Structural Engineering and Mechanics

Techno-Press (테크노프레스)
권호 표지
DOI QR코드

DOI QR Code

Fault detection in blade pitch systems of floating wind turbines utilizing transformer architecture

  • Seongpil Cho (Department of Aeronautical and Astronautical Engineering, Korea Aerospace University) ;
  • Sang-Woo Kim (Department of Aeronautical and Astronautical Engineering, Korea Aerospace University) ;
  • Hyo-Jin Kim (Department of Korean Medical Science, Kyung Hee University)
  • Received : 2024.08.04
  • Accepted : 2024.09.27
  • Published : 2024.10.25
⟨ Previous Next ⟩

Abstract

This paper proposes a fault detection method for blade pitch systems of floating wind turbines using transformer-based deep-learning models. Transformers leverage self-attention mechanisms, efficiently process time-series data, and capture long-term dependencies more effectively than traditional recurrent neural networks (RNNs). The model was trained using normal operational data to detect anomalies through high reconstruction losses when encountering abnormal data. In this study, various fault conditions in a blade pitch system, including environmental load cases, were simulated using a detailed model of a spar-type floating wind turbine, the data collected from these simulations were used to train and test the transformer models. The model demonstrated superior fault-detection capabilities with high accuracy, precision, recall, and F1 scores. The results show that the proposed method successfully identifies faults and achieves high-performance metrics, outperforming existing traditional multi-layer perceptron (MLP) models and long short-term memory-autoencoder (LSTM-AE) models. This study highlights the potential of transformer models for real-time fault detection in wind turbines, contributing to more advanced condition-monitoring systems with minimal human intervention.

Keywords

Acknowledgement

We thank Professor Torgeir Moan and Zhen Gao at Norwegian University of Science and Technology (NTNU) for the valuable discussion and comments. This research was performed by the MIT-NTNU-Statoil Wind Turbine Program funded by Equinor (formerly Statoil). This research was also supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (No. 2022R1A6A1A03056784 and 2022R1C1C2006328).

