摘要
复合绝缘子憎水性状况对于电力系统的安全稳定运行具有重要影响。为了准确、快捷、方便地识别复合绝缘子憎水性,借助MATLAB图像处理工具和GUI图形用户界面功能开发了一款复合绝缘子憎水性分析软件。该软件开发运用了模块化思想,建立了图像处理、水珠特征量提取、憎水性检测方法和数据库管理等功能模块;采用改进的Canny算子对图像进行分割,提取了水珠(或水迹)的特征;最后利用改进径向基函数(radial basis function,RBF)神经网络对绝缘子憎水性进行分级。实验结果表明,该软件能够准确、客观地识别绝缘子憎水性等级,准确识别率高达92%。避免了人为因素的影响,提高了工作效率,为现场运行人员提供了数据支持和参考。
The hydrophobic condition of composite insulators has an important influence on the security and stability of power system. In order to identify hydrophobicity of composite insulators accurately,fast and conveniently,with the tools of MATLAB image processing and GUI developing composite insulators hydrophobicity analysis software. The software was developed using a modular thinking,the establishment of image processing,the amount of drops of water feature extraction,hydrophobicity detection methods and database management modules. The improved Canny operator was used for image segmentation,extraction characteristics of water drop（ or water stains）,finally the improved RBF neural network was used to analyse the insulators hydrophobicity grade. The results of experiment show that the software was able to identify insulators hydrophobicity grade accurately and objectively and its accurate recognition rate is as high as 92%. The software can avoid the influence of artificial factors,improve the work efficiency and provide data support and reference for field operation personnel.
作者
闫康
YAN Kang(State Grid Luquan Power Supply Company,Shijiazhuang 050200,China)
出处
《电力工程技术》
2018年第5期126-131,共6页
Electric Power Engineering Technology
