基于传声器阵列的汽车鸣笛声定位算法及实现

针对违章鸣笛车辆的锁定问题,本文提出了一种基于传声器阵列空间搜索法的三维空间声源定位算法。设计计算机仿真实验,分析并确定了阵列的传声器个数。通过广义互相关方法估计传声器阵列中任意两个传声器的相对时延,然后结合估计得到的时延,采用基于5传声器阵列的空间搜索法计算得到声源空间位置。为验证算法的可行性,实际构建三维空间的点声源定位系统,进行实车定位实验。实验结果表明提出的算法可以有效地定位汽车鸣笛声。     

音频无人机定位的时延估计模拟分析

针对无人机非平稳音频信号时差定位中,广义互相关时延估计算法抗噪性差和时延估计值精度低等问题,文章采用了一种基于广义二次相关时延估计的改进算法。算法对叠加了实际噪声（如风声、雨声、汽车鸣笛声等）的无人机音频信号进行频谱细化的广义二次相关,有效抑制了噪声干扰,融合相关峰精确插值算法,提高了互相关函数的分辨率,使得时延峰值更加明显。仿真实验结果表明,改进的广义二次相关方法在不同信噪比时,比广义互相关和广义二次相关算法的时延估计精度更高,稳定性更好。改进的广义二次相关算法对无人机定位中的时延估计具有更好的性能优势,具有较强的实际应用性。     

基于时延估计改进的主动隔声耳罩语音增强算法

为实现高噪声场所中的听力保护和不同角度入射声源的语音增强，文章采用基于谱减法改进的广义互相关（Generalized Cross Correlation, GCC）时延估计方法对麦克风阵列接收信号进行同步处理，将同步后的阵列信号与基于广义旁瓣抵消器（Generalized Sidelobe Canceller, GSC）的主动隔声耳罩语音增强算法结合，实现噪声抑制和语音增强。首先以GCC算法为基础，引入多窗谱估计的谱减法作为时延估计信号的预处理环节，提高传声器接收信号的信噪比对算法进行改进。随后对加入谱减法的GCC时延估计算法，比较分析了取不同加权函数的仿真结果。最后由时延估计结果对麦克风阵列信号进行时延补偿，对改进后的主动隔声耳罩语音增强算法的噪声抑制和语音增强效果进行了分析。与改进前算法相比，在不同噪声环境下输出的语音信号质量有明显提升。     

Pattern时延差编码水声定位的时延估计研究

基于Pattern时延编码体制,针对水下超短基线定位系统设计了呼叫信号的正、负调频时延编码结构,并提出利用二次互相关技术进行时延估计。该方法在一定程度上能够提高呼叫方的可检测信噪比,同时在复杂环境下对信道中包括直达波在内的多径信号能量进行累加,能够准确地完成峰值检测,进而达到提高时延估计精度的目的。海试实验结果表明二次相关法在不同信噪比条件下能有效地抑制界面反射产生的多径干扰。     

室内噪声环境下语音信号时延估计

针对室内复杂的声学环境,结合语音信号的特点,提出了一种利用语音激励信息的时延估计算法。采用线性预测得到不同位置麦克风接收的含噪语音信号的残余误差(LPCRE),残余误差中包含激励信息,再利用残余误差的互功率谱相位(CSP),进行时延估计,性能有较大改善,且适合背景噪声为脉冲噪声或白噪声或两者混合的环境下应用。     

多途信道对弹道跟踪系统时延估计的影响分析

0引言基于时延估计的定位技术计算量小,定位精度高,而且该算法的实现对硬件要求不高,易于实现实时定位系统的开发,这些优点使其广泛应用于水下弹道跟踪系统中,时延估计的精度将直接影响到弹道跟踪的准确性。然而,复杂的水下环境使得水声信道具有传播衰减、多途、散射和频散等物理效应[1],水声信道的这些特性对高精度的时延估计构成了巨大的挑     

多通道近场广义互相关算法研究

广义互相关时延估计方法(GCC)由于计算简单和易于实现,在被动声源定位系统中得到了广泛应用,在此基础上,提出了多通道近场广义互相关算法。该算法建立时延估计矩阵,得出方位估计函数,用最小方差估计出最优的方位估计量。仿真实验表明,这种方法使估计精度得到提高。     

基于单比特量化的单频信号时延估计算法

提出了一种基于单比特量化的单频信号自适应时延估计算法,采用该算法在保持信号频谱特性不发生变化的基础上,极大地去除信号冗余量,对简化后的信号框架采用包络检测,估计时延.对接收信号进行单比特量化能够最大限度地降低数据量,同时量化结果又进一步简化了单频信号滤波器的输入响应函数结构,自适应算法则保证了时延估计的高分辨率.通过仿真实验表明,该算法在不造成频偏和频率衰减的同时,又很好保证了时延估计精度.最后利用湖上实验验证了算法的有效性和可靠性.     

