Regularization for improving the deconvolution in real-time near-field acoustic holography

Near-field acoustic holography is a measuring process for locating and characterizing stationary sound sources from measurements made by a microphone array in the near-field of the acoustic source plane. A technique called real-time near-field acoustic holography (RT-NAH) has been introduced to extend this method in the case of nonstationary sources. This technique is based on a formulation which describes the propagation of time-dependent sound pressure signals on a forward plane using a convolution product with an impulse response in the time-wavenumber domain. Thus the backward propagation of the pressure field is obtained by deconvolution. Taking the evanescent waves into account in RT-NAH improves the spatial resolution of the solution but makes the deconvolution problem "ill-posed" and often yields inappropriate solutions. The purpose of this paper is to focus on solving this deconvolution problem. Two deconvolution methods are compared: one uses a singular value decomposition and a standard Tikhonov regularization and the other one is based on optimum Wiener filtering. A simulation involving monopoles driven by nonstationary signals demonstrates, by means of objective indicators, the accuracy of the time-dependent reconstructed sound field. The results highlight the advantage of using regularization and particularly in the presence of measurement noise.     

A robust super-resolution approach with sparsity constraint in acoustic imaging

Acoustic imaging is a standard technique for mapping acoustic source powers and positions from limited observations on microphone sensors, which often causes an ill-conditioned inverse problem. In this article, we firstly improve the forward model of acoustic power propagation by considering background noises at the sensor array, and the propagation uncertainty caused by wind tunnel effects. We then propose a robust super-resolution approach via sparsity constraint for acoustic imaging in strong background noises. The sparsity parameter is adaptively derived from the sparse distribution of source powers. The proposed approach can jointly reconstruct source powers and positions, as well as the background noise power. Our approach is compared with the conventional beamforming, deconvolution and sparse regularization methods by simulated, wind tunnel data and hybrid data respectively. It is feasible to apply the proposed approach for effectively mapping monopole sources in wind tunnel tests.     

A covariance fitting approach for correlated acoustic source mapping

Microphone arrays are commonly used for noise source localization and power estimation in aeroacoustic measurements. The delay-and-sum (DAS) beamformer, which is the most widely used beamforming algorithm in practice, suffers from low resolution and high sidelobe level problems. Therefore, deconvolution approaches, such as the deconvolution approach for the mapping of acoustic sources (DAMAS), are often used for extracting the actual source powers from the contaminated DAS results. However, most deconvolution approaches assume that the sources are uncorrelated. Although deconvolution algorithms that can deal with correlated sources, such as DAMAS for correlated sources, do exist, these algorithms are computationally impractical even for small scanning grid sizes. This paper presents a covariance fitting approach for the mapping of acoustic correlated sources (MACS), which can work with uncorrelated, partially correlated or even coherent sources with a reasonably low computational complexity. MACS minimizes a quadratic cost function in a cyclic manner by making use of convex optimization and sparsity, and is guaranteed to converge at least locally. Simulations and experimental data acquired at the University of Florida Aeroacoustic Flow Facility with a 63-element logarithmic spiral microphone array in the absence of flow are used to demonstrate the performance of MACS.     

Imaging of directional distributed noise sources

This study relates to the acoustic imaging of noise sources that are distributed and strongly directional, such as in turbulent jets. The goal is to generate high-resolution noise source maps with self-consistency, i.e., their integration over the extent of the noise source region gives the far-field pressure auto-spectrum for a particular emission direction. Self-consistency is possible by including a directivity factor in the formulation of the source cross-spectral density. The resulting source distribution is based on the complex coherence, rather than the cross-spectrum, of the measured acoustic field. For jet noise, whose spectral nature changes with emission angle, it is necessary to conduct the measurements with a narrow-aperture array. Three coherence-based imaging methods were applied to a Mach 0.9 turbulent jet: delay-and-sum beamforming; deconvolution of the beamformer output; and direct spectral estimation that relies on minimizing the difference between the measured and modeled coherences of the acoustic field. The delay-and-sum beamforming generates noise source maps with strong spatial distortions and sidelobes. Deconvolution leads to a five-fold improvement in spatial resolution and significantly reduces the intensity of the sidelobes. The direct spectral estimation produces maps very similar to those obtained by deconvolution. The coherence-based noise source maps, obtained by deconvolution or direct spectral estimation, are similar at small and large observation angles relative to the jet axis.     

