Study and Analysis of Various Partial Discharge Signals Classification Using Machine Learning Application

Partial discharges (PD) are the major reason for the failure of high-voltage equipment. Detection of partial discharges is very important to check the health condition of its insulation and the severity level is detected by the classification of different PD signals. In this work, 4 different types of PD signals are collected in the high-voltage laboratory. The collected signals are also de-noised using wavelet transform and then proceed for classification. For the classification study various machine learning algorithms artificial neural network (ANN), support vector machine (SVM), and random forest (RF) have been used. Different statistical features (skewness, kurtosis, standard deviation, variance, etc.) are extracted which are used to train the ML algorithms. In this work, a total of 800 PD signals are collected which are used for training, validation, and testing of the ML model. The model performance and its’ comparison study are evaluated based on accuracy, precision, recall, etc.