基于矢量阵的运动声源柱面聚焦定位方法试验研究

本文基于被动合成孔径原理, 在建立运动声源矢量阵近场柱面聚焦测量模型的基础上, 分别研究了适用于单频线谱信号和宽带连续谱信号的矢量阵柱面聚焦定位方法, 通过数值仿真计算了该方法在多种误差条件下的定位精度, 并进一步通过舱段模型试验对该方法的工程实用性和正确性进行了详细的分析和论证. 舱段模型试验结果表明, 柱面聚焦定位结果与壳体振动能量分布规律符合较好, 该方法不仅能真实反映声源位置信息, 而且能反映不同频带内声源能量分布的相对大小, 具有良好的定位效果.     

基于波束形成的多类型多声源定位研究

为实现空压机多噪声源的准确定位,仿真对比了多种近场球面波多声源定位算法。基于时域波束形成,研究了相同声源平面、不同声源频率、不同声源纵向距离、不同声源强度下多声源定位以及声源频率、声源纵向距离和声源强度多因素联合的多声源定位仿真方法,模拟了更接近实际的噪声源类型。基于频域波束形成,仿真研究了1400 Hz,2400 Hz,3400 Hz,4400 Hz的多声源。分别利用互功率谱波束形成和除自谱的互功率谱波束形成,仿真研究了相干声源和不相干声源。开发了阵列声成像测试平台,运用频域波束形成和功率谱波束形成对空压机进行了定位试验研究。结果表明,1400 Hz下空压机的主要噪声源是气缸盖、空气滤清器和曲轴附近的机体,这可为空压机减振降噪改进设计提供依据。     

十七无十字阵水下合作目标远程定位方法

结合水声对抗的实际需求,针对海洋环境下目标定位困难且定位距离有限的问题,利用多水听器十字阵接收信号,并对信号进行加权来提高接收信号信噪比,以达到定位水下合作目标的目的。具体通过波束形成和信号相关,依据目标方位与波束形成方位一致时能量最大的原则,实现对远距离水下合作目标方位和距离的估计,并通过仿真和试验验证了所提定位方法的有效性。     

圆柱阵子阵分级处理的稳健超指向性波束形成方法

提出了圆柱阵子阵分级处理的稳健超指向性波束形成方法。首先建立了圆柱阵分两级子阵进行波束形成的信号模型,接着利用空间均匀噪声场中噪声互谱矩阵的循环特性得到基于特征波束分解与综合模型的圆柱阵超指向性的最优解,然后仿真研究了其误差敏感度函数、阵增益和波束图等性能指标,并与圆柱阵的传统全局处理方法进行了对比。提出的分两级子阵处理的超指向性方法与传统全局处理方法相比不仅降低了数据存储量和波束形成计算量,而且进一步提升了稳健性,并且在低频段的阵增益远远高于常规波束形成的值,对水下声呐阵列的设计具有一定的参考价值。     

球形传声器阵列下基于主导声源检测MUSIC群时延算法的多声源定位*

针对混响环境中,多径效应、散射、衍射等原因导致声源定位失败或分辨能力不足的现象,提出一种基于主导声源检测MUSIC群时延的邻近多声源定位方法。该方法采用球形传声器阵列,相比平面阵列可以捕获3D声场信息,利用球谐域下信号的频率分量与角度分量解耦的优势,从而可直接利用频率平滑技术处理宽带语声信号而不需要构造聚焦矩阵,并在球谐域下通过设置阈值对一组时频段进行主导声源检测,从而选择出包含直达声的一组时频块来构造MUSIC群时延空间谱。上述举措在提升波达方向估计在高混响环境下定位鲁棒性的同时,也提高了多个邻近声源的分辨能力。仿真实验结果表明,所提出的主导声源检测MUSIC群时延算法,在高混响和低信噪比条件下,仍具有更好的定位精度与更优的邻近多声源分辨效果。     

基于阵形畸变补偿的多线阵左右舷分辨研究

多线阵可解决单线阵在目标探测中的左右舷分辨模糊问题,但阵形畸变不利于准确实现对目标的左右舷分辨。本文针对多线阵在拖曳过程中出现的典型阵形畸变,定量地分析了其对目标左右舷分辨的影响,并提出了一种基于阵形畸变实时测量补偿来进行目标左右舷分辨的方法,海上实验数据分析结果表明,该方法可实现阵形畸变下的目标左右舷分辨,对于提高拖曳多线阵声纳性能具有参考价值。     

有限采样下低频声源聚焦波束形成的定位研究

依赖于基阵协方差矩阵的算法,要求有足够样本容量来实现稳健的性能。针对常规聚焦波束形成在低频段分辨能力不足以及有限的数据长度,本文研究了一种基于最小方差无失真响应(MVDR)的重叠分子阵算法。仿真实验表明,重叠分子阵MVDR算法在低快拍数、低信噪比情况下,稳健性优于全阵列MVDR算法。在海试数据中,利用一条孔径为109m的光纤水听器阵列,验证了该算法可以在低快拍数下对50m内低频段100～200Hz的目标实现分辨率为0.5m的近场定位。     