水声信号稀疏重构的高阶累积量波达方向估计

高阶累积量具有高斯噪声抑制和阵元扩展特性,将高阶累积量引入水声信号的方位估计中,提出了离格稀疏贝叶斯学习重构的高阶累积量测向算法。该方法利用高阶累积量对高斯噪声的自然盲性,计算阵列信号四阶累积量来滤除高斯噪声,使阵元在原来的结构上扩展了一倍;并构造出选择矩阵剔除了四阶累积量中的冗余项,能再一次的扩展阵元,得到的新观测模型具有更好的统计性能;最后利用空域稀疏性,推导出四阶累积量下的离格稀疏表示模型,采用贝叶斯学习解算出源信号的最大后验概率,实现了目标方位估计。数值仿真和海试实验数据表明,该方法在相邻声源方位间隔为4°的情况下分辨概率可达到95%以上,在信噪比大于-5 dB时目标方位估计的均方根误差在1°以内,可显著抑制背景噪声干扰,在多声源密集分布条件下也能准确、稳健的对水声目标方位进行估计。      

格林函数解卷积处理的阵不变量浅海无源定位

水下声源无源定位是声呐技术重要的研究方向.针对水下声源无源定位问题,本文提出了一种基于格林函数解卷积处理的阵不变量无源定位方法.该方法使用盲解卷积算法从水平阵接收信号中提取时域格林函数,然后采用空域解卷积方法处理得到的时域格林函数,获得波束时间偏移,从波束时间偏移中计算得到阵不变量,解算目标距离,从而实现声源定位.区别...     

近场反卷积聚焦波束形成声图测量

为了提高声图测量中对多个声源的分辨能力和定位精度,给出了一种近场二维反卷积聚焦波束形成声图测量方法。推导了水下声图测量的广义卷积模型,根据声图测量中点传播函数移变但可预测的特点,通过预存点传播函数字典的方式,将波束形成过程中的卷积问题转化成叠加积分问题,并应用二维Richardson-Lucy迭代算法实现了二维移变模型情况下的近场二维反卷积求解,从而实现高分辨声图测量。通过仿真和海试对比了反卷积、常规声图测量和MVDR声图测量的性能,结果表明反卷积算法在500次迭代情况下聚焦峰尺度小于另外两种算法的1/2,旁瓣级下降超过6 dB.      

Detection of number of sources and DOA estimation based on the combined information processing of pressure and particle velocity using acoustic vector sensor array

In order to solve the problem of DOA(direction of arrival)estimation of underwater remote targets,a novel subspace-decomposition method based on the cross covariance matrix of the pressure and the particle velocity of acoustic vector sensor arrays(AVSA)was proposed. Whereafter,using spatio-temporal virtual tapped-delay-line,a new eigenvector-based criteria of detection of number of sources and of subspace partition is also presented.The theoretical analysis shows that the new source detection and direction finding method is different from existing AVSA based DOA estimation methods using particle velocity information of acoustic vector sensor(AVS)as an independent array element.It is entirely based on the combined information processing of pressure and particle velocity,has better estimation performance than existing methods in isotropic noise field.Computer simulations with data from lake trials demonstrate,the proposed method is effective and obviously outperforms existing methods in resolution and accuracy in the case of low signal-to-noise ratio(SNR).     

Direction-of-arrival estimation for extensible acoustic vector sensor array

Aiming at high-resolution estimation of the direction-of-arrival of closely-spaced sources at low signal-to-noise ratio regions, this paper proposes a DOA estimation algorithm that is suitable for an extensible acoustic vector sensor array. Taking the 3D array composed of the minimum number(four) of acoustic vector sensors as the acquisition module, a virtual array having the same structure as the original array structure is extended in the three-dimensional space based on the aperture expansion characteristic of higher-order cumulants. The virtual array and the real array can construct a matrix with rotational invariance, which contains the angular information for estimating DOA. The Cramer-Rao bound of the algorithm are derived. We analyze the influence of SNR, the number of snapshots and the elevation angle on the performance of the algorithm. Simulation results show that the proposed algorithm has better noise suppression ability and higher resolution in DOA estimation than the conventional ESPRIT algorithm using the acoustic vector array. Experiments are conducted to validate the proposed algorithm.     

