时频分析在旋转机械状态监测及故障诊断中的应用

本文简要介绍了时频分析方法的发展及其在旋转机械状态监测、故障诊断领域中的应用。并以一个实际监测系统为例，对其检测信号应用时频分析的方法进行了数据处理，对用传统傅立叶变换处理的结果与采用时频分析方法处理的结果进行了对比分析．     

On the joint time-frequency characteristics of chemical oscillations

A method of time-frequency transformations of nonstationary signals has been discussed and applied to the analysis of oscillatory Belousov-Zhabotinsky (BZ) reaction. Selected registers of BZ reactions, that represent regular and complex dynamic behavior, have been submitted to the detailed spectral analysis using the short time Fourier transform (STFT). It has been demonstrated that the STFT is an accurate and suitable method for decomposition of chemical oscillations in the joint time and frequency domain. The energy distribution via STFT, presented in this article, revealed some specific characteristics of transitions between different regimes in the closed system of BZ reaction.     

多速率STFT超宽带信号瞬时频率估计研究

提出多速率短时傅里叶变换(Multi Rate Short Time Fourier Transform,MR-STFT)瞬时频率估计算法,提高了超宽带信号瞬时频率估计精度。该方法将多速率信号处理算法与短时傅里叶变换(STFT)技术相结合,兼顾采样频率和被测频率,将宽频范围进行分段采样,对分段处理结果进行拟合,构成多速率STFT算法,实现超宽带信号瞬时频率的高精度测量。论文通过对仿真信号和实测信号进行处理,研究了方法的可行性和频率估计精度,结果表明MR-STFT算法较大提高了超宽带信号瞬时频率估计精度,尤其对低信噪比的超宽带信号效果显著。     

Learning to short-time Fourier transform in spectrum sensing

The future wireless communication will come up with a strict requirement on high spectral efficiency, developing novel algorithms for spectrum sensing with deep sensing capability will be more challenging. However, traditional expert feature-based spectrum sensing algorithms are lack of sufficient capability of self-learning and adaptability to unknown environments and complex cognitive tasks. To address this problem, we propose to build up a deep learning network to learn short time-frequency transformation (STFT), a basic entity of traditional spectrum sensing algorithms. Spectrum sensing based on the learning to STFT network is supposed to automatically extract features for communication signals and makes decisions for complex cognitive tasks meanwhile. The feasibility and performances of the designed learning network are verified by classifying signal modulation types in deep spectrum sensing applications.     

Output-only identification and dynamic analysis of time-varying mechanical structures under random excitation: A comparative assessment of parametric methods

This article addresses the problem of parametric time-domain identification and dynamic analysis for time-varying (TV) mechanical structures under unobservable random excitation. The methods presented are based on time-dependent autoregressive moving average (TARMA) models, and are classified according to the mathematical structure imposed on the TV parameter evolution as unstructured parameter evolution, stochastic parameter evolution, and deterministic parameter evolution. The features and relative merits of each class are outlined. A representative method from each is then assessed through its application to the identification and dynamic analysis of a laboratory TV structure consisting of a beam with a mass moving on it. The results are mutually compared and contrasted to those obtained through “frozen-configuration” (multiple experiment) baseline identification.     

窗函数在扬声器异常声客观检测中的影响

该文研究了基于短时傅里叶变换的扬声器异常声检测方法中,窗函数对特征距离阈值设定的影响。以连续对数扫频信号激励被测扬声器单元,并对响应信号进行短时傅里叶变换。通过对比不同窗函数下,被测单元之间的特征距离来寻找合适的窗函数。结果表明,不同窗函数下得到的特征距离具有不同的离散度;选择合适的窗函数,可使特征距离具有较大离散度,从而更合理地设定阈值。     

Application of least-square estimation in white-light scanning interferometry

White-light scanning interferometry (WLSI) is a powerful tool for investigating the profile of a test object that contains sharp steps. Due to the light source used in WLSI system, it is able to overcome phase ambiguity problem, which is often encountered in monochromatic interferometry. In this paper, a new algorithm based on least-square estimation using short-time Fourier transform (STFT) is proposed to measure the profile of a test object. STFT is used to extract the peak position of the coherence envelope of a white-light interference signal and retrieve the corresponding phase values simultaneously at first. A complex phasor (CP) method is introduced to further reduce the phase noise. Then, the phase values at several positions around are utilized to achieve a more accurate peak position based on least-square line fitting. Both simulated and experimental results show that the proposed algorithm is able to accurately measure the profile of a test object.     

基于短时傅立叶变换的高频地波雷达射频干扰抑制

提出了一种适合于高频地波雷达射频干扰滤除的方法,该方法基于短时傅立叶变换(STFT)的基本原理,针对单一频率的射频干扰在线性调频雷达中的特点,利用STFT在时频面上检测出射频干扰,通过时域滤波和建立AR模型产生预测数据滤除射频干扰的影响．经过实测数据的检验,信噪比提高了将近10dB,表明该方法能有效地解决线性调频雷达中射频干扰的抑制问题．     

针对磁共振射频脉冲的时频域分析方法比较研究

射频脉冲可实现样本自旋体系的精确操控,进而产生预期的核磁共振(NMR)信号,在NMR信号产生过程中扮演重要角色.该文分别采用短时傅里叶变换(STFT)、连续小波变换(CWT)和维格纳-威利分布(WVD)几种时频域分析方法对射频脉冲(优化形状脉冲)进行特性分析和比较.结果表明,三种方法各自具有优缺点,结合各自优势对射频脉冲进行各种方法分析,可以更好地理解复杂脉冲的幅度、相位特性在时频域的分布情况.该文的研究方法将为直观理解复杂射频脉冲对自旋体系的作用机制提供参考.     

基于模式能量比的海底声衰减系数反演

简正波的分离性能对简正波频散曲线、模式能量比等物理量的提取至关重要,这些物理量可以用来反演海底地声参数。本文分析了短时傅里叶变换(STFT)和warping变换两种方法在简正波分离上的性能的差异,并利用该两种方法提取模式能量比来反演海底声衰减系数。仿真实验和2018年冬季北海实验数据处理表明,warping变换在简正波分离上具有较大的优势,尤其是在近距离接收上。