利用水平阵模态域波束形成判别声源深度*

针对近水面声源和水下声源的深度判别问题,根据近水面声源难以激发低阶模态的物理现象,研究利用声源波数谱结构和波数位置的不同来分辨近水面声源和水下声源。通过采用MVDR的谱估计方法进行模态域波束形成,补偿水平阵各阵元之间各号简正波的相位差,获得主瓣窄、旁瓣低的声源信号波数谱。波数谱的波数位置与频率呈近似线性关系,水中声速剖面、海底参数、海深都会影响波数谱的具体结构和位置。此外,声源信号的到达角估计误差同样也会影响波数谱主瓣的位置估计。数值仿真结果表明,在浅海负跃层声速剖面条件下,可利用水平阵模态域波束形成判别声源深度,区分近水面声源和水下声源。     

圆阵快速数字多波束形成

在时域上实现多波束形成,其运算量与系统时空采样点数的平方成正比.木文提出以“交迭-贮存”的方法计算输入信号与时延横向滤波器脉冲响应的卷积,再根据Farrier等人的工作,用FFT算法完全实现时域上的圆阵多波束形成.因此在节省运算量的同时,保证了结果的准确.文中给出了计算机模拟计算的结果,同时还提出了一种非整数倍采样周期时延的方法,改善了指向特性.     

基于任意阵的最小方差无失真响应聚焦波束形成的被动定位方法

将最小方差无失真响应(MVDR)聚焦波束形成近场被动定位方法应用到平面任意结构阵列上,可以很好地解决线列阵由于阵长较长,在实际使用中存在布放困难、部分阵元失效影响定位性能等问题。本文首先分析了平面任意结构阵列的近场接收信号模型,推导出阵列流型,然后将MVDR聚焦波束形成技术应用到该模型中,得到了基于任意阵的MVDR聚焦波束形成近场被动定位方法。仿真验证了该方法的可行性和有效性。     

一种利用分布式传声器阵列的声源三维定位方法

为了提高噪声和混响条件下分布式传声器阵列进行声源定位的性能,提出一种利用空间稀疏性和压缩感知原理的声源三维定位方法。该方法首先通过两次离散余弦变换方式提取出声音信号特征,并用该特征来构建稀疏定位模型,以便能够综合利用语音信号的短时和长时特性,同时降低模型维数;然后利用在线字典学习技术动态调整字典,克服稀疏模型与实际信号之间的失配问题,增强稀疏定位模型的鲁棒性;进而提出一种改进的平滑l₀范数稀疏重构算法来进行声源位置解算,以提高低信噪比条件下的重构精度。仿真结果表明该方法不仅可以实现多目标定位,而且具有较强的抗噪声和抗混响能力.     

Three-dimensional sound source localization using distributed microphone arrays

To improve the performance of sound source localization based on distributed microphone arrays in noisy and reverberant environments,a sound source localization method was proposed.This method exploited the inherent spatial sparsity to convert the localization problem into a sparse recovery problem based on the compressive sensing(CS) theory.In this method two-step discrete cosine transform(DCT)-based feature extraction was utilized to cover both short-time and long-time properties of the signal and reduce the dimensions of the sparse model.Moreover,an online dictionary learning(DL) method was used to dynamically adjust the dictionary for matching the changes of audio signals,and then the sparse solution could better represent location estimations.In addition,we proposed an improved approximate l_0norm minimization algorithm to enhance reconstruction performance for sparse signals in low signal-noise ratio(SNR).The effectiveness of the proposed scheme is demonstrated by simulation results where the locations of multiple sources can be obtained in the noisy and reverberant conditions.     

同心圆环阵列模态波束形成器设计

针对贝塞尔函数零点造成的模态波束形成器性能损失的问题,提出了基于均匀同心圆环阵列的模态波束形成器设计方法。从阵元域出发,建立了阵元域和圆谐波域之间的转换关系,推导了圆谐波域阵列信号处理表达式,其中重点推导了延时求和、最小方差无失真响应和多约束波束形成器的设计方法。理论证明了在平面各向同性噪声场中的基于同心环阵的圆谐波域最小方差无失真响应波束形成器等价于相位模式波束形成器。在该结论的基础上,综合考虑相位模式波束形成器在一定频率范围内的稳健性和指向性,提出了一种同心圆环阵列的内环半径优化方法。使用16元双环阵列对本文提出的波束形成算法进行仿真和实验,结果表明:同心圆环阵列能够较好地解决贝塞尔函数零点问题;其中多约束波束形成器能够在多个关联的性能指标之间取得合理的折衷,实现优于相位模式和延时求和波束形成器的声源定位性能。对内环半径优化方法进行仿真和分析结果表明,该方法得到了相较于传统方法更优的阵列结构。     

