混响背景下基于分数阶傅里叶变换的自适应LMS滤波算法

针对强混响背景下经典的最小均方误差(Least Mean Square,LMS)滤波算法难以有效地实现信混分离的问题,提出一种基于分数阶傅里叶变换的自适应LMS算法。首先将混响信号和自适应LMS滤波算法中的参考信号进行分数阶傅里叶变换,寻找最优变换域,并在分数阶域进行带通滤波,然后将得到的信号进行分数阶傅里叶反变换,最后将基于正态分布曲线的变步长LMS算法应用于此混响条件下进行滤波。仿真和海试数据验证结果表明,在信混比为0 dB的情况下,算法仍可以有效地滤除混响,使信混比提高6dB。     

基于分数阶Fourier变换的机械变速过程信号分析

研究了旋转机械的变速测量过程，通过引入分数阶傅立叶变换（FRFT），推导证明了分数阶傅立叶变换对扫频信号具有良好的时频聚焦性，因此应用分数阶傅立叶变换来分析旋转机械的变速过程的测量信号，可以有效地解决机械转速不恒定对信号分析的影响，提高信号分析、状态监测以及故障诊断的准确性。     

宽带体制下混响和目标空时分布特征研究

传统的空时自适应处理(Space-Time Adaptive Processing,STAP)利用窄带混响和目标在空时平面的可分辨特征抑制窄带混响。对宽带混响和目标在空时平面分布特征进行研究,并通过理论推导,得出带宽对混响和目标空时分辨特征的影响公式。结果表明:宽带混响和目标在空时平面分布特征部分重叠,导致传统STAP效果不佳。在此基础上,借鉴STAP思想,并利用线性调频(Linear Frequency Modulation,LFM)信号在分数阶Fourier变换域上的聚焦性,分析了在空-分数阶Fourier域三维空间上抑制宽带LFM混响的可行性。     

分数阶傅立叶变换的线性调频信号降噪

本文在分数阶Fourier变换原理的基础上,提出了一种基于分数阶傅里叶变换的线性调频（LFM）信号的滤波方法,利用该变换等同于对信号在时频平面进行旋转,将混迭有噪声的信号以特定的旋转角作分数阶傅里叶变换,使得信号与噪声在变换域中的交迭达到最小：然后通过窄带通滤波器对LFM信号进行抽取,再经过分数阶傅里叶反变换,恢复出原信号。     

HHT在水雷目标特征提取中的应用

Hilbert-Huang变换（HHT）是近年来发展起来的一种处理非线性非平稳的时频分析方法。水声信号具有非线性非平稳的特点,把HHT方法应用于水声信号的分析中,将典型主动声纳信号的Hilbert边际谱和Hilbert谱分别与傅里叶幅度谱和传统的时频分析进行了比较。然后对水下目标回波的亮点模型作了仿真研究,并与小波变换作比较,最后用实际目标的回波作了分类研究。仿真与实际数据都证明了HHT在水下目标特征提取中的优越性。     

基于分数阶聚能带分析的微弱故障特征提取研究

提出了一种分数阶聚能带时频累加谱方法,快速实现长数据的时频分析,突出目标分量,用于提取变速器急加速过程微弱故障特征。根据变速器输入轴转速信号及传动比确定分数阶傅里叶变换（Fractional Fourier Transform, FRFT）最佳阶次,对变速器急加速过程振动信号进行最佳阶次FRFT,根据FRFT模值谱确定聚能带,计算分数阶聚能带时频累加谱,通过对比多组正常和故障数据的分数阶聚能带时频累加谱结果和阶比谱结果,验证该方法的有效性。试验结果表明：根据转速信号能快速、准确确定FRFT最佳阶次;选取聚能带内的FRFT结果进行时频分析,计算量小,分辨率高,分数阶聚能带时频累加谱具有聚焦和局部放大的特点, 能很好地突出目标分量,抑噪噪声,是提取变速器急加速过程信号微弱故障特征的有效方法。     

