Regional forecasting of wind speed in large scale wind plants

The ultra-short-term forecasting of the wind speed in large-scale wind plants not only helps to monitor system operation but also improves the acceptance rate for the power system. The present study establishes a single unit and a regional wind speed forecasting models and achieves the fast prediction of the wind speed in the large-scale wind plant. On the basis, by combining the variational mode decomposition (VMD) and autoregressive integrated moving average (ARIMA) model, an improved VMD-ARIMA model for wind speed forecasting of the single unit is proposed, and the cross-correlation coefficient is presented to determine the number of modes in VMD. Moreover, according to the similarity of wind speed series of different units, the wind plant is divided into several sub-regions based on the correlation of different units. Finally, by calculating the proportion coefficient between the key unit and the non-key one, the wind speed of the whole wind plant is forecasted fast. To assess the performance of the proposed model, it is compared with some forecasting model. Results validated by the actual data sets for Jishanliang wind plant, Shaanxi province, China, show higher efficiency and better reliability of the proposed model in comparison with other forecasting methods.