Non-line-of-sight sound source localization using matched-field processing

Acoustic diffraction allows sound to travel around opaque objects and therefore may allow beyond-line-of-sight sensing of remote sound sources. This paper reports simulated and experimental results for localizing sound sources based on fully shadowed microphone array measurements. The generic geometry includes a point source, a solid 90° wedge, and a receiving array that lies entirely in the shadow defined by the source location and the wedge. Source localization performance is assessed via matched-field (MF) ambiguity surfaces as a function of receiving array configuration, and received signal-to-noise ratio for the Bartlett and minimum variance distortionless (MVD) MF processors. Here, the sound propagation model is developed from a Green's function integral treatment. A simple 16 element line array of microphones is tested in three mutually orthogonal orientations. The experiments were conducted using an approximate 50-to-1-scaled tabletop model of a blind city-street intersection and produced ambiguity surfaces from source frequencies between 17.5 and 19 kHz that were incoherently summed. The experimental results suggest that a sound source may be localized by the MVD processor when using fully shadowed arrays that have significant aperture parallel to the edge of the wedge. However, this performance is reduced significantly for signal-to-noise ratios below 40 dB.     

Robust tracking of moving sound source using multiple model Kalman filter

In this paper a novel method for tracking an active speaker in a noisy and reverberant environment by means of a spatially distributed microphone array is presented. Firstly, a sound source localization algorithm based on time delays of arrival (TDOA) in microphone pairs provides observed position estimates. Then these remarkably noisy estimates are filtered by a multiple model Kalman filter (MMKF) in order to obtain a smoothed trajectory of the speaker’s movement. Compared with the traditional Kalman filter (KF), simulated results prove the MMKF is more robust and effective in noisy environments.     

Robust tracking of moving sound source using scaled unscented particle filter

Tracking an active sound source involves the modeling of non-linear non-Gaussian systems. To address this problem, this paper proposed scaled unscented particle filter (SUPF) algorithm for tracking moving sound source. The particle filter part of the SUPF provides the general probabilistic framework to handle non-linear non-Gaussian systems, and the scaled unscented Kalman filter (SUKF) part of the SUPF generates better proposal distributions by taking into account the most recent observation. Meanwhile, models used in SUPF algorithm were also explored for the sound source motion, observation and the likelihood of the sound source location in the light of the Langevin process. Compared with the conventional PF approach, the simulated results demonstrated that the SUPF algorithm had superior tracking performance.     

Material sound source localization through headphones

In the present paper a study of sound localization is carried out, considering two different sounds emitted from different hit materials (wood and bongo) as well as a Delta sound. The motivation of this research is to study how humans localize sounds coming from different materials, with the purpose of a future implementation of the acoustic sounds with better localization features in navigation aid systems or training audio-games suited for blind people. Wood and bongo sounds are recorded after hitting two objects made of these materials. Afterwards, they are analysed and processed. On the other hand, the Delta sound (click) is generated by using the Adobe Audition software, considering a frequency of 44.1 kHz. All sounds are analysed and convolved with previously measured non-individual Head-Related Transfer Functions both for an anechoic environment and for an environment with reverberation. The First Choice method is used in this experiment. Subjects are asked to localize the source position of the sound listened through the headphones, by using a graphic user interface. The analyses of the recorded data reveal that no significant differences are obtained either when considering the nature of the sounds (wood, bongo, Delta) or their environmental context (with or without reverberation). The localization accuracies for the anechoic sounds are: wood 90.19%, bongo 92.96% and Delta sound 89.59%, whereas for the sounds with reverberation the results are: wood 90.59%, bongo 92.63% and Delta sound 90.91%. According to these data, we can conclude that even when considering the reverberation effect, the localization accuracy does not significantly increase.     

基于稀疏表示和特征加权的离格双耳声源定位*

基于头相关传递函数数据库的传统双耳声源定位方法的定位角度往往被限定在头相关传递函数数据库的离散测量点上。当头相关传递函数数据库的测量方位角间隔较大时,这类算法的性能会显著下降,这就是典型的离格问题。该文提出了基于加权宽带稀疏贝叶斯学习的离格双耳声源定位算法。首先该算法建立离格双耳信号的稀疏表示模型,然后利用双耳相干与扩散能量比特征对各个频点进行加权以降低噪声和混响的影响,最后通过加权宽带稀疏贝叶斯学习方法估计离格声源的方位角。实验结果表明,该算法在各种复杂的声学环境下都有着较高的定位精度和鲁棒性,特别是提高了离格条件下的声源定位性能。     

声源作圆周运动时的多普勒效应

利用自制装置演示了声源作圆周运动的多普勒效应，并应用图谱分析软件得到频率分布图。