一种自适应Chirplet分解方法及在故障诊断中应用

提出了一种自适应Chirplet信号分解过程中的参数优化估计方法，该方法首先利用最大投影分解原理结合分数阶傅立叶变换和拟牛顿方法进行参数估计，然后利用最大期望方法，进一步进行参数的优化。将提出的方法与传统的时频分析方法如短时傅立叶谱图，Wigner分布进行对比分析，仿真结果表明，提出的方法具有很高的参数估计精度、很高的时频图分辨率和抗噪能力。说明了本文提出的方法中引入EM算法的必要性。又将提出的方法应用到轴承的故障诊断中，实验结果表明，提出的自适应Chirplet分解方法是非常有效的。     

基于Kaiser窗的分数阶Fourier变换与时频分析

分数阶Fourier变换作为传统Fourier变换的推广,与传统Fourier变换分析平稳信号类似,在实现对非平稳信号的时频分析过程中往往出现同样的频谱泄漏问题。为了提高分数阶Fourier变换与时频分析的精度,依据Kaiser窗可自由选择主瓣和旁瓣宽度的特性,提出一种基于Kaiser窗的分数阶Fourier变换算法,论述了Kaiser窗在分数阶Fourier变换中的作用原理,从理论上推导出一般信号基于Kaiser窗的分数阶Fourier变换解析时频表达式以及特性,最终得到非平稳信号的时频分布与时变结构参数识别算法。通过任意线性调频信号的仿真算例以及非平稳激励三层框架结构振动台试验,对结构进行瞬时频率识别和算法的验证。结果表明,瞬时频率识别值与理论值和试验结果吻合良好,Kaiser窗可以提高分数阶Fourier变换算法时频分析的精度,体现出该方法有一定的鲁棒性。     

分数阶时频局部二值模式谱在齿轮故障诊断中的应用

针对齿轮故障信号分析,提出采用分数阶时频LBP谱表达齿轮故障信号的时频特征。首先为克服S变换对高频信号的时频分辨性能差的不足,基于分数阶Fourier变换良好的时频旋转特性设计了一种分数阶S变换,用于获取齿轮信号的二维时频表示;然后引入局部二值模式(LBP)算子,将LBP算子作用于分数阶S变换时频图,提取分数阶时频LBP谱;最后结合"uniform"模式LBP的概念和类内类间距准则,对分数阶时频局部二值模式谱进行特征优选,用于表达齿轮故障特征。对5种不同状态的齿轮信号进行了分析,结果表明优选后的分数阶时频LBP谱具有较强的特征描述能力,是齿轮故障信号的一类新的有效特征参数。     

基于线性调频信号二次相位变换的主动声呐目标速度估计

分数阶傅里叶变换常用于主动声呐线性调频信号的分析,通过搜索变换参数可以估计目标的相对速度.但是由于其参数搜索计算量偏大,在自主式水下航行器等计算力有限的小平台难以实时处理.针对该问题提出了一种基于二次相位变换的线性调频信号速度估计方法,根据多普勒频移后线性调频信号的二次相位变换的解析解,分析了二次相位变换后变换域频谱展...     

Simplified fractional Fourier transforms

The fractional Fourier transform (FRFT) has been used for many years, and it is useful in many applications. Most applications of the FRFT are based on the design of fractional filters (such as removal of chirp noise and the fractional Hilbert transform) or on fractional correlation (such as scaled space-variant pattern recognition). In this study we introduce several types of simplified fractional Fourier transform (SFRFT). Such transforms are all special cases of a linear canonical transform (an affine Fourier transform or an ABCD transform). They have the same capabilities as the original FRFT for design of fractional filters or for fractional correlation. But they are simpler than the original FRFT in terms of digital computation, optical implementation, implementation of gradient-index media, and implementation of radar systems. Our goal is to search for the simplest transform that has the same capabilities as the original FRFT. Thus we discuss not only the formulas and properties of the SFRFT's but also their implementation. Although these SFRFT's usually have no additivity properties, they are useful for the practical applications. They have great potential for replacing the original FRFT's in many applications.     

Fractional Fourier transform for quasi-periodic Bloch functions

The fractional Fourier transform (FRFT) for quasi-periodic Bloch functions is studied. An isomorphism between square-integrable functions on the real line and quasi-periodic Bloch functions is used to extend existing work on the fractional Fourier transform for the former functions to the latter. The properties of the FRFT for quasi-periodic Bloch functions are discussed, and various numerical examples are presented.     

