Feature Selection by Genetic Algorithm for Wind Power Prediction

Yıl 2022, , 1028 - 1040, 27.10.2022

Ece Çetin Yağmur , Sercan Yağmur

https://doi.org/10.35414/akufemubid.1117779

Öz

The need for renewable energy sources for sustainable development has been increasing every day. One of these sources is wind energy. Due to the stochastic nature of the wind, the estimation of wind speed and wind power has been a subject of great interest to researchers in recent years. In this study, wind power estimation was carried out for a wind turbine in Turkey, using the data set obtained by the SCADA system during 2018 and the meteorological data set shared by NASA for the same location. Wind speed, wind direction, and theoretical power curve were taken from the SCADA system as input variables; Meteorological parameters were taken from the NASA system and historical data of wind power were used. Unnecessary features in the model that cause computational complexity are removed from the model with the wrapper selection method to increase model performance. Genetic Algorithm (GA) was used as the wrapper selection method. In the study, the predictive power of different machine learning algorithms was compared according to different performance criteria and the effect of feature selection on the model was evaluated. In the final model proposed by GA, the number of variables was reduced from 47 to 9, unnecessary variables were removed from the model, and a strong prediction model with R2 value of 0.98 was obtained with the least number of variables.

Anahtar Kelimeler

Machine learning, Wind power, Renewable energy, Feature selection, Genetic algorithm

Rüzgar Gücü Tahmininde Genetik Algoritma ile Öznitelik Seçimi

Öz

Sürdürülebilir gelişim için yenilenebilir enerji kaynaklarına olan ihtiyaç her geçen gün artmaktadır. Bu kaynaklardan birisi de rüzgar enerjisidir. Rüzgarın stokastik yapısı nedeniyle rüzgar hızı ve rüzgar gücünün tahmini son yıllarda araştırmacılar tarafından oldukça ilgi çeken bir konu haline gelmiştir. Yapılan çalışmada Türkiye’de yer alan bir rüzgar türbini için 2018 yılı boyunca SCADA sistemi ile elde edilen veri seti ile aynı konum için NASA tarafından paylaşılan meteorolojik veri seti kullanılarak rüzgar gücü tahmini gerçekleştirilmiştir. Girdi değişkenleri olarak SCADA sisteminden çekilen rüzgar hızı, rüzgar yönü ve teorik güç eğrisi; NASA sisteminden çekilen meteorolojik parametreler ve rüzgar gücüne ait geçmiş veriler kullanılmıştır. Modelde yer alan ve hesaplama karmaşıklığına neden olan gereksiz öznitelikler model performansını artırmak amacıyla sarmal seçim yöntemi ile modelden çıkarılmıştır. Sarmal seçim yöntemi olarak Genetik Algoritma (GA) kullanılmıştır. Yapılan çalışmada hem farklı makine öğrenme algoritmalarının tahmin gücü, farklı performans ölçütlerine göre karşılaştırılmış hem de öznitelik seçiminin modele etkisi değerlendirilmiştir. GA ile önerilen nihai modelde değişken sayısı 47’den 9’a indirgenerek gereksiz değişkenler modelden uzaklaştırılmış ve en az sayıda değişken ile R2 değeri 0,98 olan güçlü bir tahmin modeli elde edilmiştir.

Anahtar Kelimeler

Makine öğrenmesi, Rüzgar gücü, Yenilenebilir enerji, Öznitelik seçimi, Genetik algoritma

