Stacked Autoencoder-Based Anomaly Detection for Predictive Maintenance of Wind Turbine Mechanical Components.

Najaad OubeBlika

Stacked Autoencoder-Based Anomaly Detection for Predictive Maintenance of Wind Turbine Mechanical Components.

Najaad OubeBlika
Volume 5

Abstract

Wind turbines operate under highly variable environmental and mechanical conditions, including fluctuating wind speed, turbulence, temperature changes, cyclic loading, and continuous rotational stress. These operating conditions can accelerate degradation in critical mechanical components such as gearboxes, bearings, shafts, generators, brakes, and rotor assemblies. Unexpected failures in these components may lead to costly downtime, reduced energy production, expensive corrective maintenance, and safety risks, particularly in remote or offshore wind farms where maintenance access is difficult and operational costs are high. Conventional condition monitoring and threshold-based fault detection methods often rely on fixed alarm limits and expert-defined rules, which may fail to identify subtle early-stage degradation patterns hidden within high-dimensional sensor data. To address these challenges, this paper proposes a stacked autoencoder-based anomaly detection framework for predictive maintenance of wind turbine mechanical components.

The proposed framework leverages the unsupervised feature learning capability of stacked autoencoders to model normal operating behavior and detect abnormal deviations that may indicate incipient mechanical faults. The system utilizes multi-source turbine monitoring data, including supervisory control and data acquisition (SCADA) signals, vibration measurements, acoustic emission data, bearing temperature, gearbox oil temperature, generator temperature, rotor speed, wind speed, wind direction, power output, torque, and nacelle position. A comprehensive preprocessing pipeline is applied to clean and synchronize sensor data, remove outliers, handle missing values, normalize continuous variables, filter noise, and construct operational condition-aware input representations. To reduce false alarms caused by normal environmental variability, the framework incorporates wind-speed binning, operating-regime clustering, and context-aware normalization.

The stacked autoencoder architecture is trained primarily on healthy turbine operating data to learn compressed latent representations of normal mechanical behavior. During inference, deviations between reconstructed and observed sensor patterns are quantified using reconstruction error, latent-space distance, and statistical anomaly scoring. Abnormal patterns exceeding adaptive thresholds are flagged as potential early indicators of component degradation. The model is designed to identify anomalies associated with bearing wear, gearbox tooth damage, shaft imbalance, lubrication problems, generator overheating, rotor misalignment, and drivetrain vibration irregularities. To improve robustness and diagnostic value, the framework integrates temporal smoothing, health index construction, and remaining useful life trend analysis, enabling maintenance teams to distinguish transient disturbances from persistent degradation.

The proposed framework can be evaluated using real-world wind turbine SCADA datasets, vibration monitoring datasets, and simulated mechanical fault scenarios. Performance is assessed using anomaly detection metrics such as precision, recall, F1-score, false alarm rate, detection delay, area under the precision–recall curve, reconstruction error distribution, and early warning lead time. Comparative analysis is conducted against conventional statistical process monitoring, principal component analysis, one-class support vector machine, isolation forest, shallow autoencoder, LSTM autoencoder, and rule-based condition monitoring approaches. Experimental evaluation is expected to demonstrate that the stacked autoencoder model achieves improved sensitivity to early mechanical degradation while maintaining a low false alarm rate under diverse operating conditions.

To support practical deployment, the proposed anomaly detection system can be integrated into wind farm supervisory control platforms and predictive maintenance dashboards. Explainability and diagnostic interpretation are incorporated through sensor contribution analysis, reconstruction error decomposition, latent feature visualization, and component-level health indicators. These tools allow maintenance engineers to identify which sensor channels or mechanical subsystems contribute most strongly to an anomaly, supporting targeted inspection and maintenance planning. The proposed framework provides a scalable, data-driven, and cost-effective solution for early fault detection in wind turbine mechanical components. By enabling proactive maintenance decisions, reducing unplanned downtime, extending component lifetime, and improving wind farm reliability, the system contributes to more efficient and sustainable wind energy generation.