Pattern recognition using binary masks based on the fractional Fourier transform

This paper presents two new methodologies for pattern recognition invariant to position and rotation. The two methods use the correlation of signatures obtained using a binary rings mask created from the fractional Fourier transform (FRFT). An analysis of fragmented information was performed with 21 images of fossil diatom species. The methodologies were applied to 30 species of phytoplankton to test the performance when rotation distortion is present in a target. Also, a noise analysis was performed. Higher signature correlation values are achieved by working outside the Fourier plane by changing the order of the FRFT.     

Fractional Fourier transform optical system with programmable diffractive lenses

We present experimental results for a fractional Fourier transform (FRFT) system implemented with programmable lenses written onto a liquid-crystal spatial light modulator (LCSLM). Because the focal length can be changed, different orders of the FRFT can be obtained without changing the optical setup. The LCSLM can very easily implement more complicated operations, including the realization of simultaneous orders of the FRFT and anamorphic transforms.     

Propagation of four-petal Gaussian beams in apertured fractional Fourier transforming systems

Based on the Collins diffraction integral formula and the fact that a hard-edged-aperture function can be expanded into a finite sum of complex Gaussian functions, the propagation of a four-petal Gaussian beam passing through an apertured fractional Fourier transform (FRFT) optical system has been studied in detail. Some typical numerical examples are given to illustrate the properties of the four-petal Gaussian beam in the FRFT plane. The results indicate that the intensity distributions of the beam in the FRFT plane are closely related to not only the fractional order but also the beam parameters and the aperture parameters.     

Fractional Fourier transform for beams generated by Gaussian mirror resonator

Based on the definition of the fractional Fourier transform (FRFT) in the cylindrical coordinate system and the fact that a hard-edged-aperture function can be expanded into a finite sum of complex Gaussian functions, the propagation properties of beams generated by Gaussian mirror resonator passing through FRFT optical systems have been studied in detail, and some typical numerical examples are given to compare the results obtained by the approximate analytical method with those by the numerical integral method, and it is shown that our method can significantly improve the numerical calculation efficiency. Further, the results indicate that the normalized intensity distributions in the FRFT plane are closely related to the fractional order, truncation parameter and initial beam parameters. The variation period of normalized intensity distributions with p is 2 whatever the value of the truncation parameter δ is.     

基于分数阶功率谱熵的未知水声脉冲信号检测方法

水声侦察的核心问题是在无先验知识条件下捕获其他平台发射的脉冲信号,单频(Continuous Wave,CW)信号和调频(Frequency Modulation,FM)信号是常用的水声探测脉冲.功率谱熵算法能有效检测低信噪比的CW信号,但对FM信号性能不佳,分数阶傅里叶变换(Fractional Fourier Tr...     

Efficient computation of joint fractional Fourier domain signal representation

A joint fractional domain signal representation is proposed based on an intuitive understanding from a time-frequency distribution of signals that designates the joint time and frequency energy content. The joint fractional signal representation (JFSR) of a signal is so designed that its projections onto the defining joint fractional Fourier domains give the modulus square of the fractional Fourier transform of the signal at the corresponding orders. We derive properties of the JFSR, including its relations to quadratic time-frequency representations and fractional Fourier transformations, which include the oblique projections of the JFSR. We present a fast algorithm to compute radial slices of the JFSR and the results are shown for various signals at different fractionally ordered domains.     

Fractional Fourier transforms, symmetrical lens systems, and their cardinal planes

We study the relation between optical lens systems that perform a fractional Fourier transform (FRFT) with the geometrical cardinal planes. We demonstrate that lens systems symmetrical with respect to the central plane provide an exact FRFT link between the input and output planes. Moreover, we show that the fractional order of the transform has real values between 0 and 2 when light propagation is produced between principal planes and antiprincipal planes, respectively. Finally, we use this new point of view to design an optical lens system that provides FRFTs with variable fractional order in the range (0,2) without moving the input and output planes.