Vibration Analysis of an Industrial Motor with Autoencoder for Predictive Maintenance

Cristian Nuñez*, Roberto Moreno, Victor Benitez, Jesus Pacheco

*Autor correspondiente de este trabajo

Producción científica: Capítulo del libro/informe/acta de congresoContribución a la conferenciarevisión exhaustiva

Resumen

This paper describes the implementation of a vibration analysis in an industrial servo motor for anomaly detection. For this, a test bench was built with the purpose of simulating an industrial process. The vibration analysis was performed with an accelerometer which took the acceleration data from a running engine. For the detection of anomalies, an Autoencoder was used which was trained with samples of the normal operation of the motor in order to reconstruct a “normal operation” signal. Once the model was trained, the MAE (Mean Absolute Error) is used to see the differences between the analyzed signal and the one reconstructed by the Autoencoder, if the difference is greater than a threshold, the signal is classified as an anomaly. The proposed methodology represents an alternative to perform vibration analysis in rotative machines and can be used to conduct predictive maintenance in several industrial processes.

Idioma originalInglés
Título de la publicación alojadaAdvances in Computational Intelligence - 21st Mexican International Conference on Artificial Intelligence, MICAI 2022, Proceedings
EditoresObdulia Pichardo Lagunas, Bella Martínez Seis, Juan Martínez-Miranda
EditorialSpringer Science and Business Media Deutschland GmbH
Páginas252-265
Número de páginas14
ISBN (versión impresa)9783031194955
DOI
EstadoPublicada - 2022
Evento21st Mexican International Conference on Artificial Intelligence, MICAI 2022 - Monterrey, México
Duración: 24 oct. 202229 oct. 2022

Serie de la publicación

NombreLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volumen13613 LNAI
ISSN (versión impresa)0302-9743
ISSN (versión digital)1611-3349

Conferencia

Conferencia21st Mexican International Conference on Artificial Intelligence, MICAI 2022
País/TerritorioMéxico
CiudadMonterrey
Período24/10/2229/10/22

Nota bibliográfica

Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Huella

Profundice en los temas de investigación de 'Vibration Analysis of an Industrial Motor with Autoencoder for Predictive Maintenance'. En conjunto forman una huella única.

Citar esto