声矢量圆阵宽带相干目标MVDR方位估计*

针对宽带相干目标的远程探测问题，本文提出一种基于声压振速联合处理和矢量重构的声矢量圆阵MVDR波束形成方法。该方法利用相位模态变换技术，将声矢量圆阵变换为与信号频率无关的虚拟线阵，并构建虚拟线阵声压与组合振速的互协方差矩阵，利用声压与振速各分量间的空间相关性有效地抑制各向同性环境噪声；并对宽带相干信号的互协方差矩阵进行矢量重构，即将最大特征值对应的特征向量划分为相互重叠的子向量，从而构建前/后向Hermitian矩阵；最后，基于MVDR波束形成器实现宽带相干目标的方位估计。仿真计算和实验数据处理结果表明，该方法具较强的解相干能力和噪声抑制能力以及较高的方位估计性能。     

基于压缩感知的矢量阵聚焦定位方法

本文针对噪声源近场定位识别问题,利用声源分布在空间域具有稀疏性,在压缩感知理论框架下建立了新体系下的矢量阵聚焦波束形成方法,用于解决同频相干声源的定位识别问题.新方法可在小快拍下准确获得噪声源的空间位置,且不损失对噪声源贡献相对大小的评价能力.通过详细的理论推导、仿真分析和试验验证,证明了基于压缩感知的矢量阵聚焦定位新方法本质上实现了l1范数正则化求解下的波形恢复和空间谱估计,因此具有较高的定位精度,较强的相干声源分辨能力、准确的声源贡献相对大小评价能力以及较高的背景压制能力,可应用于水下复杂噪声源的定位识别.     

一种矢量水听器基阵模拟器的设计

矢量水听器的每个阵元同时测量声场中声矢量和质点振速的3个分量,相对于声压水听器阵来说,矢量阵获取声场中更多的信息。利用矢量阵所获得速度场的信息可以去除目标方位估计中的 180°模糊。模拟器可以模拟实际海洋环境中目标的辐射特性和噪声特性,应用模拟器可以有效地缩短声纳的研制周期。本文提出一种矢量水听器基阵模拟器的设计方案,该方案解决了矢量阵中宽带信号的90°移相问题、时延精确控制问题和宽带噪声的谱状控制问题。     

Fast implementation of sparse iterative covariance-based estimation for source localization

Fast implementations of the sparse iterative covariance-based estimation (SPICE) algorithm are presented for source localization with a uniform linear array (ULA). SPICE is a robust, user parameter-free, high-resolution, iterative, and globally convergent estimation algorithm for array processing. SPICE offers superior resolution and lower sidelobe levels for source localization compared to the conventional delay-and-sum beamforming method; however, a traditional SPICE implementation has a higher computational complexity (which is exacerbated in higher dimensional data). It is shown that the computational complexity of the SPICE algorithm can be mitigated by exploiting the Toeplitz structure of the array output covariance matrix using Gohberg-Semencul factorization. The SPICE algorithm is also extended to the acoustic vector-sensor ULA scenario with a specific nonuniform white noise assumption, and the fast implementation is developed based on the block Toeplitz properties of the array output covariance matrix. Finally, numerical simulations illustrate the computational gains of the proposed methods.     

浅海波导环境失配下的协方差矩阵拟合稳健匹配场声源功率估计

实际浅海波导中环境噪声为相干噪声,最小方差匹配场声源功率估计方法能在相干噪声背景下准确估计声源辐射功率,但该方法受环境不确定性影响较大;此外,由于最小方差匹配场声源功率估计方法使用信号幅度作为中间量估计声源功率,信号幅度估计误差会二次放大并传递到声源功率估计结果中。本文提出一种协方差矩阵拟合稳健最小方差匹配场声源功率估计方法,该方法引入信道传递函数不确定集,结合协方差矩阵拟合思想将声源功率估计问题建模为在信道传递函数不确定集约束下对函数取极值的问题,使用Lagrange乘子法求解该问题得到信道传递函数估计值和声源辐射功率估计值。环境失配影响声源辐射功率估计性能的根本原因在于信道传递函数偏差较大,协方差矩阵拟合稳健匹配场声源功率估计方法有效减小了环境失配时信道传递函数的偏差,从而显著提升环境失配稳健性。此外,该方法使用权值直接估计声源功率,无需使用信号幅度作为中间量,避免了估计误差的传递。仿真验证了协方差矩阵拟合稳健匹配场声源功率估计方法的环境失配稳健性。     

一种深度学习的立体阵波达方向估计方法

针对常规波束形成主瓣宽且目标分辨能力低的问题,提出一种基于深度卷积神经网络的波达方向估计方法。算法使用常规波束形成计算二维空间功率谱,将预处理后的空间功率谱图输入深度卷积神经网络。该文利用神经网络学习解卷积映射关系,输出主瓣宽度更窄的空间功率谱图,从而实现高分辨率二维波达方向估计。该算法对阵列结构没有限制,适用于立体阵。仿真结果表明该文方法在不同目标个数、快拍数及信噪比参数下均能准确估计目标方向。该文方法目标分辨能力优于常规波束形成方法。在低快拍情况下,目标方向估计误差低于自适应波束形成方法。     

