Short-term wind forecasting and its impact on the mix-mode operated buildings

In this paper, we developed a new method to forecast 10-minutes ahead wind speed based on Heteroscdastic Gaus-sian Process and investigated the impacts of this predict-ing on a mix-mode operated building. The forecasting re-sult shows the mean absolute percentage error (MAPE) of 9.2%. The indoor air temperature, infiltration air change rate and cooling energy consumption varied 25% in av-erage compared with the baseline model where the wind speed is not from forecasting.