基于数据分布的风电机组叶片失速诊断方法研究  

Diagnosis Method of Blade Stall Based on Data Distribution for Wind Turbine

作　　者：尹业峰 张家友 娄斌 陈亚楠 YIN Yefeng;ZHANG Jiayou;LOU Bin;CHEN Yanan(CRRC Zhuzhou Institute Co.,Ltd.,Zhuzhou,Hunan 412001,China)

机构地区：[1]中车株洲电力机车研究所有限公司,湖南株洲412001

出　　处：《控制与信息技术》2025年第1期44-49,共6页CONTROL AND INFORMATION TECHNOLOGY

基　　金：湖南省十大技术攻关项目(2023GK1030)。

摘　　要：风电机组作为风能转换的关键设备,其运行稳定性和可靠性对于保障能源供应和减少维护成本至关重要。围绕其进行故障诊断与预警是业内普遍关注及研究的热点,特别是其中的叶片失速故障,直接影响风电机组的发电效率和安全性。对此,文章提出了一种用于诊断风电机组叶片失速情况的方法,其利用DBSCAN空间密度聚类算法初步筛选正常工况数据,再根据正常工况数据的空间分布规律获取数据下边界,依据数据下边界进行叶片失速与否的诊断。在若干个风场采用该方法进行了叶片失速诊断,实验结果显示,该方法能有效识别风电机组叶片失速情况,其平均诊断真正率达94.5%,漏检率为1.6%。该方法为风电机组叶片失速诊断提供了一种新的思路,可通过与主控系统协同响应,提升风电机组的发电效率,进而提高风电机组的发电性能及安全稳定性。Wind turbines serve as crucial equipment for wind energy conversion,and their operational stability and reliability are vital for ensuring energy supply and reducing maintenance costs.Fault diagnosis and warning around wind turbines is a widely concerned and researched hotspot in the industry,especially the blade stall faults,which directly affects the power generation efficiency and safety of wind turbines.This paper proposes a method for diagnosing blade stall conditions in wind turbines.Initially,operating data under normal conditions are screened using the density-based spatial clustering of applications with noise(DBSCAN)algorithm.This is followed by identifying the lower boundary of the data based on the spatial distribution pattern of the operating data under normal conditions.The final step involves diagnosing whether a blade stall has occurred according to the identified lower boundary.This method has been applied for diagnosis at several wind farms,and the results demonstrated its efficacy in identifying stalled blades in wind turbines,with an average true positive rate of 94.5% and an omission rate of 1.6%.The proposed method offers a novel approach of collaborative response with the main control system for diagnosing stalled blades in wind turbines.It can improve the power generation efficiency of wind turbines,thereby improving their power generation performance,safety,and stability.

关 键 词：风电机组 叶片失速 DBSCAN算法 数据分布 

分 类 号：TK89[动力工程及工程热物理—流体机械及工程]