基于耦合声场建模的膨胀腔消声器声学性能分析及试验

针对圆柱形膨胀腔消声器三维建模及声学性能分析问题, 提出一种基于切比雪夫变分原理的耦合声场建模方法, 建立三维圆柱形膨胀腔消声器理论模型并搭建试验台架, 传递损失试验结果验证了理论模型的准确性. 将膨胀腔消声器内部声场分解为多个子声场, 基于子声场间压力与质点振速连续性条件, 推导声场耦合变分公式, 构建子声场拉格朗日泛函. 将子声场声压函数展开为切比雪夫-傅里叶级数形式, 通过瑞利-里兹法求解膨胀腔消声器频率、声压响应及传递损失. 计算并对比分析扩张比、扩张腔长度、进出口管偏置对膨胀腔消声器消声性能的影响. 结果表明: 扩张比增大会有效提高消声器在低频段的消声性能, 进出口管的偏置对消声器消声性能影响很小.     

阻性管道消声器消声量方程的一个新算法

将阻性管道消声器消声量特征方程（组）转化为常微分方程（组），由此给出消声量计算的一个新方法。该方法毋需迭代因而不存在计算稳定性问题，计算量小，初值的迭取也很简单，适合于扁矩形、任意矩形及圆形等各种截面管道消声器，尤其对多变量、考虑平均流影响的情形更为有效。     

一种微穿孔板消声器

本文根据马大猷教授提出的“微穿孔板吸声体精确理论”，在对微穿孔板消声器结构参数与吸声特性之间的相互影响进行了分析的基础上，设计了微穿孔板消声器并将其应用在新舟60飞机APU(辅助动力装置)降噪的实际工程中，并通过编写的噪音数据采集分析程序，对降噪效果进行了测试和分析，验证出微穿孔板消声器对消除飞机的APU的排气噪声具有良好的消声效果，平均降噪7．16dB。     

GA optimization on multi-segments muffler under space constraints

Whilst the space volume of muffler in noise control system is often constrained for maintenance in practical engineering work, the maximization on muffler’s performance becomes important and essential. In this paper, a novel approach genetic algorithms (GAs) based on the principles of natural biological evolution will be used to tackle this optimization of muffler design [M. Mitchell, An Introduction to Genetic Algorithms, The MIT Press, Cambridge, MA, 1996]. Here, the shape optimization of multi-segments muffler coupled with the GA searching technique is presented. The techniques of binary genetic algorithms (BGA) together with the commercial MATLAB package [G. Lindfield, J. Penny, Numerical Method Using Matlab, second ed., Prentice Hall, Englewood Cliffs, NJ, 2000] are applied in GA searching. In addition, a numerical case of pure tone elimination with 2-5 segments on muffler is introduced and fully discussed. To achieve the best optimization in GA, several GA parameters are on trial in various values. Results show that the GA operators, including crossover mutation and elitism, are essential in accuracy. Consequently, results verify that the optimal sound transmission loss at the designed frequency of 500 Hz is exactly maximized. The GA optimization on multi-segments muffler proposed in this study surely provides a quick and correct approach.     

Sound attenuation of a periodic array of micro-perforated tube mufflers

The wave propagation in a periodic array of micro-perforated tube mufflers is investigated theoretically, numerically and experimentally. Because of the high acoustic resistance and low mass reactance due to the sub-millimeter perforation, the micro-perforated muffler can provide considerable sound attenuation of duct noise. Multiple mufflers are often used to enhance attenuation performance. When mufflers are distributed periodically in a duct, the periodic structure produces special dispersion characteristics in the overall sound transmission loss. The Bloch wave theory and the transfer matrix method are used to study the wave propagation in periodic micro-perforated tube mufflers and the dispersion characteristics of periodic micro-perforated mufflers are examined. The results predicted by the theory are compared with finite element method simulation and experimental results. The results indicate that the periodic structure can influence the performance of micro-perforated mufflers. With different periodic distances, the combination of the periodic structure and the micro-perforated tube muffler can contribute to the control of lower frequency noise with a broader frequency range or improvement of the peak transmission loss around the resonant frequency.     

