Skip to main content
SpringerLink
Log in

High-temperature power reduction state identification for wind turbines using feature correlation analysis and deep learning methods

Identifizierung des Zustands der Hochtemperatur-Leistungsreduzierung bei Windkraftanlagen mit Hilfe von Merkmalskorrelationsanalyse und Deep-Learning-Methoden

  • Originalarbeiten/Originals
  • Published:
Forschung im Ingenieurwesen Aims and scope Submit manuscript

Abstract

The excessively high temperature of gearbox oil in summer leads to power reduction of wind turbines (called high-temperature power reduction state), which does harm to the performance of the power generation performance of the wind turbine. The timely and accurate identification of high-temperature power reduction state will contribute to improving the operation efficiency of the unit. This study proposes a new method based on the vine copula model and algorithm of convolutional neural network cascading to the bidirectional long short term memory network with attention mechanism (CNN-BiLSTM-attention). Firstly, the vine copula model is used to analyze the correlation of the features in supervisory control and data acquisition (SCADA) system, and the features that can reflect the high-temperature power reduction state are extracted comprehensively. Secondly, CNN is used to mine the coupling relationship between features and extract deep spatial features. Finally, BiLSTM is used to extract the time-series information in the depth spatial feature further for high-temperature power reduction state identification, and the attention mechanism is introduced to sense and identify the relevant network weights adaptively to enhance the influence of important information. The experimental results show that the method has high identification accuracy in the identification of high-temperature power reduction state. Accurate and reliable identification results can provide reference for formulating the operation and maintenance scheme of wind turbine reasonably.

Zusammenfassung

Die übermäßig hohe Temperatur des Getriebeöls im Sommer führt zu einer Leistungsreduzierung von Windkraftanlagen (sog. Hochtemperatur-Leistungsreduzierungszustand), was der Leistung der Stromerzeugungsleistung der Windkraftanlage schadet. Die rechtzeitige und genaue Identifizierung des Hochtemperatur-Leistungsreduktionszustands wird dazu beitragen, die Betriebseffizienz der Anlage zu verbessern. In dieser Studie wird eine neue Methode vorgeschlagen, die auf dem Vine-Copula-Modell und dem Algorithmus des Faltungsneuronalen Netzes basiert, das mit dem bidirektionalen Langzeitgedächtnisnetz mit Aufmerksamkeitsmechanismus (CNN-BiLSTM-Attention) kaskadiert ist. Erstens wird das Vine-Copula-Modell verwendet, um die Korrelation der Merkmale im Überwachungs- und Datenerfassungssystem (SCADA) zu analysieren, und die Merkmale, die den Zustand der Hochtemperaturleistungsreduzierung widerspiegeln können, werden umfassend extrahiert. Zweitens wird CNN verwendet, um die Kopplungsbeziehung zwischen den Merkmalen zu ermitteln und tiefe räumliche Merkmale zu extrahieren. Schließlich wird BiLSTM verwendet, um die Zeitreiheninformationen in den räumlichen Tiefenmerkmalen zu extrahieren, um den Zustand der Hochtemperatur-Leistungsreduzierung zu identifizieren, und der Aufmerksamkeitsmechanismus wird eingeführt, um die relevanten Netzwerkgewichte adaptiv zu erkennen und zu identifizieren, um den Einfluss wichtiger Informationen zu verbessern. Die experimentellen Ergebnisse zeigen, dass die Methode eine hohe Erkennungsgenauigkeit bei der Identifizierung des Hochtemperatur-Leistungsreduktionszustands aufweist. Genaue und zuverlässige Identifizierungsergebnisse können als Referenz für die Formulierung des Betriebs- und Wartungsplans von Windkraftanlagen dienen.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11

Similar content being viewed by others

References

  1. Heydari A, Garcia DA, Fekih A, Keynia F, Tjernberg LB, De Santoli L (2021) A hybrid intelligent model for the condition monitoring and diagnostics of wind turbines gearbox. IEEE Access. https://doi.org/10.1109/ACCESS.2021.3090434

