Solar radiation estimate and forecasting by neural networks for smart grid energy management

The high rate of penetration of renewable energy in the context of smart grids and distributed generation makes the prediction of meteorological time series particularly useful for planning and management of the power grid with the aim of improving its overall efficiency and performance. On such a basis, this paper proposes an application of Artificial Neural Networks (ANNs) to the field of photovoltaic power generation. In particular, two suitably trained dynamic recurrent ANNs, i.e., the Focused Time-Delay Neural Network (FTDNN) and the Nonlinear autoregressive network with exogenous inputs (NARX), are used to develop a model for the estimate and forecast of daily solar radiation. ANNs implemented in this study show good performance since reliable and precise models of daily solar radiation, are obtained. This allows the PV output power for a given plant to be forecast as well. Finally, the potential of the proposed method in optimal sizing and energy management of electrical grids is outlined showing an example of NARX network application to electric load forecast.