欢迎访问《太阳能学报》官方网站,今天是 分享到:
ISSN 0254-0096 CN 11-2082/K

太阳能学报 ›› 2022, Vol. 43 ›› Issue (4): 409-417.DOI: 10.19912/j.0254-0096.tynxb.2020-0816

• 电化学储能安全性与退役动力电池梯次利用关键技术专题 • 上一篇    下一篇

数据驱动的风电机组变桨系统状态监测

金晓航1,2, 泮恒拓1, 徐正国3   

  1. 1.浙江工业大学机械工程学院,杭州 310023;
    2.特种装备制造与先进加工技术教育部/浙江省重点实验室(浙江工业大学), 杭州 310023;
    3.浙江大学控制科学与工程学院,杭州 310027
  • 收稿日期:2020-08-17 出版日期:2022-04-28 发布日期:2022-10-28
  • 通讯作者: 金晓航(1981—),男,博士、副教授,主要从事新能源机电装备智能运维等方面的研究。xhjin@zjut.edu.cn
  • 基金资助:
    国家重点研发计划(2022YFE0198900); 宁波市自然科学基金(2021J038); 工业控制技术国家重点实验室开放课题(ICT2021B43)

CONDITION MONITORING OF WIND TURBINE PITCH SYSTEM USING DATA-DRIVEN APPROACH

Jin Xiaohang1,2, Pan Hengtuo1, Xu Zhengguo3   

  1. 1. College of Mechanical Engineering, Zhejiang University of Technology, Hangzhou 310023, China;
    2. Key Laboratory of Special Purpose Equipment and Advanced Processing Technology, Ministry of Education and Zhejiang Province, Zhejiang University of Technology, Hangzhou 310023, China;
    3. College of Control Science and Engineering, Zhejiang University, Hangzhou 310027, China
  • Received:2020-08-17 Online:2022-04-28 Published:2022-10-28

摘要: 针对数据采集与监视控制(SCADA)系统存在误报、故障报警滞后等问题,提出一种基于单分类模型的风电机组变桨系统在线状态监测方法。首先,从SCADA数据中提取出与变桨系统相关的特征参数并进行特征重构以进一步提取出更值得关注的桨叶之间的差异化信息。其次,基于单分类支持向量机对历史数据的分析确定变桨系统运行数据的健康边界,进而通过判断实时运行数据是否位于健康边界内部来辨别变桨系统当前的运行状态。最后,以变桨系统的实际工程案例分析验证所提方法的有效性。

关键词: 风电机组, SCADA系统, 状态监测, 变桨系统, 单分类支持向量机

Abstract: To address the problems of a high false alarm rate and delayed fault alarm in supervisory control and data acquisition (SCADA) systems, an online condition monitoring method for wind turbine pitch system is proposed. Firstly, the parameters related to the pitch system are extracted from the SCADA data, and these parameters are reconstructed to extract the differential information between the blades. Secondly, the health boundary of the working status of pitch system is found based on the analysis of the historical data by using one-class support vector machine, and then the online health status of the pitch system can be assessed by judging whether the real-time data is within the health boundary. Finally, the effectiveness of the proposed method is verified by two actual engineering cases of the wind turbine’s pitch system.

Key words: wind turbines, SCADA system, condition monitoring, pitch system, one-class support vector machine

中图分类号: