A classification of the leakage current pulses of high voltage contaminated porcelain insulators using neural network a self-organized neural network

Tsarabaris, PT; Karagiannopoulos, CG; Bourkas, PD; Theodorou, NJ

dc.contributor.author	Tsarabaris, PT	en
dc.contributor.author	Karagiannopoulos, CG	en
dc.contributor.author	Bourkas, PD	en
dc.contributor.author	Theodorou, NJ	en
dc.date.accessioned	2014-03-01T01:55:30Z
dc.date.available	2014-03-01T01:55:30Z
dc.date.issued	2006	en
dc.identifier.issn	1472-8915	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/27761
dc.subject	high voltage insulators	en
dc.subject	leakage current	en
dc.subject	self-organized neural networks	en
dc.subject.classification	Computer Science, Artificial Intelligence	en
dc.subject.classification	Engineering, Electrical & Electronic	en
dc.subject.other	FLASHOVER	en
dc.subject.other	SURFACES	en
dc.subject.other	DIELECTRICS	en
dc.subject.other	INCEPTION	en
dc.subject.other	FOG	en
dc.title	A classification of the leakage current pulses of high voltage contaminated porcelain insulators using neural network a self-organized neural network	en
heal.type	journalArticle	en
heal.language	English	en
heal.publicationDate	2006	en
heal.abstract	This paper proposes a classification system for leakage current values on high voltage porcelain insulators, and presents the results, which derived. The proposed system can be divided into two parts: the first part consists of a high sampling rate and high precision measuring system and the second part consists of a self-organized neural network, which is responsible for the classification of the leakage current patterns. The analysis of the results showed that current values, which are measured, could be divided into three categories. The first category consists of low amplitude (less than 3.62 mA) leakage current values, which are located in a rather linear region of the leakage current waveform. Current values of the second and third category have medium (from 3.62 mA to 10.21 mA) and high (greater than 10.21 mA) amplitudes respectively, and belong to a strictly non-linear area of the waveform. Those values constitute short and long width pulses and are attributed to the presence of partial discharges and arcs on the insulator surface. The results of this work are being analyzed and physical interpretations on the above classification are attempted. The system can probably be helpful in the further investigation of the phenomena observed on contaminated industrial insulators as well as an alarm device for heavily polluted insulators, which are in service.	en
heal.publisher	C R L PUBLISHING LTD	en
heal.journalName	ENGINEERING INTELLIGENT SYSTEMS FOR ELECTRICAL ENGINEERING AND COMMUNICATIONS	en
dc.identifier.isi	ISI:000238062000004	en
dc.identifier.volume	14	en
dc.identifier.issue	1	en
dc.identifier.spage	39	en
dc.identifier.epage	45	en