CASIA OpenIR  > 毕业生  > 博士学位论文
Thesis Advisor谭民
Degree Grantor中国科学院大学
Place of Conferral中国科学院大学
Keyword中厚板焊缝 十宇线结构光 视觉传感器标定 焊缝特征检测 焊缝跟踪控制
Abstract      机器人在焊接行业己经使用了几十年。然而,大多数的焊接机器人仅仅应 用于简单的焊缝类型,如薄板的对接焊缝。在中厚板的焊接过程中,焊缝几何 形状比较复杂,焊缝检测和焊缝跟踪难以得到实现。本文主要研究基于视觉传 感系统的中厚板焊接机器人,主要包括三个部分,1)视觉传感系统标定;2) 焊缝特征检测;3)基于视觉传感系统的运动控制。本文主要研究内容如下:
     首先,提出了一种交叉线激光的标定技术。该方法的特点如下:1)只需 要一个普通的棋盘格,2)只需要两幅图像,3)可以实现交叉线结构光的两个 激光平面同时标定。标定结果具有较高的精度,可以应用于实际的生产制造, 以提高焊接质童。
      其次,提出了一种实时特征检测方法。使用十宇线结构光的交叉标记来精 确确定感兴趣E域。在所提出的方法中,模板可以n动创逑,并且沿着激光条 纹边缘进行匹配。实验结果表明,该方法较高的精度和较少的计算量,可以满 足实时检测系统。
      第三,提出了基于分层策略和改进卡尔蛙滤波的特征检测方法,提高实时 特征检测的稳定性。首先,用R.ANSAC算法粗略估计交叉结构光的十字标记, 然后,利用均值漂移算法精确修正十字标记的位罝。M后,使用马氏距离分类 器更新沅童卡尔蛙滤波器的状态。实验结果表明,该方法可以用于不同中厚板 的焊缝特征检测。
       第四,采用概率技术同时实现中厚板焊缝跟踪和移动焊接机器人定位。应 用扩展卡尔蛙滤波器进行焊缝跟踪控制,应用粒子滤波器进行机器人定位。实 验结果表明,本文所设计的机器人焊缝跟踪系统具有较高的准确性和稳定性。
Other Abstract
    Robots have been employed in the welding industry for many decades. However,
most of the welding robots are used for weld seams with simple shape such
as butt seams in thin plates. In thick plate welding, it is not easy to detect and
track the weld seam because of its complex geometry. This study focuses on the
development of thick plate welding robot using visual sensing system. The study
consists of three main parts, namely 1)visual sensing system calibration, 2)weld
seam features detection, and 3)motion control based on visual sensing system.
The core contents of the research are as follows.
      Firstly, we propose a particular calibration technique which is accurate, practical,
quick, and simple to calibrate a cross-line structured light. The highlights
of the method are 1)only a regular chessboard is used, 2)only two additional
images are needed, 3)the two planes of the cross-line structured light can be calibrated
concurrently. The accuracy of calibration is enough for the weld seam
tracking system. The results of this study are expected to be applied in real
manufacturing to improve the welding quality.
     Secondly, a real-time feature detection method is proposed. We use the cross
mark of the cross-line structured light as the pinpoint for designating the region
of interest(ROI). In the proposed method, templates can be created automatically,
and they are manipulated to match the edge-like locations. As a result
of the implementation, high accuracy and precision of detection are obtained.
Furthermore, the method can reduce a huge computation cost, and thus it is
adequate for a real-time detection system.
      Thirdly, a real-time and robust feature detection method using hierarchical
strategy and modi ed Kalman lter is proposed to stabilize the real-time features
detection. In the rst step, the cross mark of the cross-line structured light is
coarsely estimated by RANSAC algorithm, and then the location of the cross
mark is re ned by mean shift algorithm. In the second step, the scalar Kalman
lter is applied to the detection to improve the robustness. In the last step,
vi A Study on Thick Plate Welding Robots Using Visual Sensing System
we use Mahalanobis distance classi er to update the state of the scalar Kalman
lter. As a result of the implementation, the detection is achieved even though
the feature is translated to the di erent surface. Moreover, the low computational
cost in the proposed method meets requirements of real-time feature detection.
Thus, it satis es the requirement of the real-time weld seam tracking system.
      Fourthly, probabilistic techniques are used to track the thick plate weld
seam and localize the welding mobile concurrently. In particular, the extended
Kalman lter is employed in the weld seam tracking control; and the particle
lter is applied to the robot localization. As a result of the implementation, a
well-integrated control system is demonstrated. The accuracy and stability of
the proposed system are reliable. Moreover, the extended work of the study is
considered to be a promising one.
     Finally, all aspects of this study are summarized and discussed. Also, the
future work of the research is suggested.
Document Type学位论文
First Author AffilicationInstitute of Automation, Chinese Academy of Sciences
Recommended Citation
GB/T 7714
KIDDEE PRASARN. 基于视觉传感系统的中厚板焊接机器人研究[D]. 中国科学院大学. 中国科学院大学,2017.
Files in This Item:
File Name/Size DocType Version Access License
基于视觉传感系统的中厚板焊接机器人研究 (12176KB)学位论文 暂不开放CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
Baidu academic
Similar articles in Baidu academic
Bing Scholar
Similar articles in Bing Scholar
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.