Title :
A Principal Component Analysis Application for Flashover Recognition
Author :
Song, Young-Chul ; Han, Byung-Gil ; Kim, Min-Ho ; Choi, Doo-Hyun
Author_Institution :
Dept. of Ind. Technol., Purdue Univ., West Lafayette, IN
Abstract :
In this paper, a short-time PCA is used to flashover monitoring. At first, leakage current across a contaminated insulator is measured and digitized using specially designed hardware. PCA is applied to the leakage current of short-time window and the window is slid as time goes. We examine several principal components whether the components is useful to recognize flashover. Through experiments using real leakage data, three principal components are presented and investigated to find an algorithm to recognize flashover
Keywords :
flashover; insulator contamination; leakage currents; principal component analysis; flashover monitoring; flashover recognition; insulator contamination; leakage current; principal component analysis; Aging; Covariance matrix; Eigenvalues and eigenfunctions; Flashover; Insulation; Karhunen-Loeve transforms; Leakage current; Polymers; Principal component analysis; Signal processing algorithms; Flashover Recognition; Insulator; PCA(Principal Component Analysis); Variance;
Conference_Titel :
SICE-ICASE, 2006. International Joint Conference
Conference_Location :
Busan
Print_ISBN :
89-950038-4-7
Electronic_ISBN :
89-950038-5-5
DOI :
10.1109/SICE.2006.315738