    Article  Google Scholar 

  2. Zeng J, Chen Y, Yang P et al (2018) Review of fault diagnosis methods of large-scale wind turbines. Power Syst Technol 42(3):849–860 (in Chinese)

    Google Scholar 

  3. Ruiming F, Minling W, xinhua G, Rongyan S, Pengfei S (2020) Identifying early defects of wind turbine based on SCADA data and dynamical network marker. Renew Energy 154:625–635

    Article  Google Scholar 

  4. Jin X, Xu Z, Qiao W (2021) Condition monitoring of wind turbine generators using SCADA data analysis. IEEE Trans Sustain Energy 12:202–210

    Article  Google Scholar 

  5. Zhang D, Zheng Q (2018) Probability warning for wind turbine gearbox incipient faults based on scada data. Chin Autom Congr

  6. Huang Z, Tian L, Dong X, Wei Y (2018) Prediction of oil temperature variations in a wind turbine gearbox based on pca and an spc-dynamic neural network hybrid. J Tsinghua Univ (Sci Technol): 539–546

  7. Bangalore P, Tjernberg LB (2017) An artificial neural network approach for early fault detection of gearbox bearings. IEEE Trans Smart Grid 6:980–987

    Article  Google Scholar 

  8. Cui Y, Bangalore P, Tjernberg LB (2018) 2018 IEEE international conference on probabilistic methods applied to power systems (PMAPS)—an anomaly detection. In: 2018 IEEE International Conference on Probabilistic Methods Applied to Power Systems(PMAPS), pp 1–6

    Google Scholar 

  9. Xu Q, Fan Z, Jia W, Jiang C (2019) Quantile regression neural network-based fault detection scheme for wind turbines with application to monitoring a bearing. Wind Energy 22:1390–1401

    Article  Google Scholar 

  10. Bartelmus W, Zimroz R (2009) A new feature for monitoring the condition of gearboxes in non-stationary operating conditions. Mech Syst Signal Process 23:1528–1534

    Article  Google Scholar 

  11. Rajabalizadeh S, Tafreshi SMM (2020) A practicable copula-based approach for power forecasting of small-scale photovoltaic systems. IEEE Syst J 14:4911–4918

    Article  Google Scholar 

  12. Li X, Zhang W (2020) Long-term fatigue damage assessment for a floating offshore wind turbine under realistic environmental conditions. Renew Energy 159:570–584

    Article  Google Scholar 

  13. Yueshan H et al (2020) Improved probabilistic load flow method based on D‑vine copulas and Latin hypercube sampling in distribution network with multiple wind generators. IET Gener Transm Distrib 14:893–899

    Article  Google Scholar 

  14. Yang X, Maihemuti M, Liu S, Wang Q (2020) “State Assessment Method of Capacity Reduction at High Temperature for Wind Turbine Based on vine copula Bayesian Network Model. In: Proceedings of the CSEE, pp 3583–3592 (in Chinese)

    Google Scholar 

  15. Xv Zengjin, Yuanyuan X, Yongjiang Y, Zuoxia X (2020) Cause Analysis and Solving Measures of High Oil Temperature in a Series of Wind Turbine Gearboxes. Acta Energ Sol Sin: 74–79

  16. Zhao Z, Chen W, Xingming W, Chen PCY, Jingmeng L (2017) LSTM network: A deep learning approach for short-term traffic forecast. IET Intell Trans Syst 11:68–75

    Article  Google Scholar 

  17. Yao G, Lei T, Zhong J (2019) A review of convolutional-neural-network-based action recognition. Pattern Recognit Lett 118:14–22

    Article  Google Scholar 

  18. Moslem M, Singh J, Bravo-Imaz I, Lee J (2020) Multisensor data fusion for gearbox fault diagnosis using 2‑D convolutional neural network and motor current signature analysis. Mech Syst Signal Process 144:106861

