Evaluation of aging characteristics in wind turbine performance based on yaw power loss

Abstract

The yaw dynamics of wind turbines are crucial for ensuring their operational efficiency and maximizing wind energy capture. However, excessive yaw movements may precipitate premature aging of these turbines. Investigating how yaw behavior influences turbine performance can aid in refining yaw control strategies, thereby mitigating the rate of performance degradation. This paper analyzes five years of SCADA data from a wind farm, employs the DBSCAN algorithm to process anomalous data, and explores the correlation between state parameters and power output under varying operational conditions. The study leverages kernel density estimation and least squares approximation for univariate data processing and curve fitting. Furthermore, it introduces the concept of a yaw loss rate to assess power efficiency during yaw maneuvers quantitatively, calculates yaw-induced power losses under diverse conditions, and proposes a novel method to evaluate turbine performance by considering historical trends in power capture. The findings confirm that the proposed evaluation methodology is practical and effective, substantiated by analyzing five consecutive years of SCADA data from four turbines located in a mountainous wind farm in southern China.