Detecção de falhas de estator em motores de indução trifásicos utilizando transformada wavelet, medida RMS e potência de previsão

• Ano de defesa: 2015
• Autor(a) principal: Vitor, Avyner Lorran de Oliveira
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Tecnológica Federal do Paraná; Cornelio Procopio; Brasil; Programa de Pós-Graduação em Engenharia Elétrica; UTFPR
• Palavras-chave em Português: Motores elétricos de indução; Entropia; Redes neurais (Computação); Electric motors, Induction; Entropy; Neural networks (Computer science); CNPQ::ENGENHARIAS; Engenharia Elétrica
• Link de acesso: http://repositorio.utfpr.edu.br/jspui/handle/1/3194

The Three-Phase Induction Motor is the most utilized electrical machine in industrial envi- ronments. It is an equipment that shows robust construction, ease maintenance, low cost and reliability. However, frequently, machines are exposed to thermal, electrical and mecha- nical efforts, which, over time, result in a failure. This work aims to analize the stator short circuit fault, one of the most frequent induction motor failures. Therefor, the peculiarities of the signals that are reflected in stator current are investigated, in order to correlate the signal characteristics to the failure in question. Current signals are processed through the Fourier and Wavelet transform in order to analyse short-circuit specific frequencies. Subse- quently, the classification is performed through two artificial neural networks, the multilayer perceptron and the Radial Basis Function.