Research On Diagnosis Of Hydrophobicity Of Composite Ifisulators Based On Image Analysis

In the seriously dirty areas, composite insulators can overcome the inherentperformance defects of the traditional porcelain insulators and glass insulators exposedin use. Currently in the field of high-voltage power transmission composite insulatorshave had a large number of applications. The hydrophobic detection of compositeinsulator surface is the main means to distinguish the performance advantages anddisadvantages of composite insulators. The article will launch a special study on thehydrophobicity diagnosis method of composite insulator surface, which has greatsignificance for ensuring the safe operation of composite insulators.This paper firstly describes the formation mechanism and characteristics of thehydrophobicity of the composite insulators, and analyzes the advantages anddisadvantages of the recent detection technology of the hydrophobicity of compositeinsulators. Learning from the sprinkler grading and hydrophobic indicating functionmethod, the paper proposed an improved method that is a diagnostic method ofhydrophobicity of insulators based on image analysis, which will be able to beimproved and innovated in the practice of the process. This method can effectivelyimprove the effects of hydrophobicity diagnosis.When we got some hydrophobicity images, the surroundings have complexity andrandomness, which resulted in some fuzzy hydrophobicity images, and made the resultsof direct hydrophobicity diagnosis on the image less accurate. Taking into account thatthe traditional processing algorithms can not meet needs of the ambiguity of thehydrophobic image, we apply the improved image processing algorithms into theprocessing of the hydrophobic image, including image enhancing on hydrophobicimages based on local histogram equalization, median filtering based on self-adapting,threshold of genetic image segmentation algorithm based on maximum variance. Andby image recognition, we strike various kinds of typical condition codes associated withhydrophobicity images of composite insulators.Through the analysis of the experimental data, the paper pointed out theinadequacy of traditional classification method in hydrophobicity classification. Asself-adaptive and self-learning of neural network are strong, using classification methodbased on BP neural network can achieve objective judgment of hydrophobicityclassification, and this method can guarantee higher accurate classification and betterpracticality. Thus, a relatively complete automatic classification system ofhydrophobicity can be formed, which has important practical significance for thedetection of hydrophobicity of the surface of composite insulators. The system canpromptly discover the composite insulators that have flashover defects, and takeappropriate measures to ensure the safe and stable operation of high voltagetransmission lines.