Inverse Scheme for Acoustic Source Localization using Microphone Measurements and Finite Element Simulations

Manfred Kaltenbacher,Stefan Gombots,Barbara Kaltenbacher

doi:10.3813/aaa.919204

Abstract

Inverse Scheme for Acoustic Source Localization using Microphone Measurements and Finite Element Simulations

Highlights

Acoustic source localization is a main task in the development of new products
The beamforming map a is computed by a(g) = gT Cg with g denoting the steering vector [2], C the cross spectral matrix (CSM) of the microphone array signals and T the transposed
We demonstrate that the identification with beamforming algorithms does not work well for such a case

Summary

Introduction

Acoustic source localization is a main task in the development of new products. In the last years, considerable improvements have been achieved in acoustic source localization using microphone arrays. Kaltenbacher et al.: Inverse Scheme for Source Localization (2) Clean-SC has the best trade-off between fast computation and correct source detection; (3) for a case of two orthogonal microphone arrays only L1-GIB results in good 3D source maps at reasonable computational costs Despite these advances in beamforming techniques, it has to be mentioned that major limitations arise from the source model. The third category of methods for source localization is provided by solving the inverse-source problem based on a constrained minimization problem, where a cost functional is minimized under the constraint that the forward problem (wave equation with source term) is fulfilled This approach assumes that the material and geometric properties of the domain of interest are known, and aims at finding the position and strength of all sources.

Physical and mathematical model

Optimization based source identification

Numerical results

Conclusion

1: DefineParameters:see Table A1