Total Daily Solar Irradiance Prediction using Recurrent Neural Networks with Determinants

Total solar irradiance appears the performance of non-linear change affected by many factors, which can be divided into the major and minor factors. Correlation analysis is used to find out the main determinants. For the sake of higher accuracy, a recurrent back-propagation network is established to forecast the daily total solar irradiance with the inputs of the major factors in this paper. A discount coefficient method is adopted in updating the weights and biases of the networks so as to make the closest forecasts playing more important roles. Based on historical daily records of solar irradiance in Shanghai as samples, an example is presented with the forecasted total solar irradiance. The results of the example indicate that the method makes the forecasts much more accurate than the forecasts using the artificial neural networks without the inputs of the main determinants.