摘要
风电机组发电机滚动轴承早期故障时,其机械振动信号包含复杂的信息成分,影响微弱故障信号的提取,导致故障识别非常困难。提出一种基于最小熵反褶积(Minimum Entropy Deconvolution,MED)和变分模态分解(Variational Mode Decomposition,VMD)结合的风电机组发电机滚动轴承故障特征提取方法。先应用MED算法对采集的振动信号进行降噪,突出被噪声所掩盖的故障冲击脉冲,后应用VMD算法对降噪后的信号进行分解,得到一系列固有模态函数(Intrinsic Mode Function,IMF),应用峭度和相关性最大准则选取包含故障信息量最多的分量,即敏感模态分量,最后对敏感分量进行包络谱分析,提取故障特征频率。实验分析和工程案例分析结果表明,该方法可有效提取故障特征。
At the early fault of the wind turbine generator rolling bearing , the mechanical vibration signal contains complex information components, which affect the extraction of weak fault signal and make it very difficult to identi- fy the fault. Wind turbine generator rolling beating exaction method is proposed based on the combination of Mini- mum Entropy Deconvolution （MED） and Variational Mode Decomposition （VMD）. Firstly, the collected vibration signal was denoised by MED algorithm, so as to highlight the fault impact characteristics which is obscured by noise. Then, the VMD algorithm is used to decompose the denoised signal to obtain a series of intrinsic mode func- tions （IMF） and kurtosis and correlation maximum criterion is selected to contain the most component of informa- tion content, namely, the sensitive modal component. Finally, the envelope spectrum of sensitive modal component is analyzed, and the fault characteristic frequency is extracted. The experimental and engineering case analysis re- subs show that the method can effectively extract fault characteristics.
出处
《黑龙江电力》
CAS
2017年第5期434-440,共7页
Heilongjiang Electric Power
基金
国家科技支撑计划项目(2015BAA06B03)
关键词
风电机组发电机
滚动轴承
最小熵解卷积
变分模态分解
特征提取
wind turbine generator
rolling bearing
Minimum Entropy Deconvolution （MED）
Variational Modal Decomposition （VMD）
feature extraction
