Short-term wind power forecasting based on cloud SVM model

A CSVM(Cloud Support Vector Machine) model combining the cloud model and the SVM(Support Vector Machine) is proposed for the short-term wind power forecasting,which applies the cloud transformation to extract the qualitative attribute of wind speed data and uses SVM to build the relationship between wind speed and wind power.The forecasts for the next 24 hours’ wind power show that,the forecasts at a particular point of the presented model is a set of discrete values with stabilized bias.The backward cloud algorithm is applied to calculate the expectation of the forecast set as the deterministic prediction,which is more accurate than that forecasted by SVM model or ARIMA(Auto-Regressive Integrated Moving Average) model.The presented model is effective for short-term wind power forecasting.