Desenvolvimento de sistema para detecção de falhas em rolamentos de motores de indução a partir de seus sinais de vibração
This work brings a study and development of an approach able to identify bearing induction motor faults from the vibration signals of these machines during their operation. Vibration signals from Case Western Reserve University Bearing Data Center are used, which include motors with healthy or fault...
| Autor: | |
|---|---|
| Formato: | tesis de maestría |
| Estado: | Versión publicada |
| Fecha de publicación: | 2022 |
| País: | Brasil |
| Recursos: | Universidade Tecnológica Federal do Paraná (UTFPR) |
| Repositorio: | Repositório Institucional da UTFPR (da Universidade Tecnológica Federal do Paraná (RIUT)) |
| Idioma: | portugués |
| OAI Identifier: | oai:repositorio.utfpr.edu.br:1/30182 |
| Acesso em linha: | http://repositorio.utfpr.edu.br/jspui/handle/1/30182 |
| Access Level: | acceso abierto |
| Palavra-chave: | Motores elétricos de indução Redes Neurais (Computação) Localização de falhas (Engenharia) Electric motors, Induction Neural networks (Computer science) Fault location (Engineering) CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA Engenharia Elétrica |
| Resumo: | This work brings a study and development of an approach able to identify bearing induction motor faults from the vibration signals of these machines during their operation. Vibration signals from Case Western Reserve University Bearing Data Center are used, which include motors with healthy or faulty bearings in many situations of load and damage levels. The preprocessing tool Hilbert-Huang Transform is used, followed by extraction features in time and time-frequency domains. Having the purpose to identify and classify patterns from the extracted features, it is used an artificial neural network Multilayer Perceptron in order to predict the condition of bearings and the type of fault that happens to them. The results are measured between healthy and faulty bearings and the kind of mishappen that may be on this case, presenting 100% of accuracy in the classification of healthy and faulty patterns, 58,97% between all bearing faults, 78,57% for only drive end bearing faults, and 72,73% for fan end bearing faults. |
|---|