基于边际谱的优越性在风电机组轴承故障诊断的应用

为了保障风力发电机组的安全运行,对其进行状态监测和故障诊断是非常重要的。对于非平稳信号,传统方法傅里叶变换(FFT)不能很好地进行分析。文中提出了在希尔伯特黄变换的基础上,对希尔伯特谱进行积分,求取边际谱的方法。以某台风力发电机组轴承内圈故障为例,将该方法与传统FFT,希尔伯特时频谱进行比较。结果表明,边际谱可以很好地检测轴承故障特征频率,也验证了该方法对非平稳信号分析的有效性。

Application of Marginal Spectrum in Fault Diagnosis of Wind Turbine Bearing

In order to ensure the safe operation of the wind turbine,condition monitoring and fault diagnosis is very important.The traditional method of Fast Fourier transform(FFT)can not analyze non-stationary signal very well.This paper proposed a new method for getting the marginal spectrum by integrating the Hilbert spectrum on the basis of Hilbert-Huang transform.Taking the bearing inner ring fault of a wind turbine as an example,the paper compares this method with the traditional FFT method and Hilbert time-frequency spectrum.The result shows that marginal spectrum can detect bearing fault characteristic frequency accurately,and verifies that it is effective to analyze non-stationary signal by using the proposed method.