一种改进的声反卷积相关声源定位方法

基于传声器阵列的声源定位算法的性能优劣决定了气动噪声试验检测中气动噪声源定位成像的准确性。为了提高相关声源定位方法的稳健性和准确性，提出一种改进的声反卷积相关声源定位方法——D-MACS（Developed-Mapping of Acoustic Correlated Sources）方法。该方法的特点是：在传统MACS方法的基础上，将声反卷积模型进行改进，使模型体现出声场空间扫描点声压互谱矩阵与互谱波束形成输出矩阵之间的数学关系，其最终获得的声源定位结果，能够提高MACS方法声源定位辨识的稳健性和准确性。通过仿真信号研究和声学风洞环境中的声源定位试验研究，比较了DAS（Delay And Sum beamformming）、DAMAS-C（Deconvolution Approach for the Mapping of Acoustic Correlated Sources）、MACS和D-MACS这4种方法的声源定位性能，验证了D-MACS方法在声源定位性能上的优势。

An improved acoustic deconvolution method for localizing correlated sound sources

In aerodynamic noise tests, the performance of sound source localization algorithm determines the accuracy of the localization of aeroacoustic sources. In this paper, D-MACS(Developed-Mapping of Acoustic Correlated Sources), an improved acoustic deconvolution method for the localization of the correlated sound sources, is proposed to enhance the robustness and accuracy when localizing the acoustic correlated sources. The characteristics of this new method are as follows. Base on the traditional MACS algorithm, the acoustic deconvolution model is adapted, forming a mathematical relationship between the cross-spectrum beamforming matrix and the covariance matrix of sound pressure at the observation points. And the sound source localization result obtained by this proposed method can increase the robustness and accuracy of the MACS method. Using the simulated sound source signals and that of the experimental data obtained in an acoustic wind tunnel, four methods including DAS(Delay And Sum beamformming),DAMAS-C(Deconvolution Approach for the Mapping of Acoustic Correlated Sources), MACS and D-MACS are compared in terms of their ability to localize the sound sources, verifying the superiority of D-MACS in sound source localization.