An Ensemble Modeling Approach for Power Curve with Multivariables

Abstract

It is important to establish reliable and accurate power curve model for wind energy optimization and monitoring. Considering the difference of wind turbine power output under different influencing factors, an integrated modeling technique is established in this paper to realize power evaluation based on multiple influencing factors. To overcome the sensitivity of the model to outliers, the data filtering technology is added to improve the accuracy of the modeling. In addition, the k-medoids ++ algorithm is used to divide the data to describe the data differences theoretically. The proposed modeling technique has been verified on three wind turbines in a wind farm in Liaoning Province, China, and the results show that the modeling technique is reliable.