Short term solar irradiance forecasting using artificial neural network for a semi-arid climate in Morocco

Abstract

Knowledge of irradiance with high accuracy is of paramount importance for monitoring planning and for better exploitation and distribution of photovoltaic (PV) energy. Different methods have been developed to accurately forecast solar irradiance, in this study, we applied multilayer perception to predict GHI (Global Horizontal Irradiance) for a hot semiarid climate in Benguerir, Morocco. Ground measurements of several meteorological variables and the GHI from the meteorological station installed at the green energy park in Benguerir were used to build the database, different model architectures with various inputs were tested to choose the most efficient model. To evaluate the performance of the models we used the nMBE (normalized Mean Bias Error), nRMSE (normalized Root Mean Square Error), and CC (Correlation Coefficient). The results favored the MLP (Multilayer Perceptron) with 3 inputs and7 neurons in the hidden layer. The final model was experimented to predict GHI for clear and cloudy days, nMBE, nRMSE and CC obtained are respectively (-0.051%, 0.10% and 0.99) for clear days, and (0.14%,0.39% and 0.96%) for cloudy days.