基于局部近场声全息的噪声源辨识研究

Study on Sound Source Identification Based on Patch Near-field Acoustical Holography

基于波叠加的理论思想,提出了一种基于波叠加方法的迭代算法来实现全息数据外推以解决传统近场声全息要求全息孔径大的问题,实现了近场声全息对局部声场的重建。由于噪声源识别问题对测量误差非常敏感,采用了结合L曲线准则的Tikhonov正则化方法进行正则化滤波。在半消声室内采用音响模拟噪声源进行验证,实验结果表明,该方法在小全息孔径条件下可以准确地对辐射体进行声场重构,减少了对全息孔径大小的要求,为其在工业现场中的应用奠定了基础。

In view of the principle of wave superposition method, an iterative procedure based on wave superposition was proposed to realize the data extrapolation of measurement sound pressure and to overcome the disadvantage of the FFT--based NAH, and realized NAH to reconstruct partial sound field. Since the sound field reconstruction is very sensitive to the measurement errors, Tikhonov regularization method was used to invert it, and the value of the regularization parameter was determined by the L--curve criteria. The experimental results show that the proposed method is successful in reconstructing sound field under the conditions of small measurement aperture; so reduces the demand of the size of measurement aperture; and also lays a foundation for its industrial field application.