Keywords

References

[1] El-Kenawy, E. S. M., Mirjalili, S., Abdelhamid, A. A., Ibrahim, A., Khodadadi, N., & Eid, M. M. (2022). Meta -heuristic optimization
[2] Abdelhamid, A. A., El -Kenawy, E. S. M., Khodadadi, N., Mirjalili, S., Khafaga, D. S., Alharbi, A. H., ... & Saber, M. (2022).
[3] Alharbi, A. H., Towfek, S. K., Abdelhamid, A. A., Ibrahim, A., Eid, M. M., & Khafaga, D. S. & Saber, M.(2023). Diagnosis of
[4] Alhussan, A. A., & Towfek, S. K. (2024). 5G Resource Allocation Using Feature Selection and Greylag Goose Optimization Algori thm.
[5] Towfek, S. K., & Alhussan, A. A. (2024). Potato Production Forecasting Based on Balance Dynamic Biruni Earth Radius Algorithm for
[6] Abdelhamid, A. A., Alhussan, A. A., Qenawy, A. S. T., Osman, A. M., Elshewey, A. M., & Eed, M. (2024). Potato harvesting pred iction
[7] Eed, M., Alhussan, A. A., Qenawy, A. S. T., Osman, A. M., Elshewey, A. M., & Arnous, R. (2024). Potato Consumption Forecastin g
[8] Mahmood, S., Sun, H., Iqbal, A., Alhussan, A. A., & El -kenawy, E. S. M. (2024). Green finance, sustainable infrastructure, and green
[9] El-Kenawy, E. S. M., Alhussan, A. A., Khodadadi, N., Mirjalili, S., & Eid, M. M. (2024). Predicting potato crop yield with machi ne
[10] Radwan, M., Alhussan, A. A., Ibrahim, A., & Tawfeek, S. M. (2025). Potato leaf disease classification using optimized machine learning
[11] Ghasemi, M., Khodadadi, N., Trojovský, P., Li, L., Mansor, Z., Abualigah, L., ... & El -Kenawy, E. S. M. (2025). Kirchhoff’s law
[12] Yassen, M. A., El -Kenawy, E. S. M., Abdel -Fattah, M. G., Ismail, I., & Mostafa, H. E. D. S. (2025). Explainable artificial intelligence
[13] Radwan, M., Alhussan, A. A., Ibrahim, A., & Tawfeek, S. M. (2025). Potato leaf disease classification using optimized machine learning
[14] Mozhdehi, A. T., Khodadadi, N., Aboutalebi, M., El -Kenawy, E. S. M., Hussien, A. G., Zhao, W., ... & Mirjalili, S. (2025). Divine
[15] El-kenawy, E. S. M., Alhussan, A. A., Mattar, E. A., & Radwan, M. (2026). Feature selection and hyperparameter tuning in transfo rmer -
[16] Chen, L., Xu, C., Lim, W. H., Sharma, A., Tiang, S. S., Chong, K. S., ... & Khafaga , D. S. (2025). Transparent and reliable construction
[17] Khodadadi, N., Towfek, S. K., Zaki, A. M., Alharbi, A. H., Khodadadi, E., Khafaga, D. S., ... & Eid, M. M. (2025). Predicting
[18] Alhussan, A. A., El -Kenawy, E. S. M., Khafaga, D. S., Alharbi, A. H., & Eid, M. M. (2025). Groundwater resource prediction and
[19] Mozhdehi, A. T., Khodadadi, N., Aboutalebi, M., El -Kenawy, E. S. M., Hussien, A. G., Zhao, W., ... & Mirjalili, S. (2025). Divine
[20] Lopes, J. M., Pinho, C. S., Martins, R. C., & Bandeira, T. (2026). Sustainability Gets Smarter: Competitive Pressure and AI L eading the
[21] Radwan, M., Ibrahim, A., Abdelsalam, M. M., Alhussan, A. A., Mattar, E. A., & El -Kenawy, E. S. M. (2026). Optimizing solar and wind
[22] Alhussan, A. A., El -Kenawy, E. S. M., Eid, M. M., & Khodadadi, N. (2026). Hybrid Al -Biruni and Puma Optimization (BERPO) for
[23] El-Kenawy, E. S. M., Ibrahim, A., Alhussan, A. A., Khafaga, D. S., Ahmed, A. E., & Eid, M. M. (2026). Smart city electricity loa d
[24] El-Kenawy, E. S. M., Khodadadi, N., Mirjalili, S., Zaki, A. M., Ibrahim, A., Alhussan, A. A., ... & Eid, M. M. (2026). Glider sn ake
[25] Mekaret, F., Rabehi, A., Zebentout, B., Tizi, S., Douara, A., Bellucci, S., ... & Alhussan, A. A. (2024). A comparative study of Schottky
[26] KOUADRI, Ali, RABEHI, Abdelhalim, BENZIANE, Ali, et al. A Robust Multi -Transform Watermarking Scheme for Medical Images
[27] Tibermacine, I. E., Russo, S., Scarano, G., Tedesco, G., Rabehi, A., Alhussan, A. A., ... & Napoli, C. (2025). Conditional VA E for
[28] Kouadri, A., Benziane, A., Rabehi, A., Rabehi, A., Alhussan, A. A., Khafaga, D. S., & El -Kenawy, E. S. M. (2025). A novel hybrid
[29] Russo, S., Tibermacine, I. E., Randieri, C., Rabehi, A., Alharbi, A. H., El -Kenawy, E. S. M., & Napoli, C. (2025). Exploiting facial
[30] Bentegri, H., Rabehi, M., Kherfane, S., Nahool, T. A., Rabehi, A., Guermoui, M., ... & El -Kenawy, E. S. M. (2025). Assessment of
[31] Tibermacine, I. E., Russo, S., Citeroni, F., Mancini, G., Rabehi, A., Alharbi, A. H., ... & Napoli, C. (2025). Adversarial de noising of EEG
[32] Ouahabi, M. S., Benyounes, A., Barkat, S., Ihammouchen, S., Rekioua, T., Rabehi, A., ... & Alharbi, A. H. (2025). Real -time sensor fault
[33] Belaid, A., Guermoui, M., Khelifi, R., Arrif, T., Chekifi, T., Rabehi, A., ... & Alhussan, A. A. (2024). Assessing suitable a reas for PV
[34] Mehallou, A., M’hamdi, B., Amari, A., Teguar, M., Rabehi, A., Guermoui, M., ... & Khafaga, D. S. (2025). Optimal multiobjecti ve
[35] Rabehi, A., El -Hadi, M., Benmahmoud, S., Rabehi, A., Alharbi, A. H., & El -Kenawy, E. S. M. (2025). SOCA -CFAR Processor in A
[36] Bakria, D., Beladel, A., Korich, B., Teta, A., Mohammedi, R. D., Laouid, A. A., ... & El -kenawy, E. S. (2025). A novel enhanced Grey

Download