Araştırma Makalesi
BibTex RIS Kaynak Göster

Modeling the Wind Turbine Power Curve: A Comparison of Neural Networks and Classical Regression

Yıl 2026, Cilt: 9 Sayı: 2, 902 - 921, 16.03.2026
https://doi.org/10.47495/okufbed.1716847
https://izlik.org/JA78LY92SK

Öz

Accurate modeling of a wind turbine’s power curve – the relationship between wind speed and electrical power output – is crucial for wind energy forecasting and intelligent turbine control. This paper investigates data-driven approaches for power curve modeling using a real turbine dataset (48,000 observations of 10-minute average wind speed and power). We compare classical regression models (linear and polynomial fits) against an artificial neural network (ANN) model. Results show that the ANN can learn the non-linear S-shaped power curve more accurately (achieving higher R^2 and lower error) than linear or polynomial regression. The linear model struggles with the curve’s strong non-linearity, while a fourth-order polynomial fit achieves better accuracy but tends to deviate outside the training range. The ANN model captures turbine behavior across all wind speed regimes, including the low-power region at cut-in and the saturated region at rated power. Data-driven ANN modeling reduced root-mean-square error by about 70% compared to a simple linear fit. These results highlight the practical advantages of machine learning models over theoretical or fixed-form equations for wind turbine performance modeling. Improved power curve models enable more reliable wind power output forecasting and can be integrated into turbine control systems to optimize performance and detect anomalies.

Kaynakça

  • Carrillo C., Obando Montaño AF., Cidrás J., Díaz-Dorado E. Review of power curve modelling for wind turbines. Renewable and Sustainable Energy Reviews 2013; 21: 572-581.
  • Karaman ÖA. Prediction of wind power with machine learning models. Applied Sciences 2023; 13(20): 11455.
  • Kusiak A., Li W. The prediction and diagnosis of wind turbine faults. Renewable Energy 2011; 36(1): 16-23.
  • Lee JCY., Stuart P., Clifton A., Fields MJ., Perr-Sauer J., Williams L., Cameron L., Geer T., Housley P. The power curve Working Group’s assessment of wind turbine power performance prediction methods. Wind Energy Science 2020; 5(1): 199-223.
  • Li S., Wunsch DC., O’Hair E., Giesselmann MG. Comparative analysis of regression and artificial neural network models for wind turbine power curve estimation. Journal of Solar Energy Engineering 2001; 123(4): 327-332.
  • Lydia M., Kumar SS., Selvakumar AI., Prem Kumar GE. A comprehensive review on wind turbine power curve modeling techniques. Renewable and Sustainable Energy Reviews 2014; 30: 452-460.
  • Marugán AP., Márquez FPG., Perez JMP., Ruiz-Hernández D. A survey of artificial neural network in wind energy systems. Applied Energy 2018; 228: 1822-1836.
  • Mata SA., Pena Martínez JJ., Bas Quesada J., Palou Larrañaga F., Yadav N., Chawla JS., Sivaram V., Howland MF. Modeling the effect of wind speed and direction shear on utility-scale wind turbine power production. Wind Energy 2024; 27(9): 873-899.
  • Mushtaq K., Waris A., Zou R., Shafique U., Khan NB., Khan MI., Jameel M., Khan MI. A comprehensive approach to wind turbine power curve modeling: Addressing outliers and enhancing accuracy. Energy 2024; 304: 131981.
  • Pelletier F., Masson C., Tahan A. Wind turbine power curve modelling using artificial neural network. Renewable Energy 2016; 89: 207-214.
  • Rahman A., Smith AD. Predicting heating demand and sizing a stratified thermal storage tank using deep learning algorithms. Applied Energy 2018; 228: 108-121.
  • Saint-Drenan YM., Besseau R., Jansen M., Staffell I., Troccoli A., Dubus L., Schmidt J., Gruber K., Simões SG., Heier S. A parametric model for wind turbine power curves incorporating environmental conditions. Renewable Energy 2019; 157: 754-768.
  • Sebastiani A., Angelou N., Peña A. Wind turbine power curve modelling under wake conditions using measurements from a spinner-mounted lidar. Applied Energy 2024; 364: 122985.
  • Sohoni V., Gupta SC., Nema RK. A critical review on wind turbine power curve modelling techniques and their applications in wind based energy systems. Journal of Energy 2016; 2016: 1-18.
  • Texas Wind Turbine Dataset - Simulated. (n.d.). Retrieved October 5, 2025, from https://www.kaggle.com/datasets/pravdomirdobrev/texas-wind-turbine-dataset-simulated?select=TexasTurbine.csv
  • Villanueva D., Feijóo A. Comparison of logistic functions for modeling wind turbine power curves. Electric Power Systems Research 2018; 155: 281-288.
  • Villanueva D., Feijóo A. A review on wind turbine deterministic power curve models. Applied Sciences 2020; 10(12): 4186.
  • Wagner R., Antoniou I., Pedersen SM., Courtney MS., Jørgensen HE. The influence of the wind speed profile on wind turbine performance measurements. Wind Energy 2009; 12(4): 348-362.
  • Wang S., Zhang T., Su H. Enhanced hydrogen production from corn starch wastewater as nitrogen source by mixed cultures. Renewable Energy 2016; 96: 1135-1141.
  • Wang Z., Wang X., Liu W. Genetic least square estimation approach to wind power curve modelling and wind power prediction. Scientific Reports 2023; 13(1): 1-15.
  • Wind energy and power calculations. EM SC 470: Applied Sustainability in Contemporary Culture. (n.d.). Retrieved June 5, 2025, from https://www.e-education.psu.edu/emsc297/node/649
  • Wind turbine SCADA dataset. (n.d.). Retrieved June 5, 2025, from https://www.kaggle.com/datasets/berkerisen/wind-turbine-scada-dataset
  • Yang Y., Lou H., Wu J., Zhang S., Gao S. A survey on wind power forecasting with machine learning approaches. Neural Computing and Applications 2024; 36(21): 12753-12773.

