DESARROLLO Y COMPARACIÓN DE MODELOS PARA LA GENERACIÓN DE CURVAS FFT A PARTIR DE IMÁGENES DE TEMPERATURA–TIEMPO PARA LA CLASIFICACIÓN DE LA LUBRICACIÓN EN COJINETES DE RODILLOS
DOI:
https://doi.org/10.61164/j87vfa28Palabras clave:
Transformada Rápida de Fourier, Cojinetes de Rodillos, Lubricación, Redes Neuronales Convolucionales, Procesamiento de SeñalesResumen
Este estudio presenta el desarrollo y la comparación de cinco variantes de un modelo computacional diseñado para convertir imágenes de curvas temperatura–tiempo en representaciones espectrales mediante la Transformada Rápida de Fourier (FFT). Se utilizaron 387 imágenes generadas a partir de archivos .csv del KAIST Bearing Run-to-Failure Dataset (2020), que representan tres regímenes de lubricación en cojinetes de rodillos (healthy, marginal y starved). Cada variante aplica estrategias específicas de suavizado, ponderación energética y realce espectral para producir curvas FFT adecuadas para la clasificación automática mediante redes neuronales convolucionales (CNN). Los resultados indican que las versiones FFT-Weighted, FFT-Composite y FFT-1D-Plus ofrecen un desempeño superior, generando espectros con alta estabilidad global y buen nivel de detalle local. Este equilibrio mejora la diferenciación entre clases, proporcionando patrones texturales y contrastes espectrales consistentes para el aprendizaje convolucional. Se concluye que la aplicación de la FFT a las curvas térmicas incrementa la expresividad de los datos y constituye una etapa prometedora en la preparación de conjuntos de datos para el reconocimiento de regímenes de lubricación mediante CNN.
Descargas
Referencias
ACTUATORS. The Fault Diagnosis of Rolling Bearings Based on FFT SE TCN SVM. Actuators, v. 14, n. 3, p. 152, 2025. Disponível em: https://www.mdpi.com/2076-0825/14/3/152. Acesso em: 31 out. 2025.
CARVALHO, R. M.; CARVALHO, T. J. A.; DINIZ, B. C. Development of an Image Processing Model for Bearing Lubrication Classification Using Convolutional Neural Networks. In: XI SIINTEC – International Symposium on Innovative Technologies, 2025, Salvador. Proceedings: XI SIINTEC, 2025.
COOLEY, J. W.; TUKEY, J. W. An algorithm for the machine calculation of complex Fourier series. Mathematics of Computation, v. 19, n. 90, p. 297–301, 1965.
KOREA ADVANCED INSTITUTE OF SCIENCE AND TECHNOLOGY (KAIST). Bearing Run-to-Failure Dataset. Daejeon, South Korea, 2020.
LECUN, Y.; BENGIO, Y.; HINTON, G. Deep learning. Nature, v. 521, n. 7553, p. 436–444, 2015.
OPPENHEIM, A. V.; SCHAFER, R. W. Discrete-Time Signal Processing. 3. ed. Upper Saddle River: Prentice Hall, 2010.
SENSORS. Bearing Fault Diagnosis Using Lightweight and Robust One Dimensional Convolution Neural Network in the Frequency Domain. Sensors, v. 22, n. 15, p. 5793, 2022. Disponível em: https://www.mdpi.com/1424-8220/22/15/5793. Acesso em: 31 out. 2025.
WU, X.; YANG, C.; ZHENG, H.; LI, X. Intelligent fault diagnosis of rotating machinery based on improved CNN with frequency–domain feature fusion. Sensors, v. 23, n. 8, p. 3921, 2023.
XU, M.; YU, Q.; CHEN, S.; LIN, J. Rolling Bearing Fault Diagnosis Based on CNN-LSTM with FFT and SVD. Information, v. 15, n. 7, p. 399, 2024. Disponível em: https://www.mdpi.com/2078-2489/15/7/399. Acesso em: 31 out. 2025.
ZHANG, H.; ZHOU, Y.; WANG, Z.; LI, J. Fault diagnosis of rolling bearings based on time–frequency representation and convolutional neural networks. Mechanical Systems and Signal Processing, v. 181, p. 109484, 2022.
ZHOU, J.; QIN, Y.; KOU, L.; YUWONO, M.; SU, S. Fault detection of rolling bearing based on FFT and classification. Journal of Advanced Mechanical Design, Systems, and Manufacturing, v. 9, n. 5, 2015. Disponível em: https://www.jstage.jst.go.jp/article/jamdsm/9/5/9_2015jamdsm0056/_article. Acesso em: 31 out. 2025.
Descargas
Publicado
Número
Sección
Licencia
Derechos de autor 2025 Bruno da Cunha Diniz, Ranya Mota Carvalho, Talisson de Jesus Alves Carvalho
Esta obra está bajo una licencia internacional Creative Commons Atribución 4.0.
Authors who publish in this journal agree to the following terms:
Authors retain copyright and grant the journal the right of first publication, with the work simultaneously licensed under the Creative Commons Attribution License, which permits the sharing of the work with proper acknowledgment of authorship and initial publication in this journal;
Authors are authorized to enter into separate, additional agreements for the non-exclusive distribution of the version of the work published in this journal (e.g., posting in an institutional repository or publishing it as a book chapter), provided that authorship and initial publication in this journal are properly acknowledged, and that the work is adapted to the template of the respective repository;
Authors are permitted and encouraged to post and distribute their work online (e.g., in institutional repositories or on their personal websites) at any point before or during the editorial process, as this may lead to productive exchanges and increase the impact and citation of the published work (see The Effect of Open Access);
Authors are responsible for correctly providing their personal information, including name, keywords, abstracts, and other relevant data, thereby defining how they wish to be cited. The journal’s editorial board is not responsible for any errors or inconsistencies in these records.
PRIVACY POLICY
The names and email addresses provided to this journal will be used exclusively for the purposes of this publication and will not be made available for any other purpose or to third parties.
Note: All content of the work is the sole responsibility of the author and the advisor.
