Partial Discharge Source Classification using Time-Frequency Transformation

UHF signals generated by partial discharges due to corona activity, surface discharge and particle movement in transformer oil were acquired and used to develop a classification model. Time-Frequency transformation of UHF signal emitted by partial discharge source was used for classification of the discharges using quadratic support vector machine (SVM) learning tool. The method was validated by simulating the field condition, by studying the classification model in an oil filled closed tank along with the pressboard along the path of UHF signal, as observed in real system.