Fuzzy Clustering Analysis for PD Signals in GIS Based on the Validity Index

Cluster validity index algorithm, which can find the number of clusters in a given object set, plays an important role in clustering analysis. There have been many proposals of cluster validity index, especially for fuzzy clustering, and many of them are dependent on clustering algorithms that can use the different interpretations of similarities between objects, usually in the geometric interpretation of objects. We introduce the Gustafson-Kessel (GK) and Fuzzy C-mean (FCM) clustering algorithms in this paper to separate the four different partial discharge (PD) defects in gas insulated switchgear (GIS), according to the parameters of Skewness (Sk), Kurtosis (Ku), number of peaks (Pe), cross-correlation factor (CC) and the discharge factor Q. Copyright (C) 2010 Praise Worthy Prize S.r.l. - All rights reserved.