基于奇异值分解及形态滤波的滚动轴承故障特征提取方法*

针对滚动轴承振动信号故障特征信息往往被强背景噪声淹没的问题,提出一种基于奇异值分解和形态滤波的振动信号故障特征提取方法。该方法利用信号时间序列重构的吸引子轨迹矩阵奇异值分布特征与信号自身特征的关系,选择轨迹矩阵中主要反映冲击信息明显的奇异值进行信号重构的方法来滤除信号中的平滑信号和部分噪声,获取带噪声的冲击信号;然后利用形态滤波能有效滤除脉冲干扰噪声的特点,反其道而行之,从而提取信号的冲击故障特征的方法,并将该方法应用于轴承的振动信号的故障特征提取。仿真与实例表明,该方法能有效提取强背景信号及噪声中的弱冲击特征信号,是一种有效的弱信号特征提取方法。

Fault feature extraction method of rolling bearing based on singular value decomposition and morphological filtering

Considering the strong noise background in fault feature information of vibration signal in rolling element bearing,this paper proposed a roller bearing fault feature extraction method based on SVD and morphological filters.This method made use of the relations between the singular value distribution of the time series track matrix of attractor and the signal characteristics to select the way of reconstruction of signal by most potential reflecting singular values.This way could filter smooth information and partial noise in the signal,and got impulse information with noise in the signal,then took the advantage of the feature that morphological filters was used to extract impulse feature in fault signal to act in opposition to pick out the extract impulse fault feature in signal and applied it to fault feature extraction of bearing in vibration signal.Results of experiment show that the presented method can be used for the Abstraction of the weak feature signal that mixed in the strong background noise,which is effective to Abstract weak feature signal.