Improving the resolution of three-dimensional acoustic imaging with planar phased arrays

This paper examines and compares two methods of improving the quality of three-dimensional beamforming with phased microphone arrays. The intended application is the detection of aerodynamic noise sources on wind turbines. Both methods employ Fourier based deconvolution. The first method involves a transformation of coordinates that tends to make the response to a point source, the point spread function, more shift invariant. The result is a significant improvement in sound source imaging in the transformed coordinate system. However, the inverse transformation to Cartesian coordinates introduces range dependent resolution limitations because of the irregular distribution of the focal points. The second method combines the transformation of coordinates with an alternative scanning technique. This method can be used in near field three-dimensional acoustic imaging to produce maps free of sidelobes and with constant resolution. The robustness of the proposed methods is validated both with computer simulations and experimentally.     

Extension of deconvolution algorithms for the mapping of moving acoustic sources

Several deconvolution algorithms are commonly used in aeroacoustics to estimate the power level radiated by static sources, for instance, the deconvolution approach for the mapping of acoustic sources (DAMAS), DAMAS2, CLEAN, and the CLEAN based on spatial source coherence algorithm (CLEAN-SC). However, few efficient methodologies are available for moving sources. In this paper, several deconvolution approaches are proposed to estimate the narrow-band spectra of low-Mach number uncorrelated sources. All of them are based on a beamformer output. Due to velocity, the beamformer output is inherently related to the source spectra over the whole frequency range, which makes the deconvolution very complex from a computational point of view. Using the conventional Doppler approximation and for limited time analysis, the problem can be separated into multiple independent problems, each involving a single source frequency, as for static sources. DAMAS, DAMAS2, CLEAN, and CLEAN-SC are then extended to moving sources. These extensions are validated from both synthesized data and real aircraft flyover noise measurements. Comparable performances to those of the corresponding static methodologies are recovered. All these approaches constitute complementary and efficient tools in order to quantify the noise level emitted from moving acoustic sources.     

声矢量锥形阵的高阶累积量波达方向估计

为解决信源在较低信噪比情况下的测向分辨率问题,提出阵列可扩展的声矢量锥形阵测向算法。算法基于四阶累积量的阵列扩展和高斯噪声抑制特性,计算声矢量传声器不同输出分量的四阶累积量,使其在三维方向上扩展与原阵型结构相同的虚拟阵,从而构造包含角度信息的旋转不变矩阵进行测向。推导给出了算法的克拉美罗界,理论分析了算法性能受信噪比、采样快拍以及入射声源俯仰角的影响。仿真实验验证了该算法较常规声矢量阵ESPRIT算法有更优的噪声抑制能力及更高分辨的DOA估计性能。      

Uncertainty of array-based measurement of radiated and absorbed sound intensity

Patch near-field acoustic holography (NAH) coupled with an array of sound intensity probes allows separating the sound field incident on a surface from the one radiated by the surface itself. Although the measurement principle has been successfully used to separate the noise source contribution from disturbing sources and/or noise reflections, the method accuracy has not been investigated in the literature. We describe the results of experiments meant to evaluate the uncertainty in the identification of noise radiated by vibrating panels with different absorption characteristics in presence of an incident acoustic radiation using the statistically optimized near-field acoustic holography. Measurement errors were evaluated through tests performed in controlled acoustic conditions. Results evidenced that the measurement uncertainty depends on the accuracy of the microphone array positioning and on the incident sound field. These conclusions were in agreement with the results obtained by simulations in the phase of instrument optimization.     

A fast signal subspace approach for the determination of absolute levels from phased microphone array measurements

Phased microphone arrays are used in a variety of applications for the estimation of acoustic source location and spectra. The popular conventional delay-and-sum beamforming methods used with such arrays suffer from inaccurate estimations of absolute source levels and in some cases also from low resolution. Deconvolution approaches such as DAMAS have better performance, but require high computational effort. A fast beamforming method is proposed that can be used in conjunction with a phased microphone array in applications with focus on the correct quantitative estimation of acoustic source spectra. This method bases on an eigenvalue decomposition of the cross spectral matrix of microphone signals and uses the eigenvalues from the signal subspace to estimate absolute source levels. The theoretical basis of the method is discussed together with an assessment of the quality of the estimation. Experimental tests using a loudspeaker setup and an airfoil trailing edge noise setup in an aeroacoustic wind tunnel show that the proposed method is robust and leads to reliable quantitative results.