Correction of Wind Power Prediction Error Under the Extreme Weather Based on K-nearest Neighbor Algorithm

Extreme weather, such as cold waves and frosts, can significantly impact the output of wind farms. To address this issue, this paper proposes a short-term power prediction and correction method suitable for wind farms during extreme weather, based on the K-nearest neighbor algorithm and multiple linear regression model. By verifying that wind power output is affected by wind speed, temperature, humidity, wind power kurtosis, and other factors, the K-nearest neighbor algorithm is used to obtain the sample set of similar days on the day to be measured. The regression model of influencing factors and output errors is established to improve the power prediction accuracy. A wind farm in northwest China is used as a test example to verify the effectiveness of the output correction method.