Rüzgâr Türbini Güç Eğrisinin Modellenmesi: Yapay Sinir Ağları ve Klasik Regresyonun Karşılaştırılması

Yıl 2026, Cilt: 9 Sayı: 2, 902 - 921, 16.03.2026
https://doi.org/10.47495/okufbed.1716847
https://izlik.org/JA78LY92SK

Öz

Bir rüzgâr türbininin güç eğrisinin—rüzgâr hızı ile elektriksel güç çıkışı arasındaki fonksiyonel ilişkinin—doğru biçimde modellenmesi, güvenilir rüzgâr enerjisi tahmini ve gelişmiş türbin kontrolü için büyük önem taşımaktadır. Bu çalışmada, iki tamamlayıcı veri kümesi kullanılarak kapsamlı bir karşılaştırmalı analiz gerçekleştirilmiştir: (i) 10 dakikalık ortalamalara dayalı 48.000 gözleme sahip gerçek bir SCADA veri kümesi, ve (ii) National Renewable Energy Laboratory (NREL) tarafından oluşturulmuş, Texas bölgesine ait simüle edilmiş bir rüzgâr türbini veri kümesi. İkinci veri kümesi, 3.6 MW kapasiteli kara tipi bir General Electric türbini için tam yıllık saatlik ölçümler içermektedir. Her iki veri kümesi üzerinde üç veri odaklı model—Doğrusal Regresyon, Dördüncü Dereceden Polinom Regresyon ve Yapay Sinir Ağı (YSA)—uygulanmış, modellerin doğruluk ve genellenebilirlik performansları farklı işletim koşullarında karşılaştırılmıştır. YSA modeli, her iki veri kümesinde de en yüksek başarıyı göstermiştir (SCADA için R² = 0.993 ve RMSE = 107 kW; Texas için R² = 0.999 ve RMSE = 27 kW). Doğrusal model, türbinin karakteristik S-şekilli eğrisindeki doğrusal olmayan yapıyı yakalamakta yetersiz kalırken, polinom model doğruluğu artırmakta ancak eğitim aralığı dışında kararsız sonuçlar üretmektedir. Gerçek ve simüle edilmiş verilerin birlikte değerlendirilmesi, YSA tabanlı modellemenin farklı rüzgâr rejimlerinde sağlam ve uyarlanabilir olduğunu ortaya koymaktadır. Sonuçlar, makine öğrenmesi tabanlı yaklaşımların yalnızca klasik regresyon yöntemlerinden daha yüksek doğruluk sağlamakla kalmayıp, aynı zamanda farklı veri kaynakları arasında daha tutarlı tahminler sunduğunu göstermektedir. Bu bulgular, veri odaklı hibrit modellemenin rüzgâr gücü tahmini, performans optimizasyonu ve akıllı türbin kontrol sistemlerinde anormallik tespiti gibi alanlarda önemli potansiyel taşıdığını ortaya koymaktadır..

