A comparative study of estimating solar radiation using machine learning approaches: DL, SMGRT, and ANFIS

ABSTRACT

Solar energy has a key role in producing clean and emissions-free power compare to conventional methods. However, sustainable development also requires a reliable and predictable energy source. It also needs methods to measure and predict predictable supply. The main aim of the study is to improve reliable and precise solar radiation prediction models on monthly mean daily basis using various machine learning techniques. Simple Membership Function and Fuzzy Rule Generating Technique (SMGRT), which does not require error and trial for model adjustment, is the first-choice model in this study. Experience and observations about the model will greatly reduce the volume of processing for the fuzzy SMGRT model. On the other hand, Deep Learning (DL) and Adaptive Neuro-Fuzzy Inference System (ANFIS) have become increasingly popular in understanding nonlinear data structures and solving complex problems. Therefore, DL and ANFIS were also applied to estimate solar radiation. The data set used in the study were created using sunshine duration (s), extra-terrestrial solar radiation (H₀), relative humidity (RH), cloudiness (C), air temperature (T) and soil temperature (ST) parameters. Estimation performance of models was evaluated by using several statistical indicators which are Mean Bias Error (MBE), Mean Square Error (MSE), Root Mean Square Error (RMSE), and Correlation Coefficient (R²). When the performances of the models were compared, it was seen that all three models obtained remarkable results. In addition, it was shown that the models performed well based on the metrics in the testing phase. The SMGRT model has slightly better performance than DL and ANFIS for different input combinations. SMGRT Model 1 (with inputs H₀, s, and T) shows the best statistical performance (MBE = 0.156, MSE = 1.878, RMSE = 1.371, and R² = 0.960) not only in SMGRT models but also in others.

KEYWORDS:

• ANFIS
• deep learning
• machine learning
• SMGRT
• solar radiation

Nomenclature

MBEMean bias error, MJ/m²

MPEMean percentage error, ͦ

RMSERoot mean square error, MJ/m²

R²Correlation coefficient

□□Sunrise hour angle, ͦ

$\delta$Declination angle, ͦ

$\mathrm{\varnothing }$Latitude of location, ͦ

$G_{sc}$Solar Constant, W/m²

KtClearness index

STSoil Temperature, °C

RHRelative Humidity, %

TTemperature, °C

CCloudiness, 8 Okta

DL Deep Learning

SMGRTSimple Membership Function and Fuzzy Rule Generation Technique

ANFISAdaptive Neuro-Fuzzy Inference System

H_m,iMeasured solar radiation, MJ/m²

H_c,iCalculated solar radiation, MJ/m²

H₀Extraterrestrial solar radiation, MJ/m²

Additional information

Notes on contributors

İsmail Üstün

Ismail Üstün, He received his BS in Mechanical Engineering from the Iskenderun Technical University in 2015. He received his MSc in Mechanical Engineering from the Iskenderun Technical University in 2018. He started his PhD in Mechanical Engineering at Iskenderun Technical University in 2018. His research interests are renewable energy, artificial neural network, thermodynamics and energy.

Fatih Üneş

Fatih Üneş, He received his BSc degree in Civil Engineering Department of Akdeniz University, Turkey in 1989. He received his MSc degree from the Department of Civil Engineering of Çukurova University, Institute of Science and Technology Turkey in 1996. He received PhD degree from Department of Civil Engineering of Istanbul Technical University in 2004. He is presently serving as a Professor in the Department of Civil Engineering Faculty of Iskenderun Technical University, Turkey. He has authored/co-authored more than 15 technical articles.

İlker Mert

İlker Mert, He received his BSc in Electronic Engineering from Erciyes University, Turkey, in 2002. He received PhD degree from Department of Mechanical Engineering of Iskenderun Technical University in 2018.His major research involves meteorology-related renewable energy sources, stochastic behavior of wind, wind power and power generation system under changing climate in Turkey. Besides artificial neural networks, fuzzy networks, and other evolutionary computation techniques are his other interests. He is presently serving as an lecturer in the Osmaniye Korkut Ata Üniversitesi.

Cuma Karakuş

Cuma Karakuş, He received his BSc and MSc degree in Mechanical Engineering from the Çukurova University, Turkey in 1992, and 1997, respectively. He received PhD degree from Department of Mechanical Engineering of Çukurova University in 2007. He is presently serving as an Assoc. Professor in the department of Mechanical Engineering in İskenderun Technical University, Turkey.  His current research interests include partice image velocimetry applied to fluid mechanic and wind energy applications.