Competitive Learning for Self Organizing Maps used in Classification of Partial Discharge |
||
Publicación: |
Programación Matemática y Software |
PDF (294 KB) |
Ruben Jaramillo-Vacio, Carlos Alberto Ochoa Ortiz Zezzatti, Julio Cesar Ponce Gallegos |
Comisión Federal de Electricidad-Laboratorio de Pruebas a Equipos y Materiales (LAPEM) |
Recibido: 8 de septiembre de 2011 Aceptado: 22 de noviembre de 2012 Publicado en línea: 6 de noviembre de 2013 |
Abstract. This paper different competitive learning algorithms for Self Organizing Map (SOM) are experimentally examined, the characterization of the obtainable results in terms of quality of SOM. The competitive learning algorithms showed to SOM algorithm are Winner-takes-all, Frequency Sensitive Competitive Learning and Rival Penalized Competitive Learning. As a case study: the performance in classification of partial discharge on power cables. |
Keywords: Competitive learning, Self Organizing Maps, Partial Discharge, Quality Measurements, Diagnosis. |
Rubén Jaramillo-Vacio(Autor de correspondencia) |
Email:ruben.jaramillo@cfe.gob.mx |