Short-term hybrid forecasting model of wind power based on EEMD-ARIMA-LSTM

Inhalt:
Under the dual-carbon goal, the rapid development of renewable energy and the increasing penetration of wind power pose a challenge to the safe and stable operation of the power grid. High-precision wind power prediction is essential. A large number of studies have shown that the decomposition ensemble model has good predictive ability. A hybrid prediction model based on Ensemble Empirical Mode Decomposition (EEMD), Autoregressive Integrated Moving Average model (ARIMA), and long short-term memory network (LSTM) is proposed to improve the prediction accuracy of wind power. The model uses EEMD to decompose the original wind power data into a series of IMF components with different bandwidths to reduce the non-stationarity and nonlinearity of the data, and uses the frequency index zerocrossing rate to reconstruct the sub-sequence into a low-frequency component and a high-frequency component. In this paper, ARIMA model and LSTM model are established to predict the high and low frequency components respectively, and the final prediction results are obtained by superimposing the prediction values of each component. By selecting wind power data from a wind farm in southwest China for simulation experiments, the results show that the EEMD-ARIMA-LSTM model has better tracking effect and prediction accuracy than other models.