A Novel Method for Visualizing the Pollution Distribution of Insulators

Knowing more about insulator pollution is of guiding significance to the development of anti-pollution strategies. Traditional methods can be used to determine only the pollution degree of insulators. However, the pollution distribution also has a substantial influence on the flashover voltage of insulators even under the same equivalent salt deposit densities (ESDD). In this article, an ESDD spectral density method and a visualization method based on hyperspectral technology are proposed. The proposed method presents the ESDD distribution using hyperspectral images. A detection model based on characteristic wavelengths is established with partial least squares regression to detect the insulator ESDDs. Based on the ESDD detection results, the features of the ESDD spectral density process, including the building of the ESDD probability graph, the overall ESDD, and visualization of the ESDD, are presented. Example analyses of field-operating insulators were conducted, revealing that the ESDD spectral density can be used to track the pollution distribution of insulators and that the pollution degree can be calculated from the overall ESDD. The error of the overall ESDD calculated by this method is within 10%. In addition, visualization of the pollution distribution can more intuitively record the degree of pollution and the nonuniform distribution characteristics. Both the pollution degree and distribution can be obtained by this method. Therefore, hyperspectral technology can overcome the limitations of traditional methods and convert the ESDD results into an image to determine the ESDD distribution. These research results provide theoretical instruction for guiding insulation design and anti-pollution strategies for operating transmission lines.