Research on short-term wind power combined forecasting and its Gaussian cloud uncertainty to support the integration of renewables and EVs

Under the pressure of environmental pollution and energy shortage, wind power generation and EVs with clean and pollution-free characteristics have developed rapidly. However, the randomness of EVs charging and the volatility of wind power output will bring great challenges to the reliability and economy of grid operation. Especially the accuracy and range of wind power forecasting are critical to the operation of the power system with a high proportion of renewable energy and EVs. Aiming at improving the accuracy of short-term wind power forecasting and its uncertainty, this paper puts forward a combined forecasting model, including BP, Wavelet, and RVM by information fusion strategy, Gaussian Cloud model is used to reflect the uncertainty in the forecasting process. According to the measured data of two units, the results of short-term wind power forecasting are analyzed and compared with the single forecasting method. It's found that the combined forecasting model can improve the forecasting accuracy with more reasonable confidence interval. The power grid can guide the EVs to dynamically adjust the EVs charging time according to the forecasting wind power and EVs charging power curves, so as to maximize the absorption of wind power, achieve the economic operation and reduce pollution emissions. (C) 2020 Elsevier Ltd. All rights reserved.