风电变桨减速器振动监测及变桨动作识别
Vibration Monitoring and Pitching Action Identification of Wind Power Pitching Reducer
武英杰 1代福峰 1田野 1赵瑞 2曲文涛 3刘少康 4辛红伟 1杨彦军 1王建国1
作者信息
- 1. 东北电力大学自动化工程学院 吉林,132012
- 2. 中广核陆丰核电有限公司 汕尾,516545
- 3. 吉林吉电新能源有限公司 长春,130117
- 4. 华北电力大学控制与计算机工程学院 北京,102200
- 折叠
摘要
由于变桨减速器随轮毂做圆周运动且伴随间歇性、回转动作,导致变桨有效振动数据识别存在困难,针对此问题,提出了粒子群优化变分模态分解结合滑动中值滤波的变桨动作识别方法.首先,通过现场变桨减速器振动信号分析,将其划分为静态未变桨、静态变桨、动态变桨和动态未变桨4部分,提出基于包络信号的变桨动作识别思路;其次,针对减速器随轮毂旋转导致的正弦分量和趋势分量,利用优化后的变分模态分解进行去除;然后,提出基于信号包络的变桨动作识别思路,采用滑动中值滤波平滑包络信号消除结构激振产生的脉冲干扰;最后,利用静态未变桨数据和3σ准则计算阈值,将平滑后的包络信号曲线与该阈值比较实现变桨动作识别.现场应用表明,该方法可准确识别变桨动作,且与其他方法相比具有明显优势,为风电机组变桨振动识别与状态监测提供参考.
Abstract
Because the pitch reducer moves circularly with the hub and rotates intermittently,it is difficult to identify the effective vibration data of pitch reducer.In this paper,a pitch motion identification method combin-ing particle swarm optimization variational mode decomposition(PSO-VMD)and moving median filtering(MMF)is proposed.Firstly,the vibration signal of the pitch reducer in the field is analyzed and divided into four parts:static unpitch,static pitch,dynamic pitch and dynamic unpitch.In view of the low-frequency compo-nents in the signal,the PSO-VMD is used to decompose the signal to eliminate the sinusoidal component and trend component caused by the rotation of the reducer with the hub.The envelope signal-based pitch action rec-ognition idea is proposed,and the MMF is used for smoothing the envelope signal to eliminate the impulse inter-ference generated by structural excitation.Finally,the static unpitch data and the 3σ criterion are used to calcu-late the threshold value,and the smoothed envelope signal curve is compared with the threshold value to achieve the pitching action.The field application shows that this method can accurately identify the pitch action,and has obvious advantages compared with other methods,providing a reference for wind turbine pitch vibration identifi-cation and condition monitoring.
关键词
信号采集/变桨减速器/间歇性动作/粒子群算法/变分模态分解/变桨识别Key words
signal acquisition/pitch reducer/intermittent motion/particle swarm optimization/variational mode decomposition/pitch recognition引用本文复制引用
基金项目
吉林省科技发展计划重点科技研发项目(20220203077SF)
吉林省教育厅科研项目(JJKH20230129KJ)
出版年
2024
NETL
NSTL
万方数据