Modeling of global horizontal irradiance in the United Arab Emirates with artificial neural networks

摘要: Abstract This paper employs ANN (Artificial Neural Network) models to estimate GHI (global horizontal irradiance) for three major cities in the UAE (United Arab Emirates), namely Abu Dhabi, Dubai and Al-Ain. City data are then used develop a comprehensive global model other nearby locations UAE. The use MLP (Multi-Layer Perceptron) RBF (Radial Basis Function) techniques with training algorithms, architectures, different combinations of inputs. tested validated against individual city available from Solar Atlas good agreement as attested by computed statistical error parameters. optimal is MLP-based requires four mean daily weather parameters; namely, maximum temperature, wind speed, sunshine hours, relative humidity. parameters MLP-ANN relation measured three-cities (referred data) MBE (mean bias error) = −0.0003 kWh/m 2 , RMSE = 0.179 kWh/m R  = 99%, NSE (Nash-Sutcliffe Efficiency coefficient) = 99%, t-statistic = 0.005 at 5% significance level. Results prove suitability estimating monthly