Anomaly Detection Framework for Rotary Equipment Using ContinuousWavelet Transform and U-Net Autoencoders
##plugins.themes.bootstrap3.article.main##
##plugins.themes.bootstrap3.article.sidebar##
Abstract
Recent advances in data-driven methods, particularly deep learning, have transformed predictive maintenance for rotary machinery. These methods enable intelligent, sensor-based condition monitoring from unlabeled operational data, even under rare-fault conditions. This study proposes an unsupervised anomaly detection framework for rotary equipment that utilizes continuous wavelet transform (CWT) to transform unlabeled, multichannel vibration signals into stacked time-frequency scalograms using complex Morlet wavelet. These scalograms are then processed by an enhanced U-Net deep convolutional autoencoder (CWT-U-Net CAE), which learns features of healthy operational conditions and detects anomalies by identifying significant deviations in reconstruction error. Coupled with its edge-compatibility, the framework enables scalable real-time condition monitoring in industrial environments. A custom test bench with an induction motor was used to obtain realistic vibrational signatures under normal operating conditions, assessing the effectiveness of the proposed approach.
##plugins.themes.bootstrap3.article.details##
Anomaly Detection Framework, Condition Monitoring, U-Net Autoencoder, Wavelet Transform, Predictive Maintenance
Bernitsas, E., & Kourkoutos-Ardavanis, N. (2021). The emerging role of scalogram-based convolutional neural networks in epileptic seizure detection. Brain Sciences, 11(11), 1424. doi: 10.3390/brainsci11111424
Chen, J., & Wu, H. (2024). Convolutional autoencoderbased anomaly detection using MEMS vibration sensors. IEEE Sensors Journal, 24(6), 10245-10256.
Chou, H., & Wang, Y. (2025). YOLO-based fault detection using time–frequency representations of vibration signals. In Proceedings of the IEEE international conference on prognostics and health management (phm).
Guo, S., Yang, T., Gao, W., & Zhang, C. (2018). A novel fault diagnosis method for rotating machinery based on a convolutional neural network. Sensors, 18(5), 1429.
He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR) (pp. 770-778).
Huang, W., & Sun, L. (2025). Engineering-driven fault diagnosis using vibration and current signal fusion. Mechanical Systems and Signal Processing, 200, 110945.
Jain, M., Patel, V., & Raj, T. (2020). Vibration monitoring of CNC machinery using mems sensors. Journal of Vibroengineering, 22(4), 899-910. doi: 10.21595/jve.2020.21125
Kanungo, P. (2025). Edge computing in healthcare: Realtime patient monitoring systems. World Journal of Advanced Engineering Technology and Sciences, 15(1), 001-009. doi: 10.30574/wjaets.2025.15.1.0168
Li, C., & Zhou, Y. (2023). Multi-sensor fusion with autoencoder models for industrial anomaly detection. IEEE Transactions on Industrial Informatics, 19(4), 5640-5651.
Malviya, A., & Singh, P. (2022). Edge computing for predictive maintenance: FPGA-based anomaly detection with lightweight autoencoders. In Proceedings of the IEEE international conference on edge computing (pp. 45-52).
Pang, G., Shen, C., Cao, L., & van den Hengel, A. (2021). Deep learning for anomaly detection: A review. ACM Computing Surveys, 54(2), 1-38. doi: 10.1145/3439950
Park, J., & Kim, D. (2023). Multi-head attention networks for weak fault diagnosis in industrial motors. Neural Computing and Applications, 35, 15187-15199.
Tang, Y., & Han, J. (2024). Anomaly detection in rotating machinery using residual CNNs with temporal modeling. Mechanical Systems and Signal Processing, 190, 110289.
Ventricci, A., & Marino, S. (2024). Motor fault classification using low-cost MEMS accelerometers and deep learning. Sensors, 24(3), 865.
Yan, R., Gao, R. X., & Chen, X. (2014). Wavelets for fault diagnosis of rotary machines: A review with applications. Signal Processing, 96, 1-15.
Yang, B., & Li, X. (2021). Bearing fault diagnosis under variable speed conditions using CNN–LSTM networks. IEEE Access, 9, 97890-97902.
Yedurkar, D. P., et al. (2023). Early fault diagnosis of rolling bearing based on threshold acquisition U-Net. Machines, 11(1), 119. doi: 10.3390/machines11010119
Zhang, L., & Chen, M. (2023). Multi-scale convolutional networks with CWT for intelligent fault diagnosis of bearings. Mechanical Systems and Signal Processing, 185, 109735.
Zheng, H., & Xu, Y. (2023). Correlation-aware feature learning for vibration-based anomaly detection. IEEE Transactions on Industrial Electronics, 70(7), 7412-7421.
This work is licensed under a Creative Commons Attribution 3.0 Unported License.