Wind power forecast and error analysis based on Forgetting Factor Least Square method

This paper reports an application of Forgetting Factors Least Square (FFLS) in the wind power forecasting field. The FFLS algorithm is applied to obtain the estimated parameter of the correction model, and then the properly structured correction model was used to correct the forecasting errors. Satisfactory experimental results are obtained for day-ahead forecast by using actual wind power data in the north of China. To make a comparison with the proposed approach, similar experiments using ordinary least squares (OLS) are also performed.