Forecasting Solar Radiation Based on Meteorological Data Using Machine Learning Techniques: A Case Study of Isparta

Buğra Güzel; Onur Sevli; Ersan Okatan

Makale detayları

Forecasting Solar Radiation Based on Meteorological Data Using Machine Learning Techniques: A Case Study of Isparta

Yazarlar:

Buğra Güzel | Onur Sevli | Ersan Okatan

Yayıncı:

Kirikkale University

Ana Başlıklar:

Engineering

Dergi:

International Journal of Engineering Research and Development Vol.15 Issue 2, pp.704-713

Belge Tipi:

Makale

Yayın Tarihi:

7/2023

Belge Dili:

İngilizce

Document ID (Org No.):

2001284767

pISSN:

1308-5506

Anahtar kelimeler:

Özet:

 Solar energy systems which is one of renewable energy sources takes more interest and gains prevalence day by day. As in other many renewable energy sources, a significant problem in solar energy systems is the unstability of the energy that the system will provide. Prediction of the energy to be obtained is very important in this respect. In this study, solar radiation is predicted using meteorological data taken from the General Directorate of Meteorology for Isparta. For predictions, the random forest RF, KNN k-Nearest Neighbor, ANN Artificial Neural Networks and Deep Learning DL methods are used. In addition, the results of dummy variable usage for time data are examined with these different methods. According to the findings obtained, it is seen that the dummy variable usage increases performance for ANN and DL methods but decreases performance for random forest and KNN methods. Best results are obtained for the prediction of the solar radiation with ANN and DL.

 Solar energy systems which is one of renewable energy sources takes more interest and gains prevalence day by day. As in other many renewable energy sources, a significant problem in solar energy systems is the unstability of the energy that the system will provide. Prediction of the energy to be obtained is very important in this respect. In this study, solar radiation is predicted using meteorological data taken from the General Directorate of Meteorology for Isparta. For predictions, the random forest RF, KNN k-Nearest Neighbor, ANN Artificial Neural Networks and Deep Learning DL methods are used. In addition, the results of dummy variable usage for time data are examined with these different methods. According to the findings obtained, it is seen that the dummy variable usage increases performance for ANN and DL methods but decreases performance for random forest and KNN methods. Best results are obtained for the prediction of the solar radiation with ANN and DL.

 Solar energy systems which is one of renewable energy sources takes more interest and gains prevalence day by day. As in other many renewable energy sources, a significant problem in solar energy systems is the unstability of the energy that the system will provide. Prediction of the energy to be obtained is very important in this respect. In this study, solar radiation is predicted using meteorological data taken from the General Directorate of Meteorology for Isparta. For predictions, the random forest RF, KNN k-Nearest Neighbor, ANN Artificial Neural Networks and Deep Learning DL methods are used. In addition, the results of dummy variable usage for time data are examined with these different methods. According to the findings obtained, it is seen that the dummy variable usage increases performance for ANN and DL methods but decreases performance for random forest and KNN methods. Best results are obtained for the prediction of the solar radiation with ANN and DL.

 Yenilenebilir enerji kaynaklarından olan güneş enerji sistemleri her geçen gün daha fazla ilgi görmekte ve kullanımı yaygınlaşmaktadır. Diğer birçok yenilenebilir enerji kaynaklarında olduğu gibi güneş enerji sistemlerindeki önemli bir sorun sistemin sağlayacağı enerjinin sürekli olmamasıdır. Elde edilecek enerjinin tahmin edilebilmesi bu bakımdan oldukça önemlidir. Bu çalışmada Meteoroloji Genel Müdürlüğünden Isparta ili için alınan meteorolojik veriler kullanılarak güneş ışınımı tahmini yapılmıştır. Tahmin işlemi için Rastgele Orman RF, k-EYK k-En Yakın Komşu, YSA Yapay Sinir Ağları ve Derin Öğrenme yöntemleri kullanılmıştır. Ayrıca zaman verileri için kukla değişken kullanımının bu farklı metotlar ile oluşturduğu sonuçlar incelenmiştir. Elde edilen bulgulara göre kukla değişken kullanımının YSA ve Derin Öğrenme yöntemlerinde performansı arttırdığı, Rastgele orman ve k-EYK yöntemlerinde ise performansı düşürdüğü görülmüştür. YSA ve derin öğrenme ile güneş ışınımı tahmininde en iyi sonuçlar elde edilmiştir.

Makale detayları

Forecasting Solar Radiation Based on Meteorological Data Using Machine Learning Techniques: A Case Study of Isparta

Yazarlar:

Buğra Güzel | Onur Sevli | Ersan Okatan

Yayıncı:

Kirikkale University

Ana Başlıklar:

Engineering

Dergi:

International Journal of Engineering Research and Development Vol.15 Issue 2, pp.704-713

Belge Tipi:

Makale

Yayın Tarihi:

7/2023

Belge Dili:

İngilizce

Document ID (Org No.):

2001284767

pISSN:

1308-5506

Artificial neural networks

deep learning

dummy variable

random forest

solar radiation prediction

derin öğrenme

kukla değişken

rastgele orman

güneş ışınımı tahmini

yapay sinir ağları

Sayı

Makale detayları

Yazarlar: Buğra Güzel | Onur Sevli | Ersan Okatan

Yayıncı: Kirikkale University

Ana Başlıklar: Engineering

Dergi: International Journal of Engineering Research and Development Vol.15 Issue 2, pp.704-713

Belge Tipi: Makale

Yayın Tarihi: 7/2023

Belge Dili: İngilizce

Document ID (Org No.): 2001284767

pISSN: 1308-5506

Sayı

Yazarlar:

Buğra Güzel | Onur Sevli | Ersan Okatan

Yayıncı:

Kirikkale University

Ana Başlıklar:

Engineering

Dergi:

International Journal of Engineering Research and Development Vol.15 Issue 2, pp.704-713

Belge Tipi:

Makale

Yayın Tarihi:

7/2023

Belge Dili:

İngilizce

Document ID (Org No.):

2001284767

pISSN:

1308-5506