Research on Cleaning Method of Wind Power Data Based on Improved Quartile Method

A method is proposed to address the challenge of anomalous wind speed and power data in wind farms, aiming to clean the wind power data. This method utilizes an enhanced quartile approach and involves two main steps. Initially, the Bin zoning method is employed to preprocess the extensive data by leveraging the actual operational data from wind turbines in Jilin Wind Farm. Subsequently, the enhanced quartile method is implemented to cleanse the wind power data post zoning. In this step, the improved quartile method cleans the scattered data, and the MeanShift algorithm is employed to clean the folded data. Finally, the proposed algorithm is compared with the data purified by the quartile method and the MeanShift method, respectively.