Analysis of Insulator Leakage Current and Discharge State Classification Based on HHT Leakage Current and Discharge State Classification Based on HHT

Leakage current is a crucial parameter for online monitoring of polluted insulators and is closely related to pollution discharge. This study conducts artificial pollution discharge tests on porcelain insulators, utilizing the Hilbert-Huang Transform (HHT) to analyze the intrinsic mode function (IMF) components, Hilbert marginal spectrum, and time-frequency entropy of leakage current at various stages of pollution flashover. Fifteen characteristic quantities are extracted from the time-frequency domain and waveform details. By combining Principal Component Analysis (PCA) with a Least Squares Support Vector Machine (LS-SVM) classifier, the identification of pollution discharge states is carried out. The results indicate that the IMF components of the leakage current are more numerous during the initial discharge and flashover stages. The main components of the leakage current in the Hilbert marginal spectrum are distributed primarily in the 0-150 Hz and 200-250 Hz ranges. Moreover, the time-frequency entropy value of the leakage current before flashover is always greater than that after flashover. When the number of training samples is sufficient, the comprehensive judgment accuracy of the classifier exceeds 99%, which makes it feasible to use this classifier for identifying pollution discharge states. These findings provide a basis for establishing an insulator pollution flashover early warning system based on the Hilbert-Huang Transform.