Ensemble Nonlinear Autoregressive Exogenous Artificial Neural Networks for Short-Term Wind Speed and Power Forecasting
Figure 10
In-sample “predictions” and out-of-sample multi-step-ahead forecasts of the wind power for the best-fitting ANN obtained from using the CFD wind speed data as the exogenous input (a) and discrepancies between the forecasted and measured wind power (b). These predictions were obtained with a NARX ANN model structure.