References

  1. Bach-Andersen, M., Romer-Odgaard, B. and Winther, O. (2018), "Deep learning for automated drivetrain fault detection", Wind Energy, 21(1), 29-41. https://doi.org/10.1002/we.2142. 
  2. Bakdi, A., Kouadri, A. and Mekhilef, S. (2019), "A data-driven algorithm for online detection of component and system faults in modern wind turbines at different operating zones", Renew. Sustain. Energy Rev., 103, 546-555. https://doi.org/10.1016/j.rser.2019.01.013. 
  3. Carroll, J., McDonald, A. and McMillan, D. (2016), "Failure rate, repair time and unscheduled O&M cost analysis of offshore wind turbines", Wind Energy, 19, 1107-1119. https://doi.org/10.1002/we.1887. 
  4. Chang, M., Kim, J.K. and Lee, J. (2019), "Hierarchical neural network for damage detection using modal parameters", Struct. Eng. Mech., 70(4), 457-466. https://doi.org/10.12989/sem.2019.70.4.457. 
  5. Cho, K., Van Merrienboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H. and Bengio, Y. (2014), "Learning phrase representations using RNN encoder-decoder for statistical machine translation", arXiv preprint arXiv:1406.1078. https://doi.org/10.48550/arXiv.1406.1078. 
  6. Cho, S. (2024), "Anomaly detection in blade pitch systems of floating wind turbines using LSTM-Autoencoder", J. Aerosp. Syst. Eng., 18(4), 43-52. https://doi.org/10.20910/JASE.2024.18.4.43. 
  7. Cho, S., Bachynski, E., Nejad, A.R., Gao, Z. and Moan, T. (2020), "Numerical modeling of hydraulic blade pitch actuator in a spar-type floating wind turbine considering fault conditions and their effects on global dynamic responses", Wind Energy, 23(2), 370-390. https://doi.org/10.1002/we.2438. 
  8. Cho, S., Choi, M., Gao, Z. and Moan, T. (2021), "Fault detection and diagnosis of a blade pitch system in a floating wind turbine based on Kalman filters and artificial neural networks", Renew. Energy, 169, 1-13. https://doi.org/10.1016/j.renene.2020.12.116. 
  9. Cho, S., Gao, Z. and Moan, T. (2018), "Model-based fault detection, fault isolation and fault tolerant control of a blade pitch system in floating wind turbines", Renew. Energy, 120, 306-321. https://doi.org/10.1016/j.renene.2017.12.102. 
  10. Cho, S., Park, J. and Choi, M. (2021), "Fault classification of a blade pitch system in a floating wind turbine based on a recurrent neural network", J. Ocean Eng. Technol., 35(4), 287-295. https://doi.org/10.26748/KSOE.2021.018. 
  11. Choi, J.H. (2014), "A review on prognostics and health management and its applications", J. Aerosp. Syst. Eng., 8(4), 7-17. https://doi.org/10.20910/JASE.2014.8.4.007. 
  12. Cui, Y., Bangalore, P. and Tjernberg, L.B. (2021), "A fault detection framework using recurrent neural networks for condition monitoring of wind turbines", Wind Energy, 24(11), 1249-1262. https://doi.org/10.1002/we.2628. 
  13. de Andrade Melani, A.H., de Carvalho Michalski, M.A., da Silva, R.F. and de Souza, G.F.M. (2021), "A framework to automate fault detection and diagnosis based on moving window principal component analysis and Bayesian network", Reliab. Eng. Syst. Saf., 215, 107837. https://doi.org/10.1016/j.ress.2021.107837. 
  14. Dervilis, N., Choi, M., Taylor, S.G., Barthorpe, R.J., Park, G., Farrar, C.R. and Worden, K. (2014), "On damage diagnosis for a wind turbine blade using pattern recognition", J. Sound Vib., 333(6), 1833-1850. https://doi.org/10.1016/j.jsv.2013.11.015. 
  15. Dhibi, K., Mansouri, M., Bouzrara, K., Nounou, H. and Nounou, M. (2022), "Enhanced recurrent neural network for fault diagnosis of uncertain wind energy conversion systems", Proceedings of International Conference on Control, Decision and Information Technologies (CoDIT), 1, 1330-1335. https://doi.org/10.1109/CoDIT55151.2022.9804119. 
  16. Dinwoodie, I., McMillan, D., Revie, M., Lazakis, I. and Dalgic, Y. (2013), "Development of a combined operational and strategic decision support model for offshore wind", Energy Procedia, 35, 157-166. https://doi.org/10.1016/j.egypro.2013.07.169. 
  17. Finotti, R.P., Gentile, C., Barbosa, F. and Cury, A. (2022), "Structural novelty detection based on sparse autoencoders and control charts", Struct. Eng. Mech., 81(5), 647-664. https://doi.org/10.12989/sem.2022.81.5.647. 
  18. Gayo, J.B. (2011), "Reliability-focused research on optimizing wind energy system design, operation and maintenance: Tools, proof of concepts, guidelines & methodologies for a new generation", Final Publishable Summary of Results of Project ReliaWind. 
  19. Hameed, Z., Hong, Y.S., Cho, Y.M., Ahn, S.H. and Song, C.K. (2009), "Condition monitoring and fault detection of wind turbines and related algorithms: A review", Renew. Sustain. Energy Rev., 13, 1-39. https://doi.org/10.1016/j.rser.2007.05.008. 
  20. Harrou, F., Kini, K.R., Madakyaru, M. and Sun, Y. (2023), "Uncovering sensor faults in wind turbines: An improved multivariate statistical approach for condition monitoring using SCADA data", Sustain. Energy Grids Netw., 35, 101126. https://doi.org/10.1016/j.segan.2023.101126. 
  21. Hasselmann, K., Barnett, T.P., Bouws, E., Carlson, H., Cartwright, D.E., Enke, K., ... & Walden, H. (1973), "Measurements of wind-wave growth and swell decay during the Joint North Sea Wave Project (JONSWAP)", Ergaenzungsheft zur Deutschen Hydrographischen Zeitschrift, Reihe A. 
  22. Hochreiter, S. and Schmidhuber, J. (1997), "Long short-term memory", Neur. Comput., 9(8), 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735. 
  23. International Electrotechnical Commission (2005), IEC 61400-1: Wind Turbines-Part 1: Design Requirements. 
