Extensions of the orthogonal beamforming to identify multiple multipole sources

Xingjian PAN,Weikang JIANG
2019-01-01
Abstract:Identification of sources with complex directivities is attracting great attentions in aeroacoustic measurements these years. A multipole orthogonal beamforming algorithm was proposed to identify multipoles with arbitrarily oriented axes with no prior knowledge about source types. Independent sources were separated by the eigenvalue decomposition of the cross spectral matrix of signals measured by microphone arrays. The influence of source directivities on source imaging was eliminated by normalizing and squaring the eigenvectors. The delay-and-sum procedure was performed on each modified eigenvector to localize sound sources. The proposed algorithm was verified by simulations and experimental trials using a combination of a loudspeaker and an Aeolian tone dipole source. The accuracy of estimated source positions is improved and the resolution of the proposed algorithm is comparable to that of deconvolution approaches.
What problem does this paper attempt to address?