噪声相关条件下多传声器时延多尺度估计

提出一种基于小波分解的多尺度时延估计方法,解决非稳态、有色噪声且通道间噪声存在相关性的条件下提高时延估计精度问题,并把这一理论应用于声阵列对声源的定位.利用小波良好的时间-频率分解性能将非稳态信号分解为多个稳态信号,同一尺度内进行互相关获得多尺度时延值,并证明了小波分解不会改变信号之间的固有时延值.针对不同尺度内时延值不一致、无法识别出正确时延值这一问题,提出用三传声器之间时延匹配关系消除噪声相关性,来识别正确时延,从而实现非稳态、噪声相关条件下的时延估计.最后通过正四方形声阵列试验检验了该方法,试验证明:该方法不但提高了识别率、细化和突出了相关函数的波峰,还提高了时延估计的稳定性和精度.     

一种水下声信号未知频率的时延差估计方法

针对目标辐射线谱信号频率未知时的时延差(TDD)估计问题,介绍一种基于时延差方差加权的时延差估计方法。首先对接收信号进行快速傅里叶变换(FFT)分析,然后采用引导信号对每一个频率单元进行频域互相关得到每一个频率单元时延差估计初值,最后利用噪声对应频率单元互相关谱最大值随机,目标对应频率单元互相关谱最大值基本一致的特点,统计各频率单元的时延差估计结果得到最终时延差估计值。理论分析和实验结果表明:该方法具有较好的有效性,对信噪比的宽容性远好于频域互相关法,该方法为弱线谱时延差估计提供了一个参考思路。     

吸声系数的传递函数法仿真计算

为得到吸声材料的吸声系数,提出了一种基于Virtual.Lab Acoustics平台的阻抗管模拟方法,建立阻抗管和试件的声学模型,提取声学计算得到的传声器测点处的复声压,然后基于传声器间距、传声器与试件距离,得到传声器测点位置的传递函数及吸声系数。将该方法的计算结果与文献值进行对比,验证了文中传递函数及吸声系数计算方法的正确性。通过仿真分析了阻抗管直径、传声器测点位置对计算吸声系数的影响,说明传声器测点位置可极大影响计算的吸声系数。对阻抗管内试件倾斜角度、试件厚度和试件后侧空气柱等对吸声系数的影响进行了定量计算。     

基于PolyMAX的声固耦合模态试验研究

白车身的结构模态频率和模态振型反映了汽车车身结构的固有特性,对车内噪声有重要影响。车内空腔跟车身结构一样,同样拥有模态频率和模态振型。采用LMS数据采集系统对某国产SUV进行车内空腔声学模态试验。首先基于传声器阵列的方法获取响应点的信号,然后利用PolyMAX方法提取声学模态频率及振型。将声学模态频率与白车身结构模态频率进行对比分析,结果表明：车内空腔的第一、二阶声学模态分别跟白车身的第四、十阶结构模态有很强的耦合。最后通过实车测试验证了声固耦合共振时低频轰鸣的存在。可以在关键部件增加板厚、顶盖和地板附加阻尼层、顶盖加加强筋等方式改变车身结构的局部模态来破坏车身结构模态和声腔模态的强耦合状态,降低车内的低频轰鸣声     

Searching for momentum enhancement in hypervelocity impacts

In conjunction with the Los Alamos National Laboratory hypervelocity microparticle impact (HMI) team effort to produce higher impact velocities and to understand the physics of crater formation and momentum transfer, we have implemented a low noise microphone as a momentum detector on both the 6 MV Van de Graaff and 85 KV “test stand” particle accelerators. Calculations are presented showing that the impulse response of a circular membrane. When used as a momentum impulse detector, the microphone theoretically may detect impulses as small as 8.8 × 10⁻¹⁵ N s. Sensitivity of the microphone in this application is limited by the noise threshold of the electronic amplifiers and the ambient microphonic vibration of the system. Calculations lead us to anticipate detection of particles over the full range of the Van de Graaff acceleration capability and up to 7 km/s on the test stand. We present momentum enhancement data in the velocity range between 10 km/s and 20 km/s. Preliminary work is presented on momentum impulse calibration of the microphone using laser-pulse photon momentum as an impulse source.     

Quantitative signal analysis in pulsed resonant photoacoustics

The pulsed excitation of acoustic resonances was studied by means of a high-Q photoacoustic resonator with different types of microphone. The signal strength of the first radial mode was calculated by the basic theory as well as by a modeling program, which takes into account the acoustic impedances of the resonator, the acoustic filter system, and the influence of the microphone coupling on the photoacoustic cavity. When the calculated signal strength is used, the high-Q system can be calibrated for trace-gas analysis without a certified gas mixture. The theoretical results were compared with measurements and show good agreement for different microphone configurations. From the measured pressure signal (in pascals per joule), the absorption coefficient of ethylene was calculated; it agreed within 10% with literature values. In addition, a Helmholtz configuration with a highly sensitive 1-in. (2.54-cm) microphone was realized. Although the Q factor was reduced, the sensitivity could be increased by the Helmholtz resonator in the case of pulsed experiments. A maximum sensitivity of the coupled system of 341 mV/Pa was achieved.     