  24. International Electrotechnical Commission (2005), IEC 61400-3: Wind Turbines-Part 3: Design Requirements for Offshore Wind Turbines. 
  25. Isermann, R. (2006), Fault-diagnosis Systems: An Introduction from Fault Detection to Fault Tolerance, Springer Science & Business Media. 
  26. Jiang, G., Xie, P., He, H. and Yan, J. (2017), "Wind turbine fault detection using a denoising autoencoder with temporal information", IEEE/ASME Trans. Mechatron., 23(1), 89-100. https://doi.org/10.1109/TMECH.2017.2759301. 
  27. Jiang, G.Q., He, H.B., Yan, J. and Xie, P. (2018), "Multiscale convolutional neural networks for fault diagnosis of wind turbine gearbox", IEEE Trans. Ind. Electron., 66, 3196-3207. https://doi.org/10.1109/TIE.2018.2844805. 
  28. Johannessen, K., Meling, T. and Haver, S. (2001), "Joint distribution for wind and waves in the northern North Sea", Int. J. Offshore Polar Eng., 12(1), 1-8. 
  29. Jonkman, J. (2010), "Definition of the Floating System for Phase IV of OC3", Technical Report, NREL/TP-500-47535, USA. 
  30. Jonkman, J. and Kilcher, L. (2012), TurbSim User's Guide, Technical Report, NREL, USA. 
  31. Jonkman, J., Butterfield, S., Musial, W. and Scott, G. (2009), "Definition of a 5-MW reference wind turbine for offshore system development", Technical Report, NREL/TP-500-38060, USA. 
  32. Kingma, D.P. and Jimmy, B. (2014), "Adam: A method for stochastic optimization", arXiv, 1412.6980v9. https://doi.org/10.48550/arXiv.1412.6980/ 
  33. Kusiak, A. and Verma, A. (2011), "A data-driven approach for monitoring blade pitch faults in wind turbines", IEEE Trans. Sustain. Energy, 2, 87-96. https://doi.org/10.1109/TSTE.2010.2066585. 
  34. Kusiak, A. and Verma, A. (2012), "Analyzing bearing faults in wind turbines: A data-mining approach", Renew. Energy, 48, 110-116. https://doi.org/10.1016/j.renene.2012.04.020. 
  35. Lei, J., Liu, C. and Jiang, D. (2019), "Fault diagnosis of wind turbine based on Long Short-term memory networks", Renew. Energy, 133, 422-432. https://doi.org/10.1016/j.renene.2018.10.031. 
  36. Liu, J., Yang, G., Li, X., Wang, Q., He, Y. and Yang, X. (2023), "Wind turbine anomaly detection based on SCADA: A deep autoencoder enhanced by fault instances", ISA Trans., 139, 586-605. https://doi.org/10.1016/j.isatra.2023.03.045. 
  37. Germanischer Lloyd (2010), Guideline for the Certification of Wind Turbines, GL Report.
  38. Lu, B., Li, Y., Wu, X. and Yang, Z. (2009), "A review of recent advances in wind turbine condition monitoring and fault diagnosis", 2009 IEEE Power Electronics and Machines in Wind Applications, 1-7. 
  39. Malhotra, P., Vig, L., Shroff, G. and Agarwal, P. (2015), "Long short term memory networks for anomaly detection in time series", Proceedings of European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN), 89-94. 
  40. Marquez, F.P.G., Tobias, A.M., Perez, J.M.P. and Papaelias, M. (2012), "Condition monitoring of wind turbines: Techniques and methods", Renew. Energy, 46, 169-178. https://doi.org/10.1016/j.renene.2012.03.003. 
  41. Nguyen, H.D., Bui Tien, T., De Roeck, G. and Abdel Wahab, M. (2021), "Damage detection in structures using modal curvatures gapped smoothing method and deep learning", Struct. Eng. Mech., 77(1), 47-56. https://doi.org/10.12989/sem.2021.77.1.047. 
  42. Nguyen, H.D., Tran, K.P., Thomassey, S. and Hamad, M. (2021), "Forecasting and anomaly detection approaches using LSTM and LSTM autoencoder techniques with the applications in supply chain management", Int. J. Inf. Manage., 57, 102282. https://doi.org/10.1016/j.ijinfomgt.2020.102282. 
  43. Rahimilarki, R., Gao, Z., Jin, N. and Zhang, A. (2022), "Convolutional neural network fault classification based on time-series analysis for benchmark wind turbine machine", Renew. Energy, 185, 916-931. https://doi.org/10.1016/j.renene.2021.12.056. 
  44. Schlag, I., Irie, K. and Schmidhuber, J. (2021), "Linear transformers are secretly fast weight programmers", Proceedings of International Conference on Machine Learning, PMLR, 9355-9366. 
  45. Shin, S., Hyun, C., Cho, S. and Lee, P.S. (2023), "Multi-sensor data-based anomaly detection and diagnosis of a pumped storage hydropower plant", Struct. Eng. Mech., 88(6), 569-581. https://doi.org/10.12989/sem.2023.88.6.569. 
  46. SINTEF Ocean (2018), RIFLEX 4.15.0 User Guide. 
  47. SINTEF Ocean (2018), SIMO 4.15.0 User Guide. 
  48. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., ... and Polosukhin, I. (2017), "Attention is all you need", Advances in Neural Information Processing Systems, https://doi.org/10.48550/arXiv.1706.03762. 
  49. Wang, L., Zhang, Z., Long, H. and Xu, J. (2016), "Wind turbine gearbox failure identification with deep neural networks", IEEE Trans. Ind. Inform., 13(3), 1360-1368. https://doi.org/10.1109/TII.2016.2607179. 
  50. Wu, H., Triebe, M.J. and Sutherland, J.W. (2023), "A transformer-based approach for novel fault detection and fault classification/diagnosis in manufacturing: A rotary system application", J. Manuf. Syst., 67, 439-452. https://doi.org/10.1016/j.jmsy.2023.02.018. 
  51. Zaher, A.S., McArthur, S.D., Infield, D.G. and Patel, Y. (2009), "Online wind turbine fault detection through automated SCADA data analysis", Wind Energy, 12(6), 574-593. https://doi.org/10.1002/we.319. 
  52. Zhou, K.L., Fu, C. and Yang, S.L. (2016), "Big data driven smart energy management: From big data to big insights", Renew. Sustain. Energy Rev., 56, 215-225. https://doi.org/10.1016/j.rser.2015.11.050.