Wind Speed Power Data Cleaning Method for Wind Turbines Based on Fan Characteristics and Isolated Forests

Abstract

To address the situations that wind turbine operating conditions are complex and variable, the wind speed and power data are confusing, and there are many anomalies with significant differences, which make data analysis difficult, a method of cleaning wind speed and power data of wind turbines based on turbine characteristics and isolated forests is established. Firstly, the wind turbine power data is initially screened with the actual wind turbine characteristics. Then, isolated trees are constructed based on the initial filtered data. Afterward, the built isolated trees are used to calculate the power anomalies to be cleaned and to distinguish the anomalies, which finally achieves effective and efficient data cleaning. The experimental analysis results with the actual unit data of a wind farm show that this method is more efficient and more general than that of the traditional cleaning methods.