Kaynakça

  • Carrillo C., Obando Montaño AF., Cidrás J., Díaz-Dorado E. Review of power curve modelling for wind turbines. Renewable and Sustainable Energy Reviews 2013; 21: 572-581.
  • Karaman ÖA. Prediction of wind power with machine learning models. Applied Sciences 2023; 13(20): 11455.
  • Kusiak A., Li W. The prediction and diagnosis of wind turbine faults. Renewable Energy 2011; 36(1): 16-23.
  • Lee JCY., Stuart P., Clifton A., Fields MJ., Perr-Sauer J., Williams L., Cameron L., Geer T., Housley P. The power curve Working Group’s assessment of wind turbine power performance prediction methods. Wind Energy Science 2020; 5(1): 199-223.
  • Li S., Wunsch DC., O’Hair E., Giesselmann MG. Comparative analysis of regression and artificial neural network models for wind turbine power curve estimation. Journal of Solar Energy Engineering 2001; 123(4): 327-332.
  • Lydia M., Kumar SS., Selvakumar AI., Prem Kumar GE. A comprehensive review on wind turbine power curve modeling techniques. Renewable and Sustainable Energy Reviews 2014; 30: 452-460.
  • Marugán AP., Márquez FPG., Perez JMP., Ruiz-Hernández D. A survey of artificial neural network in wind energy systems. Applied Energy 2018; 228: 1822-1836.
  • Mata SA., Pena Martínez JJ., Bas Quesada J., Palou Larrañaga F., Yadav N., Chawla JS., Sivaram V., Howland MF. Modeling the effect of wind speed and direction shear on utility-scale wind turbine power production. Wind Energy 2024; 27(9): 873-899.
  • Mushtaq K., Waris A., Zou R., Shafique U., Khan NB., Khan MI., Jameel M., Khan MI. A comprehensive approach to wind turbine power curve modeling: Addressing outliers and enhancing accuracy. Energy 2024; 304: 131981.
  • Pelletier F., Masson C., Tahan A. Wind turbine power curve modelling using artificial neural network. Renewable Energy 2016; 89: 207-214.
  • Rahman A., Smith AD. Predicting heating demand and sizing a stratified thermal storage tank using deep learning algorithms. Applied Energy 2018; 228: 108-121.
  • Saint-Drenan YM., Besseau R., Jansen M., Staffell I., Troccoli A., Dubus L., Schmidt J., Gruber K., Simões SG., Heier S. A parametric model for wind turbine power curves incorporating environmental conditions. Renewable Energy 2019; 157: 754-768.
  • Sebastiani A., Angelou N., Peña A. Wind turbine power curve modelling under wake conditions using measurements from a spinner-mounted lidar. Applied Energy 2024; 364: 122985.
  • Sohoni V., Gupta SC., Nema RK. A critical review on wind turbine power curve modelling techniques and their applications in wind based energy systems. Journal of Energy 2016; 2016: 1-18.
  • Texas Wind Turbine Dataset - Simulated. (n.d.). Retrieved October 5, 2025, from https://www.kaggle.com/datasets/pravdomirdobrev/texas-wind-turbine-dataset-simulated?select=TexasTurbine.csv
  • Villanueva D., Feijóo A. Comparison of logistic functions for modeling wind turbine power curves. Electric Power Systems Research 2018; 155: 281-288.
  • Villanueva D., Feijóo A. A review on wind turbine deterministic power curve models. Applied Sciences 2020; 10(12): 4186.
  • Wagner R., Antoniou I., Pedersen SM., Courtney MS., Jørgensen HE. The influence of the wind speed profile on wind turbine performance measurements. Wind Energy 2009; 12(4): 348-362.
  • Wang S., Zhang T., Su H. Enhanced hydrogen production from corn starch wastewater as nitrogen source by mixed cultures. Renewable Energy 2016; 96: 1135-1141.
  • Wang Z., Wang X., Liu W. Genetic least square estimation approach to wind power curve modelling and wind power prediction. Scientific Reports 2023; 13(1): 1-15.
  • Wind energy and power calculations. EM SC 470: Applied Sustainability in Contemporary Culture. (n.d.). Retrieved June 5, 2025, from https://www.e-education.psu.edu/emsc297/node/649
  • Wind turbine SCADA dataset. (n.d.). Retrieved June 5, 2025, from https://www.kaggle.com/datasets/berkerisen/wind-turbine-scada-dataset
  • Yang Y., Lou H., Wu J., Zhang S., Gao S. A survey on wind power forecasting with machine learning approaches. Neural Computing and Applications 2024; 36(21): 12753-12773.
Toplam 23 adet kaynakça vardır.

