Statistical Method of Low Efficiency Wind Turbine Generating Wind Power Curve Based on Operating Data

Abstract

Aiming at the low efficiency of wind turbines in service, this paper proposes a statistical method of low efficiency turbines based on wind power curve generated from operating data. According to the data distribution characteristics of the wind-power data scatter diagram, a data preprocessing model was established based on the quartile method and quadratic clustering method, which can effectively mark the data of shutdown, invalidity, undergeneration, overgeneration and power limit, and draw the wind power curve which can accurately represent the fan’s performance. On this basis, a mining method for low efficiency units based on the difference rate of lost electric quantity and power consistency coefficient is established, which can make statistics of high loss electric quantity units, units deviating from the designed power curve and units with degraded performance regularly, and evaluate the generation improvement before and after maintenance. Through the actual operation and maintenance data of all 814 wind turbines of a certain company, the effectiveness of the method is verified.