STUDY OF SOUND SOURCE IDENTIFICATION BASED ON BEAMFORMING AND PATCH NEAR-FIELD ACOUSTIC HOLOGRAPHY

Abstract

A study for applying data extrapolation technique based on wave superposition method (WSM) is proposed to overcome the disadvantages of near-field acoustical holography (NAH). Unlike conformal NAH, where the measurement surface surrounds the entire structure, in the patch holography the measurement surface needs only be approximately as large as the patch on the structure surface where the reconstruction is required. At first, a microphone array is used to acquire the sound pressure field radiated from a vibrator; then a beamforming method is adopted to locate sound sources; after that, a serial of equivalent sources are collocated around these sound sources; at last, a data extrapolation technique based on WSM is applied to extend the measurement aperture and reconstruct the sound field. Since the data extrapolation algorithm requires the inversion of Green's function matrix which may be ill-conditioned, Tikhonov regularization method is used to invert it, and the value of the regularization parameter is determined by the L-curve criteria. The effectiveness of this method is demonstrated by numerical simulation with two pulse ball model, and also experiment is carried out in a semi-anechoic chamber using two sound boxes. The results confirm that the exterior sound field can be accurately reconstructed with few iterative times and the reconstruction error is sufficiently suppressed by the proposed method.