Diagnosis of Operating Insulator Aging Status Based on Image Processing and Hyperspectral Analysis

In this paper, an optical imaging detection method of aging status diagnosis for operating composite insulators is proposed, which uses image processing as well as hyperspectral imaging technology. Firstly, slices of the silicone rubber insulator were made to execute acid aging experiment. Hyperspectral imaging was used to shoot umbrella skirt slices with varying degrees of acid aging status. Secondly, hyperspectral analysis was performed to obtain reflected spectral characteristics at 400nm to 1000nm. Next, Principle Component Analysis (PCA) was performed to reconstructed the grayscale images. Finally, the reconstructed grayscale image was transformed linearly and rendered to obtain visual detection results and diagnostic criteria. Laboratory verification experiments show that the accuracy rate was 85%. It means that the aging status and characteristic spectrum of insulators could be connected with each other, which has great potential of engineering application.