利用径向温度分布特征检测绝缘子污秽等级  

Insulator Pollution Level Detection Using Radial Temperature Distribution Features

作　　者：陈华林[1] 姚建刚[1] 李佐胜[2] 黄诗文 李文杰[1] 

机构地区：[1]湖南大学电气与信息工程学院,长沙410082 [2]湖南省电力公司超高压管理局,长沙410014 [3]湖南湖大华龙电气与信息技术有限公司,长沙410012

出　　处：《电力系统及其自动化学报》2010年第5期56-61,共6页Proceedings of the CSU-EPSA

基　　金：湖南省科技计划资助项目(2006GK3043)

摘　　要：为了实现绝缘子污秽状况的红外热像在线测量,提出一种基于径向温度分布特征的绝缘子污秽等级检测新方法。人工截取绝缘子图像区域,采用复小波广义交叉确认GCV阈值去噪方法抑制白噪声干扰,通过直方图包络线提取阈值分割图像,利用二值图像水平扫描采样绝缘子盘面边缘点,应用最小二乘拟合盘面边缘椭圆方程,获得椭圆内长轴以上图像区域,并提取径向温度分布特征表征绝缘子污秽程度,采用最近邻湿度条件下的最小距离分类器评定待检测绝缘子的污秽等级。试验结果验证了该方法的可行性和有效性。In order to on-line detect the contamination condition of insulator by infrared thermal imaging,a new method for insulator pollution level detection was proposed based on radial temperature distribution features.The image region of insulator was intercepted manually and denoised by complex wavelet generalized crossing validation（GCV） threshold method to suppress white-noise interference.According to the segmentation threshold extracted from the histogram envelope line,the edge points of the ceramic piece surface of insulator were sampled by binarizated horizontal scan of the segmented image,which is applied to fit the elliptic equation of disc surface edge by least square method.Finally,the ellipse image region above its long axis was abstracted,from which radial temperature distribution features were extracted to represent the pollution severity of the insulator.The insulator pollution level was evaluated by minimum distance classifier under the nearest humidity condition.Experiment results proved the feasibility and effectiveness of the method.

关 键 词：绝缘子污秽等级 红外热像 径向温度分布特征 图像分割 最小距离分类器 

分 类 号：TM85[电气工程—高电压与绝缘技术] TP18[自动化与计算机技术—控制理论与控制工程]