Kaynakça

• Azimi R, Ghofrani M, Ghayekhloo M, 2016, A hybrid wind power forecasting model based on data mining and wavelets analysis. Energy conversion and management, 127, 208-225.
• Belkin M, Hsu D, Ma S, Mandal S, 2019, Reconciling modern machine-learning practice and the classical bias–variance trade-off. Proceedings of the National Academy of Sciences, 116, 32, 15849-15854.
• Breiman L, 2001, Random forests. Machine learning, 45, 1, 5-32.
• Demolli H, Dokuz AS, Ecemis A, Gokcek M, 2019, Wind power forecasting based on daily wind speed data using machine learning algorithms. Energy Conversion and Management, 198, 111823.
• Deng Y-C, Tang X-H, Zhou Z-Y, Yang Y, Niu F, 2021, Application of machine learning algorithms in wind power: a review. Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 1-22.
• Foley AM, Leahy PG, Marvuglia A, McKeogh EJ, 2012, Current methods and advances in forecasting of wind power generation. Renewable energy, 37, 1, 1-8.
• Heinermann J, Kramer O, 2016, Machine learning ensembles for wind power prediction. Renewable Energy, 89, 671-679.
• Higashiyama K, Fujimoto Y, Hayashi Y, 2018, Feature extraction of NWP data for wind power forecasting using 3D-convolutional neural networks. Energy Procedia, 155, 350-358.
• Hoerl AE, Kennard RW, 1970, Ridge regression: Biased estimation for nonorthogonal problems. Technometrics, 12, 1, 55-67.
• Holland JH, 1992. Adaptation in natural and artificial systems: an introductory analysis with applications to biology, control, and artificial intelligence, A Bradford Book MIT Press, p.
• Kuhn M, Johnson K, 2013. Applied predictive modeling, New York: Springer, 26, 13.
• Le TT, Fu W, Moore JH, 2020, Scaling tree-based automated machine learning to biomedical big data with a feature set selector. Bioinformatics, 36, 1, 250-256.
• Liu H, Chen C, 2019, Data processing strategies in wind energy forecasting models and applications: A comprehensive review. Applied Energy, 249, 392-408.
• Liu X, Zhang H, Kong X, Lee KY, 2020, Wind speed forecasting using deep neural network with feature selection. Neurocomputing, 397, 393-403.
• Lu P, Ye L, Zhao Y, Dai B, Pei M, Li Z, 2021, Feature extraction of meteorological factors for wind power prediction based on variable weight combined method. Renewable Energy, 179, 1925-1939.
• Melkumova L, Shatskikh SY, 2017, Comparing Ridge and LASSO estimators for data analysis. Procedia Engineering, 201, 746-755.
• Olson RS, Moore JH, TPOT: A tree-based pipeline optimization tool for automating machine learning, Workshop on automatic machine learning, 66-74.
• Renani ET, Elias MFM, Rahim NA, 2016, Using data-driven approach for wind power prediction: A comparative study. Energy Conversion and Management, 118, 193-203.
• Salcedo-Sanz S, Pastor-Sánchez A, Prieto L, Blanco-Aguilera A, García-Herrera R, 2014, Feature selection in wind speed prediction systems based on a hybrid coral reefs optimization–Extreme learning machine approach. Energy Conversion and Management, 87, 10-18.
• Shi J, Wang L, Lee W-J, Cheng X, Zong X, 2019, Hybrid Energy Storage System (HESS) optimization enabling very short-term wind power generation scheduling based on output feature extraction. Applied Energy, 256, 113915.
• Sun S, Qiao H, Wei Y, Wang S, 2017, A new dynamic integrated approach for wind speed forecasting. Applied Energy, 197, 151-162.
• Theofilatos A, Chen C, Antoniou C, 2019, Comparing machine learning and deep learning methods for real-time crash prediction. Transportation research record, 2673, 8, 169-178.
• Tibshirani R, 1996, Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society: Series B (Methodological), 58, 1, 267-288.
• Tso GK, Yau KK, 2007, Predicting electricity energy consumption: A comparison of regression analysis, decision tree and neural networks. Energy, 32, 9, 1761-1768.
• Wang H, Sun J, Sun J, Wang J, 2017, Using random forests to select optimal input variables for short-term wind speed forecasting models. Energies, 10, 10, 1522.
• Wang K, Qi X, Liu H, Song J, 2018, Deep belief network based k-means cluster approach for short-term wind power forecasting. Energy, 165, 840-852.
• WCED SWS, 1987, World commission on environment and development. Our common future, 17, 1, 1-91.
• Zhang S, 2012, Nearest neighbor selection for iteratively kNN imputation. Journal of Systems and Software, 85, 11, 2541-2552.
• Zou H, Hastie T, 2005, Regularization and variable selection via the elastic net. Journal of the royal statistical society: series B (statistical methodology), 67, 2, 301-320.
• İnternet kaynakları 1- https://sdgs.un.org/goals, (16.04.2022)
• 2- https://www.globalgoals.org/goals/7-affordable-and-clean-energy/ , (16.04.2022)
• 3- https://www.ren21.net/wp-content/uploads/2019/05/GSR2021_Full_Report.pdf , (14.04.2022)
• 4- https://www.kaggle.com/datasets/berkerisen/wind-turbine-scada-dataset, (16.05.2022)
• 5- https://www.tureb.com.tr/, (16.05.2022)
• 6- https://www.renewables.ninja/, (16.05.2022)