Application of Machine Learning in Discharge Classification

Condition monitoring of electric insulators is essential to maintain reliable power transmission and distribution. This paper presents a study to use a commercial acoustic sensor along with different machine learning algorithms to classify different types of discharges in outdoor electric insulation systems. Support vector machine (SVM) based classifiers have been utilized to distinguish between five common electrical discharges that were generated under controlled conditions. The classification problem expands to include outdoor ceramic insulators with three defects i.e., a crack in the ceramic disc, surface pollution discharge, and corona near the insulator surface. For both controlled samples and ceramic insulators, an average recognition rate of more than 91% has been achieved.