    Article  Google Scholar 

  19. Guo S, Yang T, Hua H, Cao J (2021) Coupling fault diagnosis of wind turbine gearbox based on multitask parallel convolutional neural networks with overall information. Renew Energy 178:639–650

    Article  Google Scholar 

  20. Xiang S, Qin Y, Zhu C, Wang Y, Chen H (2020) LSTM networks based on attention ordered neurons for gear remaining life prediction. ISA Trans 106:343–354

    Article  Google Scholar 

  21. Lei J, Liu C, Jiang D (2019) Fault diagnosis of wind turbine based on Long Short-term memory networks. Renew Energy 133:422–432

    Article  Google Scholar 

  22. Qianyu W et al (2021) Ultra-short-term multi-step wind power forecasting based on CNN-LSTM. IET Renew Power Gener 15:1019–1029

    Article  Google Scholar 

  23. Liu S, Lee K, Lee I (2020) Document-level multi-topic sentiment classification of Email data with BiLSTM and data augmentation. Knowl Based Syst 197:105918

    Article  Google Scholar 

  24. Kurowicka D, Cooke RM (2007) Sampling algorithms for generating joint uniform distributions using the vine copula method. Comput Stat Data Anal 51:2889–2906

    Article  MathSciNet  Google Scholar 

  25. Akaike H (1987) Factor analysis and AIC. Psychometrika 52(3):317–332

    Article  MathSciNet  Google Scholar 

  26. Dißmann J, Brechmann EC, Czado C, Kurowicka D (2013) Selecting and estimating regular vine copulae and application to financial returns. Comput Stat Data Anal 59:52–69

    Article  MathSciNet  Google Scholar 

  27. Brechmann EC, Czado C, Aas K (2012) Truncated regular vines in high dimensions with application to financial data. Can J Stat 40:68–85

    Article  MathSciNet  Google Scholar 

  28. Czado C, Schepsmeier U, Min A (2012) Maximum likelihood estimation of mixed C‑vines with application to exchange rates. Stat Modelling 12:229–255

    Article  MathSciNet  Google Scholar 

  29. Chen W (2008) The nonparametric testing of the fisher effect in China. Huazhong University of Science and Technology, Wuhan (in Chinese)

    Google Scholar 

  30. Che C, Wang H, Fu Q, Ni X (2019) Combining multiple deep learning algorithms for prognostic and health management of aircraft. Aerosp Sci Technol 94:105423

    Article  Google Scholar 

  31. Zhou H, Fang Z, Gao Y, Huang B, Zhong C, Shang R (2020) Feature fusion network based on attention mechanism for 3D semantic segmentation of point clouds. Pattern Recognit Lett 133:327–333

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. School of Control and Computer Engineering, North China Electric Power University, Beijing, China

    Xiyun Yang, Xinxin Huang, Xiaxiang Gao & Zhun Tao

Authors
  1. Xiyun Yang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Xinxin Huang
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Xiaxiang Gao
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Zhun Tao
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Xinxin Huang.

Ethics declarations

Conflict of interest

X. Yang, X. Huang, X. Gao and Z. Tao declare that they have no competing interests. I hereby certify that to the best of my knowledge, the authors have no relevant financial or non-financial interests to disclose. The authors have no conflicts of interest to declare that are relevant to the content of this article. All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript. The authors have no financial or proprietary interests in any material discussed in this article.

Appendix

Appendix

1.1 Appendix A

Table 9 Calculation result by AIC
Full size table

1.2 Appendix B

Table 10 The best copula functions for pair copulas
Full size table

1.3 Appendix C

Table 11 τk, ρs and τi,sum corresponding to pair copula
Full size table

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Yang, X., Huang, X., Gao, X. et al. High-temperature power reduction state identification for wind turbines using feature correlation analysis and deep learning methods. Forsch Ingenieurwes 86, 225–239 (2022). https://doi.org/10.1007/s10010-022-00586-y

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10010-022-00586-y

Search

Navigation