Deconvolution sound source imaging method for identifying coherent and incoherent sound sources

Abstract

The invention discloses a deconvolution sound source imaging algorithm suitable for coherent and non-coherent sound sources. The algorithm includes the steps of obtaining sound pressure data of a measurement surface by using a sensor and performing denoising processing to obtain denoised sound pressure data; calculating the delay summation output at focus points on a focus surface for the denoisedsound pressure data by using a delay summation algorithm; and constructing a new point spread function matrix corresponding to a delay summation beam-forming result by using a sound pressure Green function and an array steering vector, and establishing a convolution relationship among the delay summation beam-forming output result, the sound source intensity distribution, and the new point spreadfunction matrix. The deconvolution method is used to solve the sound source intensity so as to achieve accurate positioning of a noise source. The deconvolution sound source imaging algorithm in theinvention can be simultaneously applied to the recognition of coherent sound sources and non-coherent sound sources, and has good noise robustness and a much higher calculating efficiency than that ofthe existing deconvolution sound source imaging methods that can be used for coherent sound sources. Therefore, the algorithm has a wider range of application and a practical significance.

Description

technical field [0001] The invention relates to the field of identification and location of noise sources, more specifically a deconvolution sound source imaging method suitable for identifying coherent and non-coherent sound sources. Background technique [0002] Deconvolution Acoustic Source Imaging (DAMAS) is a high-resolution noise source identification and localization method based on acoustic arrays. This method is based on the incoherent assumption of spatial sound sources, and establishes the output results of cross-spectral imaging beamforming and the real sound source distribution. And the convolution relationship between the array point spread function (PSF) matrix, the real sound source distribution is solved by deconvolution calculation, so as to eliminate the influence of non-ideal PSF on the beamforming output results, and effectively reduce the main lobe width and side lobe level, for the purpose of significantly improving the spatial resolution of sound sour...

AI Technical Summary

Problems solved by technology

It can be seen that due to the low computational efficiency of the DAMAS-C algorithm, it is not practical in practice

In addition, Yardibi et al. proposed a sparse constrained covariance matrix fitting algorithm (CMF-C) for coherent sound source identification. This algorithm has the same idea as the DAMAS-C algorithm. All the elements (including cross terms) of are solved as unknowns to be sought, but the deconvolution calculation process is realized by minimizing the fitting error of the covariance matrix, so although the CMF-C algorithm is slightly higher in recognition accuracy It is based on the DAMAS-C algorithm, but it is equivalent to the DAMAS-C algorithm in terms of computational efficiency, and it is also not very practical

Images