Title :
Competitive learning for Self Organizing Maps used in classification of partial discharge
Author :
Jaramillo-Vacio, Rubén ; Ochoa-Zezzatti, Alberto ; Rios-Lira, Armando ; Ponce, Julio
Author_Institution :
Lab. de Pruebas a Equipos y Mater. (LAPEM), Comision Fed. de Electr., Mexico
Abstract :
This paper presents some competitive learning algorithms for Self Organizing Map (SOM). The competitive learning algorithms showed to self organizing map algorithm are winner-takes-all, Frequency Sensitive Competitive Learning and Rival Penalized Competitive Learning. The result shows the performance in classification of partial discharge on power cables using SOM.
Keywords :
learning (artificial intelligence); partial discharges; power cables; power engineering computing; self-organising feature maps; SOM; competitive learning algorithms; frequency sensitive competitive learning; partial discharge classification; power cables; rival penalized competitive learning; self-organizing maps; winner-takes-all; Discharges (electric); Insulation; Partial discharge measurement; Partial discharges; Power cables; Training; Vectors;
Conference_Titel :
Electrical Communications and Computers (CONIELECOMP), 2012 22nd International Conference on
Conference_Location :
Cholula, Puebla
Print_ISBN :
978-1-4577-1326-2
DOI :
10.1109/CONIELECOMP.2012.6189910
