中文摘要：

针对风电机组结构复杂、滚动轴承早期故障特征信号往往易受正常信号和现场噪声的干扰而不易识别以及信噪比低的问题,在原有盲源分离方法的基础上提出一种新的故障特征提取方法,该方法首先对原始振动信号进行包络解调分析和小波去噪,有效抑制信号的高频干扰,再采用基于最大信噪比的盲源分离方法对得到的小波包络解调信号进行分离,最后对分离后的信号进行频谱变换,从频谱图上可以清晰地观察出轴承的故障特征频率。实例分析表明,使用此方法对实测的风机主轴承故障振动信号进行分析能够有效提取出轴承的故障特征,有助于实现轴承的在线故障诊断。

英文摘要：

Aiming at the problems such as complicated wind turbine, difficult identification of early fault signal caused by interference of normal signals and noises, low signal to noise ratio and so on, a new fault feature extraction method of wind turbine main bearing is proposed based on blind source separation（BSS） method.Firstly,the original vibration signal is processed by analysis of envelope demodulation and wavelet denoise to restrict the high frequency interference effectively.Then the blind source separation method based on the maximum signal to noise ratio is applied to separate the signals that are gotten at the first step.At last the spectrum transform of separated signal is implemented; the fault feature frequencies are clearly shown in the spectrogram.And the example analysis shows that this method can extract the fault feature to realize the on-line fault diagnosis of the main bearings by analyzing the fault vibration signal.