基于Logistic函数与分位数回归的风电机组功率曲线建模方法  

Wind turbine power curve modeling method with Logistic functions based on quantile regression

作　　者：王勃 孙勇 李振元 王铮 荆博 WANG Bo;SUN Yong;LI Zhenyuan;WANG Zheng;JING Bo(State Key Laboratory of Operation and Control of Renewable Energy&Storage System,China Electric Power Research Institute,Beijing 100192,China;State Grid Jilin Electric Power Company Limited,Changchun 130021,China;Beihang University,Beijing 100191,China)

机构地区：[1]中国电力科学研究院有限公司新能源与储能运行控制国家重点实验室,北京100192 [2]国网吉林省电力有限公司,长春130021 [3]北京航空航天大学,北京100191

出　　处：《电测与仪表》2025年第3期112-120,共9页Electrical Measurement & Instrumentation

基　　金：国家电网有限公司科技项目(5100-201929190A-0-0-00)。

摘　　要：风电机组功率曲线建模是风电功率预测、状态监测、性能评估的关键环节。文章提出了一种基于lo-gistic函数和分位数回归的风电机组功率曲线建模算法。为解决风电功率的不确定性,文中在logistic函数中嵌入了分位回归损失函数,建立了分位数回归logistic模型(quantile regression logistic function,QRLF),并采用了三种优化算法进行优化;为降低原始数据中异常值的影响,提出了基于QRLF算法的自适应异常筛选方法;在三个风电场的SCADA(supervisory control and data acquisition)数据中进行了实例验证。文中采用五种评价指标对所提方法进行评估。结果表明,相比传统的风电机组功率曲线建模方法,文中所提方法可以同时提供较好的确定性功率曲线和概率性功率曲线结果。The wind turbine power curve(WTPC)modeling is of great significance for wind power forecasting,condition monitoring,and performance assessment.This paper proposes a novel WTPC modelling method with lo-gistic functions based on quantile regression(QRLF).Quantile regression logistic functions(QRLF)is embedded in the logistic function,and the QRLF model was established,so that the proposed method can describe the uncer-tainty of wind power.In order to reduce the effect of outliers in original data,an adaptive outlier filtering method is developed based on QRLF.Supervisory control and data acquisition(SCADA)data collected from wind turbines in three wind farms are used to evaluate the performance of the proposed method.Five evaluation metrics are applied for the comparative analysis.Compared with typical WTPC models,QRLF has better fitting performance in both de-terministic and probabilistic power curve modeling.

关 键 词：logistic函数 分位数回归 异常筛选 风电机组功率曲线 

分 类 号：TM933[电气工程—电力电子与电力传动]