Automatic Classification of mechanical vibration patterns in rotating machinery affected by unbalance

In this paper, we present an algorithm for automatic classification of vibration patterns on rotating machinery affected by unbalance from spectral analysis. We developed this algorithm using case-based reasoning and various descriptors. The raised descriptors were: The root mean square value (RMS),...

Full description

Autores:
Tipo de recurso:
Article of journal
Fecha de publicación:
2013
Institución:
Universidad Antonio Nariño
Repositorio:
Repositorio UAN
Idioma:
spa
OAI Identifier:
oai:repositorio.uan.edu.co:123456789/10437
Acceso en línea:
https://revistas.uan.edu.co/index.php/ingeuan/article/view/361
https://repositorio.uan.edu.co/handle/123456789/10437
Palabra clave:
Análisis vibracional
reconocimiento de patrones
descriptores de falla en maquina rotativa
espectro de Fourier
Vibrational analysis
pattern recognition
failure descriptors on rotating machine
Fourier spectrum
Rights
License
https://creativecommons.org/licenses/by-nc-sa/4.0
Description
Summary:In this paper, we present an algorithm for automatic classification of vibration patterns on rotating machinery affected by unbalance from spectral analysis. We developed this algorithm using case-based reasoning and various descriptors. The raised descriptors were: The root mean square value (RMS), the energy of Fourier spectra, the Higher Order frequency moments and the maximum value of the Fourier spectra. The job was to induce imbalance to a universal motor, taking the vibration signal in time domain by 3300 XL 8mm Proximity sensors and through a data acquisition card NI USB 6008, bringing data to the computer where we implemented a virtual instrument for capturing data and its subsequent transformation to obtain frequency spectrum. Consequently, we developed the algorithm in Matlab to automatically identify the imbalance present in the machine, using the technique of case-based reasoning, based on the calculation of the descriptors and the application of these within the algorithm implemented using the Euclidean distance as part of the decision mechanism among patterns without unbalancing vibration. The results show the RMS as the best performing descriptor for classification showed.