The research on semi-active muffler device of controlling the exhaust pipe’s low-frequency noise

In order to control low frequency noise in exhaust pipe, this paper puts forward a new concept of H-Q tube based semi-active muffler device. The semi-active muffler device and bench testing system have been designed and operated. Finite element simulation study on semi-active muffler and experimental study on semi-active muffler and passive muffler have been carried on. The effect of simulation and experiment are consistent. The semi-active muffler device acts well in low frequency band, especially between 50 Hz and 150 Hz. The average level of noise reduction is around 35 dB, which is much better than passive muffler. Between 150 Hz and 350 Hz, semi-active muffler has a better performance than passive muffler; above 350 Hz, it has worse performance compared with the passive muffler.     

磁流变液效应对远端流场的扰动

研究了磁流变液效应对远离磁极的流道中流动的扰动问题.通过双腔型结构器件的动态力学响应以及混合流模式的流动分析,揭示了当磁流变液离开磁极后,由于剩磁的影响,磁极后方的管道中的流动将受影响,主要表现在流过管道的介质的体积比浓度将发生显著变化.     

Shock wave interaction with area changes in ducts

Whitham's approximation for handling shock wave propagation in area changes (reductions) in a duct was checked in comparison with a numerical solution. Also the Whitham approximation for shock wave propagation from a constant cross-sectional duct to a duct of a smaller cross-sectional area was studied and compared with a numerical solution. It was found that for modest incident shock Mach numbers and modest area reductions the Whitham approximation provided a fair solution for the shock Mach number and for the post-shock pressure. For higher shock Mach numbers and/or area reductions, large discrepancies exit between the approximate and exact solutions. A wider range of applicability of the Whitham approximation is found for the monotonical area reduction case; it is quite narrow for the passage of a shock wave from a wider to a narrower duct case. In addition, the effect of the extent of the area change region on the time required for reaching a quasi-steady flow was studied. It was shown that the longer the area change segment is, the longer it takes to reach a quasi-steady flow.This article was processed using Springer-Verlag T_EX Shock Waves macro package 1.0 and the AMS fonts, developed by the American Mathematical Society.     

电子器件散热风扇气动噪声管道声学模态截止控制技术

深度学习输入特征的选择直接影响其分类性能,为了进一步提高基于深度学习的鸟类物种识别模型的分类性能,该文提出一种多特征融合识别方法。该方法首先通过短时傅里叶变换、梅尔倒谱变换和线性调频小波变换分别计算得到鸣声信号的3种语图样本集,然后分别利用3种语图样本集训练3个基于VGG16迁移的单一特征模型,将3个模型的输出进行自适应加权求和实现融合,并修正了加权交叉熵函数以克服样本不平衡的问题,最后对语图进行分类实现鸟类物种的识别。以ICML4B鸣声库的35种鸟类为研究对象,对比了4种模型的平均识别准确率(MAP),结果表明特征融合模型较单一特征模型的MAP最大提高了0.307;选择输入语图的持续时间分别为100 ms、300 ms以及500 ms,对比不同持续时间下4种模型的测试MAP值,结果表明持续时间为300 ms时4种模型的MAP值均为最高;对比了不同信噪比下4种模型的识别效果,多特征融合模型的识别准确率随着信噪比的下降降低最少。说明在选择合适的语图持续时间后,该文提出的特征融合模型能得到更高的识别准确率,具有一定的抗噪能力,且训练参数少,更适合于少样本鸟类的识别。     

基于低频吸声超构材料的复合消声器研究*

针对管道内低频噪声难以抑制的问题，本文基于亥姆霍兹共振腔（HR）阵列吸声板和穿孔管消声器组合，设计了一种复合式宽带消声器。首先利用有限元法仿真分析传统穿孔管消声器，发现中低频消声能力较差，通过嵌入HR阵列吸声板吸收中低频噪声。采用仿真与实验的方式研究吸声板的声学性能：在400-1000 Hz频段内的平均吸声系数达到了0.88。然后对复合式消声器进行数值模拟及3D打印阻抗管实验测试对比：复合式消声器在400-1718Hz频率范围内的平均传递损失为18.15 dB ，最终实现了管道内全频带噪声有效控制。