Current status of wind energy forecasting and a hybrid method for hourly predictions

• A comprehensive review for accurate forecasting is provided.
• Extensive tables listing the prediction errors and used methods for different horizons are given.
• A prediction method for 1 h ahead predictions is described.
• The method gives markedly low errors for daily average wind speeds.

Generating accurate wind energy and/or power forecasts is crucially important for energy trading and planning. The present study initially gives an extensive review of recent advances in statistical wind forecasting. Numerous prediction methods for varying prediction horizons from a few seconds to several months are listed. Then in the light of accurate results in the literature, the present study combines the adaptive neuro-fuzzy inference system (ANFIS) and an artificial neural network (ANN) for 1 h ahead wind speed forecasts. The performance results show the mean absolute percentage errors (MAPE) of 2.2598%, 3.3530% and 3.8589% at three different locations for daily average wind speeds.