@article { author = {Ghasemi Mobtaker, Hassan and Ajabshirchi, Yahya and Ranjbar, Seyed Faramarz and Matloobi, Mansour and Taki, Morteza}, title = {Estimation of Monthly Mean Daily Global Solar Radiation in Tabriz Using Empirical Models and Artificial Neural Networks}, journal = {Journal of Renewable Energy and Environment}, volume = {3}, number = {3}, pages = {21-30}, year = {2016}, publisher = {Materials and Energy Research Center (MERC) Iranian Association of Chemical Engineers (IAChE)}, issn = {2423-5547}, eissn = {2423-7469}, doi = {10.30501/jree.2016.70089}, abstract = {Precise knowledge ofthe amount of global solar radiation plays an important role in designing solar energy systems. In this study, by using 22-year meteorologicaldata, 19 empirical models were tested for prediction of the monthly mean daily global solar radiation in Tabriz. In addition, various Artificial Neural Network (ANN) models were designed for comparison with empirical models. For this purpose, the meteorological data recorded by Iran Meteorological Organization (1992–2013) was used. These data include: monthly mean daily sunshine duration, monthly mean ambient temperature, monthly mean maximum and minimum ambient temperature and monthly mean relative humidity.Theresults showed that the yearly average solar radiation in the region was 16.37 MJ m .Among the empirical models, the best result was acquired for model (19) with correlation coefficient (r) of 0.9663. Results also showed that the ANN model trained with total meteorological data in input layer (ANN5) produces better results in comparison to others. Root Mean Square Error (RMSE) and r for this model were1.0800 MJ m-2 and 0.9714, respectively. Comparison betweenthe model 19 and ANN5, demonstrated that modeling the monthly mean daily global solar radiationthrough the use of the ANNtechnique, yields better estimates. Mean Percentage Errors (MPE) for these models were 7.4754% and 1.0060%, respectively. -2 day-1}, keywords = {Solar Energy,Meteorological Data Sunshine Hours,prediction,Artificial Neural Networks}, url = {https://www.jree.ir/article_70089.html}, eprint = {https://www.jree.ir/article_70089_e90209589f20f0c9ba458e09720af50e.pdf} }