基于机器视觉的锅炉关键管道宏观位移检测方法研究及应用

发布时间：2018-09-11 10:11

【摘要】：锅炉蒸汽管道的蠕变和宏观位移是锅炉安全运行需要监测的重要方面,上述位移的实时监测可以反映锅炉蒸汽管道内部的应力应变情况。为了能够实时在线准确获取管道宏观位移,本文从视觉测量技术研究出发,利用支吊架与管道的结构关系,提出一种基于视觉的锅炉关键管道的宏观位移以及蠕变情况的测量方法,并自主设计一套应用此方法的在线检测系统。提出锅炉蒸汽管道蠕变情况及宏观位移检测方法,该方法通过在弹簧支吊架下节点焊两个标志圆,以标志圆圆心为标志点,根据标志圆圆心的三维空间坐标,拟合弹簧支吊架下节所在直线的直线方程,联立蒸汽管道的圆柱体曲线方程,得到二者的交点,根据该交点坐标的变化量即可得出该点宏观位移,在此基础上,进一步结合多个支吊架,将支吊架与管道的交点转换到同一个坐标系中,则可根据彼此交点距离的变化量,得到在两个或多个支吊架之间管道的蠕变情况。为了实时准确获得标志圆圆心的三维坐标,采用了标志圆视觉检测算法,即先对图像进行灰度化及灰度直方图分析,接着进行图像去噪处理,之后对图像进行二值化,并通过边缘特征提取标志圆的轮廓,用Hough变换检测标志圆,获得标志圆圆心的像素坐标,最后根据针孔成像模型得到标志圆圆心的三维坐标。实验验证了该方法的可行性。针对强光照、标志圆表面积灰等复杂环境,常规方法检测误差大甚至无法检测的现象,提出基于HSV模型的图像处理算法,该方法通过对HSV中H分量、S分量和V分量进行的一系列数学组合运算和色调、饱和度及亮度配比系数的调整,可有效解决强光照,表面积灰等情况,仿真实验结果证明了该方法的有效性。设计了一套锅炉管道宏观位移在线检测系统,运行于珠海电厂1号机组。软件系统采用VC在VS2013开发环境下进行开发,试验结果表明该系统能够全天候对锅炉管道宏观位移进行实时在线检测,精度高,该方法的成功应用为锅炉蒸汽管道失效概率和剩余寿命计算模型的建立奠定了基础。
[Abstract]:The creep and macroscopic displacement of boiler steam pipeline are important aspects of boiler safety operation. The real-time monitoring of the above displacement can reflect the stress and strain situation in boiler steam pipeline. In order to accurately obtain the macroscopic displacement of pipeline in real time and online, this paper, based on the research of visual measurement technology, makes use of the structural relationship between the support hanger and the pipeline. This paper presents a vision-based measurement method for macroscopic displacement and creep of boiler critical pipes, and designs an on-line detection system using this method. The creep condition and macroscopic displacement detection method of boiler steam pipeline are put forward. By spot welding two mark circles under the spring support and hanger, the method takes the mark circle center as the mark point, and according to the three dimensional space coordinate of the mark circle center, By fitting the straight line equation of the lower node of the spring support hanger and the cylinder curve equation of the steam pipeline, the intersection point of the two points can be obtained. According to the change of the coordinate of the intersection point, the macroscopic displacement of the point can be obtained. The creep of the pipe between two or more supports and hangers can be obtained according to the variation of the distance between the points of intersection between the two or more supports and hangers by converting the intersection points of the support and hanger to the same coordinate system. In order to get the 3D coordinate of the mark circle center in real time, this paper adopts the sign circle vision detection algorithm, that is, the image is first grayscale and grayscale histogram analysis, then the image is de-noised, and then the image is binarized. The contour of mark circle is extracted by edge feature, and the pixel coordinate of mark circle is obtained by Hough transform. Finally, the 3D coordinate of mark circle center is obtained according to the pinhole imaging model. The feasibility of the method is verified by experiments. A new image processing algorithm based on HSV model is proposed in view of the complex environment such as strong illumination, marked circle surface area ash and so on, in which the detection error is large or even undetectable by conventional methods. This method can effectively solve the problems of strong illumination, surface area ash and so on by a series of mathematical combination operations and adjustment of hue, saturation and luminance ratio coefficient of H component S component and V component in HSV. The simulation results show that the method is effective. A set of boiler pipeline macro displacement measurement system is designed, which runs in No. 1 unit of Zhuhai Power Plant. The software system is developed by using VC in the VS2013 development environment. The experimental results show that the system can detect the macroscopic displacement of boiler pipeline in real time and online all the time, and the precision is high. The successful application of this method has laid a foundation for the establishment of failure probability and residual life model of boiler steam pipeline.
【学位授予单位】：长沙理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.41;TK228

【参考文献】