当前位置: X-MOL 学术Measurement › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Longitudinal synchroextracting transform: A useful tool for characterizing signals with strong frequency modulation and application to machine fault diagnosis
Measurement ( IF 5.6 ) Pub Date : 2021-06-16 , DOI: 10.1016/j.measurement.2021.109750
Yong Lv , Site Lv , Rui Yuan , Hewenxuan Li

In this paper, we propose a new time–frequency analysis (TFA) method termed longitudinal synchrosqueezing transform (LSST). As a post-processing time–frequency analysis method, the theory of this method is based on the Reassignment method (RM) and Fourier-based synchrosqueezing transform (FSST). LSST combines the advantages of RM and FSST in signal processing and avoids their drawbacks. Compared with RM and FSST, this method can achieve compact time–frequency representation (TFR) while retaining the ability of modes extraction and reconstruction. However, when addressing strong frequency modulation (FM) signals, there will still be energy smear in the TFR generated by LSST. Therefore, longitudinal synchroextracting transform (LSET) is further proposed to cope with signals with strong FM components and generate TFR with high energy concentration. On the basis of LSST, we only keep the energy most relevant to the instantaneous frequency (IF) of the signal through the Dirichlet function to obtain LSET. LSET estimates the signal's instantaneous frequency using a second-order approximation, which enables an accurate characterization of fast time-varying features of the signal at large window size range and it is more robust against noise. In addition, the proposed method is applied to the early fault diagnosis of rotor rub-impact and the fault feature extraction of rolling bearings under variable speed. Compared with other advanced TFA technologies, the simulated and measured signals can verify the effectiveness and competitiveness of the proposed method.



中文翻译:

纵向同步提取变换:用于表征具有强频率调制的信号并应用于机器故障诊断的有用工具

在本文中,我们提出了一种新的时频分析 (TFA) 方法,称为纵向同步压缩变换 (LSST)。作为一种后处理时频分析方法,该方法的理论基于重新分配方法(RM)和基于傅立叶的同步压缩变换(FSST)。LSST 结合了 RM 和 FSST 在信号处理方面的优点,避免了它们的缺点。与 RM 和 FSST 相比,该方法可以在保留模式提取和重建能力的同时实现紧凑的时频表示(TFR)。然而,在处理强调频 (FM) 信号时,LSST 产生的 TFR 中仍然会有能量拖尾。因此,进一步提出纵向同步提取变换(LSET)来处理具有强FM分量的信号并产生具有高能量集中的TFR。在LSST的基础上,我们只保留与信号瞬时频率(IF)最相关的能量,通过Dirichlet函数得到LSET。LSET 使用二阶近似估计信号的瞬时频率,这可以在大窗口尺寸范围内准确表征信号的快速时变特征,并且对噪声更加稳健。此外,该方法还应用于转子碰摩的早期故障诊断和变速滚动轴承的故障特征提取。与其他先进的 TFA 技术相比,模拟和测量的信号可以验证所提出方法的有效性和竞争力。我们只保留与信号的瞬时频率 (IF) 最相关的能量,通过 Dirichlet 函数来获得 LSET。LSET 使用二阶近似估计信号的瞬时频率,这可以在大窗口尺寸范围内准确表征信号的快速时变特征,并且对噪声更加稳健。此外,该方法还应用于转子碰摩的早期故障诊断和变速滚动轴承的故障特征提取。与其他先进的 TFA 技术相比,模拟和测量的信号可以验证所提出方法的有效性和竞争力。我们只保留与信号的瞬时频率 (IF) 最相关的能量,通过 Dirichlet 函数来获得 LSET。LSET 使用二阶近似估计信号的瞬时频率,这可以在大窗口尺寸范围内准确表征信号的快速时变特征,并且对噪声更加稳健。此外,该方法还应用于转子碰摩的早期故障诊断和变速滚动轴承的故障特征提取。与其他先进的 TFA 技术相比,模拟和测量的信号可以验证所提出方法的有效性和竞争力。这可以在大窗口尺寸范围内准确表征信号的快速时变特征,并且对噪声更加鲁棒。此外,该方法还应用于转子碰摩的早期故障诊断和变速滚动轴承的故障特征提取。与其他先进的 TFA 技术相比,模拟和测量的信号可以验证所提出方法的有效性和竞争力。这可以在大窗口尺寸范围内准确表征信号的快速时变特征,并且对噪声更加鲁棒。此外,该方法还应用于转子碰摩的早期故障诊断和变速滚动轴承的故障特征提取。与其他先进的 TFA 技术相比,模拟和测量的信号可以验证所提出方法的有效性和竞争力。

更新日期:2021-06-22
down
wechat
bug