基于FSWT和TT变换的滚动轴承故障诊断方法

Rolling bearing fault diagnosis method based on FSWT and TT-transform

文中提出了基于频率切片小波变换(Frequency Slice Wavelet Transform,FSWT)的一种双时域谱的分析方法,FSWT-TT变换(Frequency Slice Wavelet-TT Transform)。首先将非平稳的振动信号通过FSWT分解重构为若干个不同频率范围的分量,依据峭度准则确定包含故障信息最多的频带;再对该频带进行FSWT-TT变换,提取TT谱中的对角线序列;最后对该对角线序列进行包络分析,提取振动信号的故障特征。依次分析了仿真信号、试验信号和工程信号,结果表明:基于FSWT和TT变换的故障诊断方法能够有效地增强振动信号的冲击特征,识别故障特征信息,相比传统的TT(TT Transform)变换,该方法借助FSWT的选频特性,在滤波效果和计算效率上具有一定的优势。

In this paper,a dual time-domain spectrum analysis method based on frequency slice wavelet transform is proposed.Firstly,the non-stationary vibration signal is decomposed and reconstructed into several components in different frequency ranges through FSWT,and the frequency band containing the most fault information is determined according to kurtosis criterion.Then,the diagonal sequence in TT spectrum is extracted by FSWT-TT transform.Finally,the envelope analysis of the diagonal sequence is carried out to extract the fault characteristics of the vibration signal.The simulation signal,experimental signal and engineering signal are analyzed in turn.The results show that the fault diagnosis method based on FSWT and TT transform can effectively enhance the impact characteristics of vibration signal,identify the fault feature information,and solve the problems of vibration signal being easily disturbed and difficult to process.Compared with the TT(TT Transform)transformation proposed earlier,this method has certain advantages in filtering effect and computing efficiency by virtue of the frequency selection characteristics of FSWT.