Elman Neural Network for Solar Radiation Components Forecasting based on the Desired Tilt Angle

The tilt angle of solar systems affect directly the conversion of solar radiation, regardless of weather conditions. To extract the maximum power from the solar panels, the sunlight should fall with a sharp angle. Therefore, the optimum angle of inclination of fixed solar panels to the horizontal plane should be changed regularly every month and every season. The prime aim of this study is to offer for solar systems engineers and designers an efficient model based on multi-input multi output (MIMO) Elman artificial neural network (EL-ANN). To predict the three components of hourly solar radiation time series (global, direct, diffuse radiation), relative to the site of Marrakesh city, Morocco, received by different tilted solar panel surface. The prediction is based on the tilt angle desired by the user and the meteorological data, including of time variable. The obtained results prove the robustness and the efficiency of the developed model to estimate accurately the three components of hourly solar radiation time series for different tilt angles.