基于FFT-OMP-DAMAS波束形成方法的汽车前围板隔声薄弱部位识别

为实现汽车前围板隔声薄弱部位的准确识别，文章提出了基于快速傅里叶变换(Fast Fourier Transform,FFT)和正交匹配追踪(Orthogonal Matching Pursuit,OMP)的反卷积(Deconvolution Approach for the Mapping of Acoustic Sourc‐es,DAMAS)波束形成方法(FFT-OMP-DAMAS)。该方法基于声源稀疏分布假设，利用正交匹配追踪思想求解反卷积问题，并进一步结合傅里叶变换和点扩散函数空间转移不变假设降低计算维度。在混响室-消声室内，分别利用延迟求和方法，DAMAS方法和FFT-OMP-DAMAS方法进行了某汽车前围板隔声薄弱部位识别试验，结果表明：FFTOMP-DAMAS方法能够有效抑制旁瓣和伪源，有效缩减主瓣宽度，从而准确识别汽车前围板隔声薄弱部位，且相较于传统的DAMAS方法，文中提出的FFT-OMP-DAMAS方法能获得更清晰的成像结果，计算效率有了明显提高。

Weak acoustic insulation area identification of car dash panel based on FFT-OMP-DAMAS beamforming

In order to achieve the accurate identification of the weak acoustic insulation areas of a car dash panel, a deconvolution approach to the mapping of acoustic sources (DAMAS) based on fast Fourier transform (FFT) and orthogonal matching pursuit (OMP) is proposed in this paper, which is named as FFT-OMP-DAMAS for short. Its core is to utilize the orthogonal matching pursuit based on the source sparse distribution assumption to solve the deconvolution problem, and further combine the Fourier transform and the spatial shift invariant assumption of the point spread function (PSF) to reduce the computational dimension. In this paper, the delay and sum (DAS), DAMAS and FFT-OMP-DAMAS beamforming methods are used to identify the weak acoustic insulation areas of a car dash panel in the reverberation chamber-anechoic chamber. The results show that the FFT-OMP-DAMAS beamforming can effectively suppress side lobes and spurious sources, and reduce the width of main lobe, so as to accurately identify the weak acoustic insulation areas of a car dash panel. In addition, compared with the traditional DAMAS method, the FFT-OMP-DAMAS method proposed in this paper can obtain clearer imaging results and significantly improve the computational efficiency.