基于压缩奇异值分解等效源法的结构板件声源识别

为改善空间连续型声源的声场重建与声源识别性能,基于压缩感知(compressed sensing,CS)和等效源法(equivalent source method,ESM)的基本理论,提出了一种压缩奇异值分解等效源法(CSVDESM)。CSVDESM通过奇异值分解法获取声场的一系列正交基,在ESM和CS框架的基础上实现对声场的重构。将CSVDESM与高阶矩阵函数波束形成理论结合,通过提高阶次值,不断缩小识别到的声学中心覆盖范围,进一步提高声源识别定位精度。数值仿真分析和实验应用均验证了CSVDESM的有效性与实用性。

The sound source identification of structural panels based on compressed singular value decomposition equivalent source method

To improve the performance of sound field reconstruction and sound source identification for the spatially extended source, a compressed singular value decomposition equivalent source method (CSVDESM), based on compressed sensing (CS) theory and the equivalent source method (ESM), is proposed in this paper. The CSVDESM obtains a series of orthogonal basis of the source field by the singular value decomposition (SVD), and reconstructs the sound field on the basis of the ESM and the CS framework. In addition, when combined with the high-order matrix function beamforming, CSVDESM can further improve the accuracy of sound source identification by increasing order value to narrow the identified acoustic center coverage continuously. Numerical simulation and experiment verify the validity and practicality of CSVDESM.