A hybrid model for wind speed prediction using empirical mode decomposition and artificial neural networks

Wind speed forecasting is one of the most important technologies to guarantee the wind energy integrated into the whole power system smoothly. In this paper a hybrid model named EMD–ANN for wind speed prediction is proposed based on the Empirical Mode Decomposition (EMD) and the Artificial Neural Networks (ANNs). To choose the best training algorithm for the ANN model, several experimental simulations with different training algorithms are made. To estimate the performance of the EMD–ANN model, two forecasting cases are completed and the results are both compared with the ANN model and the Autoregressive Integrated Moving Average (ARIMA) model respectively. To avoid the randomness caused by the ANN model or the ANN part of hybrid EMD–ANN model, all simulations in this study are repeated at least 30 times to get the average. The results show that: (1) the performance of the proposed model is highly satisfactory; and (2) the proposed EMD–ANN hybrid method is robust in dealing with jumping samplings in non-stationary wind series.

► A new hybrid EMD–ANN model is proposed to forecast wind speed.
► In EMD–ANN, the EMD is adopted to decompose the original data.
► In EMD–ANN, the ANN is used to build the forecasting models for sub-series.
► A comparison of the proposed method, the ANN and the ARIMA are provided.
► The EMD–ANN model has better performance than the ANN and the ARIMA.