Partial discharge classification on high voltage equipment with K-means

Abstract

This document proposes a statistical classification model using k-means cluster analysis to classify partial discharge (PD) patterns into four categories listed as corona at high voltage side in air, corona at low voltage side in air, surface in air, and internal discharge. The independent variables in this k-means cluster analysis model are skewness, kurtosis, asymmetry, and cross correlation following the φ - q - n PD patterns obtained from the fingerprint analysis which is a digital signal processing technique for PD measurement. The experiments were set to simulate all four PD patterns to obtain statistical parameters into 10 independent variables from the fingerprint analysis. This document recommends using 8 final clusters instead of using 4 final clusters to achieve the accuracy of using this k-means cluster approach to 100 percent. © 2006 IEEE.