基于波束形成的声源识别方法及其风洞内应用

Sound Source Identification Based on Conventional Beamforming and Its Application in the Wind Tunnel

摘要:

基于波束形成的声源识别方法是风洞内研究汽车气动噪声问题重要的实验手段。为了更好地研究及应用这一测试技术，首先通过仿真计算，考察麦克风阵型、声源频率、测试距离、声波入射角等因素对阵列性能的影响，并通过系统设计完成了实验验证。结果表明：随着测试距离和入射角增大，系统空间分辨率变差；随着声源频率增大，系统空间分辨率更好，但动态范围变差；随着测试距离偏差的增大，识别的声源强度不断降低；不同声源频率、测试距离与阵列阵型对测试距离偏差引起的声源幅值误差的敏感度不同，声源频率越高，测试距离越小则敏感度越高；多声源的相互影响使阵列的有效动态范围降低，同时声源强度增大。然后，基于DrivAer整车油泥模型，在整车气动声学风洞内应用波束形成技术进行车外声源识别，分析车外气动噪声源的分布。最后，应用声源相减方法，分析了不同造型后视镜声源识别结果的差异。

Abstract:

Sound source identification based on beamforming is an important experimental method for the study of vehicle aerodynamic noise in wind tunnel. In order to better understand and apply this test technique， a series of simulation were conducted. The influence of factors such as array shape， source frequency， distance between array and testing object， and incidence angle of sound wave on the performance of the array was studied and validated with designed corresponding experiments. The results show that the increases of test distance and incidence angle will worsen the spatial resolution. As source frequency increases， the system spatial resolution becomes better， but the dynamic range worsens. As deviation of the distance between array and testing object increases， the identified source intensity decreases. Furthermore，the sensitivity of different source frequencies， testing distances， and array shapes to sound source amplitude errors caused by testing distance deviations varies. As the source frequency goes higher， testing distance becomes shorter， the sensitivity becomes higher accordingly. The mutual influence of multiple sound sources reduces the effective dynamic range of the array， while increasing the intensity of the sound sources. Afterwards， based on the full-scale DrivAer clay model， the sound sources distribution of the vehicle exterior was identified with beamforming array. Finally，the difference in sound source recognition results of rearview mirrors with different shapes was analyzed using the sound source subtraction method， which demonstrates that this technique can provide a powerful support for the development of vehicle wind noise.