structural Light system Calibration technique / 3) system error Correction technique. In the verification experiment, the mechanical surface of 450mm 脳 450mm is scanned by one-word laser and precision guideway displacement, and the effect is equivalent to that of linear structured light. The process of data processing is based on the principle of pinhole imaging, taking the center of the projection fringe of the calibrated plate as the reference, comparing the difference between the center of the fringe and the center of the datum, the relative height of the object space is obtained by the magnification relation after calibration. The coordinate information of the measured plane can be obtained by mapping the 3D information in turn. After the 3D reconstruction of the detected surface, the ideal plane is obtained by the least square fitting algorithm, and the surface shape information of the measured plane is further obtained, thus the flatness evaluation of the measured plane is completed. As a contrast experiment, the plane of the sample detected by laser tracker is 0.177mm, the result of this method is 0.189mm, the difference is 0.012mm, and the height point-to-point subtraction obtained by the two methods is 0.014mm. Thus, the feasibility of the grazing incident structural light detection method is verified.
【学位授予单位】：中国科学院大学(中国科学院光电技术研究所)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【参考文献】