An iterative reassignment based energy-concentrated TFA post-processing tool and application to bearing fault diagnosis
An iterative reassignment based energy-concentrated TFA post-processing tool and application to bearing fault diagnosis
复制标题
基于迭代重新分配的能量集中TFA后处理工具及其在轴承故障诊断中的应用
DOI：
10.1016/j.measurement.2022.110953
复制
发表时间：
2022-03-12
期刊：
影响因子：
5.6
通讯作者：
Yu, Gang
中科院分区：
文献类型：
--
作者：
Li, Yueyang;Shi, Zhenjin;Yu, Gang
在时频分析（TFA）后处理方法中，时间重分配同步压缩变换（TSST）和二阶同步压缩变换（STSST）由于群时延（GD）估计精度的限制，对强GD调制信号的分析效果不理想。为了解决这一问题，提出了高阶TSST（HTSST）。然而，由于HTSST对加噪声信号敏感的事实，TF表示不如STSST集中。为了进一步提高TF的分辨率，本文提出了一种新的TFA后处理方法--局部最大高阶时间迭代同步压缩（LHTIS）方法。这种技术的基本思想可以分为两个步骤。首先，根据迭代过程，高阶频率变化信号的TF系数被重新分配到估计的高阶GD轨迹上。其次，通过最大化第一步迭代重新分配后的局部TF系数，我们重建了一个GD估计算子，以挤压所获得的TF结果。通过数值仿真验证了LHTIS方法的TF聚集和抗噪声性能。通过对滚动轴承故障信号的处理，验证了LHTIS的有效性。
Among the time frequency (TF) analysis (TFA) post-processing methods, limited by the accuracy of the estimated group delay (GD), the time-reassigned synchrosqueezing transform (TSST) and the second-order TSST (STSST) cannot provide a satisfied TFA performance for analyzing the signal with strong GD modulation. To settle this downside, the high order TSST (HTSST) is proposed. However, due to the truth that HTSST is sensitive to the plus-noise signal, the TF representation is not concentrated than the STSST. For further improving the TF resolution, a new TFA post-processing tool is proposed in this study, which named as local maximum high order time iterative synchrosqueezing (LHTIS) method. The basic idea of this technique can be divided as two steps. First, according to an iterative procedure, the TF coefficients of the high order frequency-varying signal are reassigned onto the estimated high order GD trajectory along time direction. Second, via maximizing the local TF coefficients after iterative reassignment in the first step, we rebuild a GD estimation operator to squeeze the obtained TF results. Numerical simulations are adopted to demonstrate the TF aggregation and the anti-noise property of LHTIS methodology. Via dealing with the fault signals of rolling bearings, the effectiveness of LHTIS is verified.