Título: Redes Neurais Aplicadas ao Diagnóstico de Falhas Incipientes em Transformadores de Potência Imersos em Óleos
Autores: Lima, Sanderson E. U. de; Reis, Laurinda Lucia N.; Oliveira, Juliana Carvalho; Coelho, Leandro dos Santos; Almeida, Otacílio M.
Resumo: Incipient fault in power transformer is closely related to isolation condition assessment. Dissolved gas analysis in transformer insulating oil is a well-known diagnostic technique. Produced gases serve as indicators of the type and severity of electrical stress. This paper presents a comparative study of multilayer perceptron and radial basis function neural networks for fault identification in oil-immersed power transformers. The neural networks are trained by using improved learning algorithms. The used data in experiments are taken out chromatographic analysis of COELCE (Electrical Company of Ceara) transformers. The simulation results indicate the potentialities of the tested neural networks for diagnostic in oil-immersed power transformers.
Palavras-chave: Neural Network; Artificial Intelligence; Fault Diagnosis; Power Transformer; Dissolved Gas-in-oil Analysis (DGA)
Código DOI: 10.21528/CBRN2005-202
Artigo em PDF: CBRN2005_202.pdf
Arquivo BibTex: CBRN2005_202.bib