A WT-ARMA based method for wind power ramp events forecasting

摘要

In an electric power system with high penetration of wind power, the sudden increase or rapid decrease of power output in a short time, known as wind power ramp event, poses a serious threat to the power system. The wind power ramp events forecasting can help the grid operators to minimize this impact in advance by electric grid scheduling. In this paper, a forecasting model, WT-ARMA (Wavelet Transform-Auto Regressive and Moving Average), was proposed. Firstly, WT was applied to decompose the observed ramp series into several sub-signals; then these sub-signals were reconstructed into two new series; next, the future values of each new series was predicted respectively based on ARMA; finally ramp events were predicted according to the predicted values of the series above. The effectiveness of WT-ARMA was tested with a wind farm from northern China, and the results showed that the forecast accuracy and ramp capture rate can both be improved with the proposed model.