Prediction of distribution transformer no-load losses using the learning vector quantization neural network

Hatziargyriou Nikos, D; Georgilakis Paul, S; Paparigas Dimitrios, G; Bakopoulos John, A

dc.contributor.author	Hatziargyriou Nikos, D	en
dc.contributor.author	Georgilakis Paul, S	en
dc.contributor.author	Paparigas Dimitrios, G	en
dc.contributor.author	Bakopoulos John, A	en
dc.date.accessioned	2014-03-01T02:41:33Z
dc.date.available	2014-03-01T02:41:33Z
dc.date.issued	1998	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/30529
dc.subject	Artificial Neural Network	en
dc.subject	Learning Vector Quantization	en
dc.subject	Quality Control	en
dc.subject	Success Rate	en
dc.subject	Neural Network	en
dc.subject	Product Line	en
dc.subject.other	Electric losses	en
dc.subject.other	Electric transformer loads	en
dc.subject.other	Learning algorithms	en
dc.subject.other	Magnetic cores	en
dc.subject.other	Vector quantization	en
dc.subject.other	Distribution transformer no load losses	en
dc.subject.other	Learning vector quantization	en
dc.subject.other	Neural networks	en
dc.title	Prediction of distribution transformer no-load losses using the learning vector quantization neural network	en
heal.type	conferenceItem	en
heal.identifier.primary	10.1109/MELCON.1998.699420	en
heal.identifier.secondary	http://dx.doi.org/10.1109/MELCON.1998.699420	en
heal.publicationDate	1998	en
heal.abstract	This paper presents an artificial neural network (ANN) approach to classification of distribution transformer no-load losses. The Learning Vector Quantization (LVQ) neural network architecture is applied for that purpose. The ANN is trained to learn the relationship among data obtained from previous completed transformer constructions. For the creation of the training and testing set actual industrial measurements are used. Data comprise grain oriented steel electrical characteristics, cores constructional parameters, quality control measurements of cores production line and transformers assembly line measurements. It is shown that ANNs are very suitable for this application since they present classification success rates between 78% and 96% for all the situations examined.	en
heal.publisher	IEEE, Piscataway, NJ, United States	en
heal.journalName	Proceedings of the Mediterranean Electrotechnical Conference - MELECON	en
dc.identifier.doi	10.1109/MELCON.1998.699420	en
dc.identifier.volume	2	en
dc.identifier.spage	1180	en
dc.identifier.epage	1184	en