Ayrıntılar

Birincil Dil İngilizce
Konular Makine Öğrenme (Diğer)
Bölüm Araştırma Makalesi
Yazarlar

Mahmud Asilsoy 0000-0003-1718-5075

Gönderilme Tarihi 10 Haziran 2025
Kabul Tarihi 26 Ekim 2025
Yayımlanma Tarihi 16 Mart 2026
DOI https://doi.org/10.47495/okufbed.1716847
IZ https://izlik.org/JA78LY92SK
Yayımlandığı Sayı Yıl 2026 Cilt: 9 Sayı: 2

Kaynak Göster

APA Asilsoy, M. (2026). Modeling the Wind Turbine Power Curve: A Comparison of Neural Networks and Classical Regression. Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi, 9(2), 902-921. https://doi.org/10.47495/okufbed.1716847
AMA 1.Asilsoy M. Modeling the Wind Turbine Power Curve: A Comparison of Neural Networks and Classical Regression. Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi. 2026;9(2):902-921. doi:10.47495/okufbed.1716847
Chicago Asilsoy, Mahmud. 2026. “Modeling the Wind Turbine Power Curve: A Comparison of Neural Networks and Classical Regression”. Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi 9 (2): 902-21. https://doi.org/10.47495/okufbed.1716847.
EndNote Asilsoy M (01 Mart 2026) Modeling the Wind Turbine Power Curve: A Comparison of Neural Networks and Classical Regression. Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi 9 2 902–921.
IEEE [1]M. Asilsoy, “Modeling the Wind Turbine Power Curve: A Comparison of Neural Networks and Classical Regression”, Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi, c. 9, sy 2, ss. 902–921, Mar. 2026, doi: 10.47495/okufbed.1716847.
ISNAD Asilsoy, Mahmud. “Modeling the Wind Turbine Power Curve: A Comparison of Neural Networks and Classical Regression”. Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi 9/2 (01 Mart 2026): 902-921. https://doi.org/10.47495/okufbed.1716847.
JAMA 1.Asilsoy M. Modeling the Wind Turbine Power Curve: A Comparison of Neural Networks and Classical Regression. Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi. 2026;9:902–921.
MLA Asilsoy, Mahmud. “Modeling the Wind Turbine Power Curve: A Comparison of Neural Networks and Classical Regression”. Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi, c. 9, sy 2, Mart 2026, ss. 902-21, doi:10.47495/okufbed.1716847.
Vancouver 1.Mahmud Asilsoy. Modeling the Wind Turbine Power Curve: A Comparison of Neural Networks and Classical Regression. Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi. 01 Mart 2026;9(2):902-21. doi:10.47495/okufbed.1716847

23487




196541947019414  

1943319434 19435194361960219721 19784  2123822610 23877

* Uluslararası Hakemli Dergi (International Peer Reviewed Journal)

* Yazar/yazarlardan hiçbir şekilde MAKALE BASIM ÜCRETİ vb. şeyler istenmemektedir (Free submission and publication).

* Yılda Ocak, Mart, Haziran, Eylül ve Aralık'ta olmak üzere 5 sayı yayınlanmaktadır (Published 5 times a year)

* Dergide, Türkçe ve İngilizce makaleler basılmaktadır.

*Dergi açık erişimli bir dergidir.

Creative Commons License

Bu web sitesi Creative Commons Atıf 4.0 Uluslararası Lisansı ile lisanslanmıştır.