基于传声器阵列的电泵噪声源识别研究

某型电泵的噪声较大,为了准确辨识出该型电泵的噪声源,得到其声场辐射特性,采用基于时域信号的"延迟与求和"的"波束成型"的阵列信号处理算法,建立了由29个传声器组成的" "字型平面传声器阵列,首先在实验室用音箱进行了声源辨识的验证和分析试验,然后分析某型电泵的辐射噪声,得到了电泵辐射噪声源的分布特性和频谱特性,分析了形成噪声的主要原因.     

应用于传声器阵列定位校准的空间点声源声场拟合方法

介绍了一种用于传声器阵列声源定位精度校准的空间点声源声场模拟方法,并基于该方法设计了一套空间点声源模拟系统,完成了一个传声器阵列的定位位置精度校准。文章采用多通道点声源空间声场合成算法模拟了一个位于自由场空间的点声源,根据传声器阵列中每一个传声器的空间位置坐标,计算出传感器所处位置声场的动态声信号。通过耦合腔标准声源将对应的多通道电压信号输入被校准阵列系统,完成点声源的模拟。然后,该阵列运用波束形成算法进行声源定位,得出点声源的位置,并与模拟点声源的位置进行比对,实现对阵列定位准确性的校准。     

Strategies for Reducing Stress in Ultrasonic Cutting Systems

Abstract:  Ultrasonic cutting is an established technology in the cutting of food products. As the high-cost components of an ultrasonic cutting system are the ultrasonic generator and transducer, some designs have concentrated on running several cutting blades from a single power source. This strategy is undoubtedly economic, but problems remain with reliability. Blade failure during cutting is a common problem in ultrasonic systems and therefore strategies are proposed to reduce stress levels at the failure locations. Many alternative blade profiles are investigated using finite element (FE) models, which are validated by experimental modal analysis (EMA) using a 3D laser Doppler vibrometer (LDV). Combining three strategies can reduce the stress level at the failure location on cutting blades. The first investigates the blade profile at the failure location with the aim of incorporating sufficient gain at reduced stress levels. The second concentrates on detuning the block horn and blades in order to move the longitudinal node away from the highest stressed section of the blade. Finally, a redesign of the block horn is studied in order to eliminate the effects of blade flexural vibrations in the longitudinal cutting mode.     

一种可作为声源定位与声场空间特性分析的测量系统

提出了一种由重复发射的声源配合缓慢转动的传声器组成的、等效于圆形传声器阵列的室内声学测量系统。阐述了这种测量系统的特点及其理论基础,介绍了它在室内声场中若干涉及声源定位及声场空间特性测量等方面的应用,并结合实例作了讨论。系统具有较强的适应性及实用性,在室内声学测量中具有广泛的应用前景。     

Effect of Power Line Interference on Microphone Calibration Measurements Made at or Near Harmonics of the Power Line Frequency

The electrical measurements required during the primary calibrations of laboratory standard microphones by the reciprocity method can be influenced by power line interference. Because of this influence, the protocols of international inter-laboratory key comparisons of microphone calibrations usually have not included measurements at power line frequencies. Such interference has been observed in microphone output voltage measurements made with a microphone pressure reciprocity calibration system under development at NIST. This system was configured for a particular type of standard microphone in such a way that measurements of relatively small signal levels, which are more susceptible to the effect of power line interference, were required. This effect was investigated by acquiring microphone output voltage measurement data with the power line frequency adjusted to move the frequency of the interference relative to the center frequency of the measurement system passband. These data showed that the effect of power line interference for this system configuration can be more than one percent at test frequencies harmonically related to the power line frequency. These data also showed that adjusting the power line frequency to separate the interference and test frequencies by as little as 1.0 Hz can reduce the effect of the interference by at least an order of magnitude. Adjustment of the power line frequency could enable accurate measurements at test frequencies that otherwise might be avoided.     

一种声场空间特性测量系统的性能研究

室内声学测量中,在声源重复发射声信号的同时,将单个无指向性传声器缓慢而匀速地转动,其测量结果等效于以转盘转轴为中心的圆形传声器阵列,相应的测量系统简称为RRS。通过消声室测量,着重分析此测量系统在方向脉冲响应测量中的性能及主要影响因素：结果表明,在满足传声器旋转半径大于1／4信号波长的条件下,系统可具有令人满意的方向分辨能力。