首页|基于X射线数字成像的GIS设备缺陷无损检测方法

基于X射线数字成像的GIS设备缺陷无损检测方法

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
GIS设备的安全性和可靠性对电力体系的平稳运行具有重要意义;因此，为提高对GIS设备缺陷的检测效果、提高设备运行的安全性，在X射线数字成像的基础上，提出一种针对GIS设备的缺陷无损检测方法;通过X射线数字成像的方式采集GIS设备图像，并对图像中存在的泊松噪声实施去噪处理，以提高图像质量;针对处理后的图像，利用二维主成分分析法，通过将复杂的图像数据转换为简单的主成分来表示原始数据，提取出最具代表性的特征;将提取结果输入到BP神经网络分类器中，通过特征分类完成对GIS设备缺陷的无损检测;实验结果表明:应用该方法后，图像识别清晰度较高，对不同类型缺陷的检测效果良好;该方法的优势在于使用先进的图像处理和机器学习技术，能够有效地识别和定位GIS设备中存在的缺陷，通过及时发现并修复这些缺陷，可以提高GIS设备的安全性和可靠性，从而确保电力系统的平稳运行。
NDT Method for GIS Equipment Defects Based on X-ray Digital Imaging
The safety and reliability of gas insulated switchgear(GIS)equipment are of great significance for the smooth operation of power systems.Therefore,in order to improve the detection effect of GIS equipment defects and improve the safety of equipment operation,a non-destructive testing(NDT)method for GIS equipment defects is proposed based on X-ray digital imaging.The meth-od collects the GIS equipment images through X-ray digital imaging,and denoises the Poisson noise in the images to improve the image quality.For the processed images,two-dimensional principal component analysis is used to represent the original data by converting complex image data into simple main components,and extract the most representative features.The extracted results are input into the BP neural network classifier,and the non-destructive testing of GIS equipment defects is achieved through the feature classifica-tion.The experimental results show that after applying this method,it has a high image recognition clarity and good detection effect for different types of defects.This method has the advantages of the advanced image processing and machine learning techniques,which can effectively identify and locate the defects in the GIS equipment.By discovering and repairing these defects in time,it can improve the safety and reliability of GIS equipment,thereby ensuring the smooth operation of power systems.

X-ray digital imagingGIS equipmentNDT of defectsPoisson noiseblind source separation and de-noisingtwo dimensional main component analysisfeature extractionback propagation(BP)neural network classifier

张志刚、张岩、吴文平、马贵荣

展开 >

国能朔黄铁路发展有限责任公司,山西原平 034100

西南交通大学电气工程学院,成都 610031

X射线数字成像 GIS设备 缺陷无损检测 泊松噪声 盲源分离去噪 二维主成分分析法 特征提取 BP神经网络分类器

朔黄铁路公司科技创新项目

SHYP-22-12

2024

10.16526/j.cnki.11-4762/tp.2024.06.006

计算机测量与控制
中国计算机自动测量与控制技术协会

计算机测量与控制

CSTPCD
影响因子：0.546
ISSN：1671-4598
年,卷(期)：2024.32(6)