Classification of Arc Fault in Sphere-Gap and Rod-Gap Using Stockwell Transform and Machine Learning Based Approach

Abstract

There are different types of arc may occur depending on the arcing conditions and involved surfaces. The severity of the arc is determined by the involved arcing surface and the arcing current flowing path. In this study, arc in Sphere-Gap and Rod-Gap surfaces is considered for the time-frequency domain analysis. The voltage characteristics for both the arc events are recorded in different voltage levels and gap length. A Stockwell Transform (ST) based approach is applied on the arc signals for the harmonic decomposition. Further, K-Nearest Neighbor (KNN) machine learning algorithm is applied on the ST coefficients for the classification of real-time arc signals of different arcing conditions. The results obtained using ST and KNN algorithm successfully classifies different arc faults due to rod-gap and sphere-gap by their harmonic signature.