A Statistical Approach of Wind Power Forecasting for Grid Scale

Abstract

Grid Scale Wind Power Forecast is an effective way to increase acceptable wind power capacity and improve the Reliability and Security of power system integrated with large scale of wind power. In this paper, modelling technology for Grid scale wind power forecasting is discussed firstly. Using correlation matrix of output power and forecast accuracy coefficient, representative wind farms are selected and weight coefficient of each farm is determined. Then grid scale wind power forecast method based on the power output of representative farm is presented. And then the method is validated by the running data of a grid connected wind farm. The experiment result indicates the feasibility and effectiveness of the method presented in this paper, with RMSE as 12.91%, MAE as only 9.9%, and usable data rate reached 87.2%. The forecasting accuracy meets the application demands of projects and far ahead of the demands of industrial standard. This method is able to economize the computing resource, reduce the dependency to the data integrity of single wind farm of grid scale wind power forecast modelling, and has high practical value.