Research On Testing Technology Of Microphone Array Acoustic Imaging

Acoustic imaging technology based on microphone array is widely used in the fields of noise control,aeroacoustics,monitoring and tracking,etc.acoustic imaging technology uses beamforming algorithm and image fusion to image the distribution information of spatial sound source,which can quickly identify and locate the noise source on the intuitive image.Positioning accuracy and spatial resolution are the criteria to evaluate the performance of acoustic imaging,then array formation and beamforming algorithm are the key factors to affect the performance.This paper focuses on the array formation advantages and disadvantages of spiral microphone array and deconvolution beamforming.Finally,the application of acoustic imaging technology in the capture of illegal car whistle is also studiedThe main content of this paper is divided into four parts:In the first part,the basic theory of acoustic imaging is studied,in which array vector and cross spectrum matrix are two key physical quantities.Array vector reflects the response performance of microphone array.Cross spectrum moment matrix contains the spatial information and energy intensity information of sound source.Beamforming algorithm recovers the distribution of actual sound source by array vector and cross spectrum matrix.The second part studies the optimized microphone array design from the dynamic range of the beam and the 3dB bandwidth.The simulation compares the performance of the regular array and the spiral array,and finds that the spiral array can effectively suppress the side lobe.The optimization idea of spiral array is: the high concentration of microphone near the center of array will get a wide dynamic range,and the high concentration of microphone near the edge of aperture will get a high resolution,and the optimal array performance can be obtained by balancing the two factors.The third part deeply studies various beamforming algorithms.DAMAS and DAMAS-C assume that the beamforming result is a linear combination of point spread function model and point source,restore the distribution of incoherent and coherent sound sources by inverse solution.CLEAN-PSF and CLEAN-C beamforming eliminate sidelobe interference by CLEAN algorithm.Functional beamforming uses matrix function method to suppress sidelobe and improve resolution.The deconvolution beamforming algorithm represented by DAMAS greatly improves the resolution of the array,but the calculation speed is limited by the number of scanning grids.For this reason,sparse deconvolution beamforming is introduced,which uses the characteristics of the sparse distribution of the sound sources on the scanning grid to quickly restore the sound source distribution by sparse approximation.In the fourth part,the application of acoustic imaging technology in the capture of illegal whistle is studied.If the whistle is detected,the location of the whistle vehicle will be quickly located,then merge the beamforming map and the video image to generate evidence of illegal whistle.In this paper,an improved algorithm DAMAS-FAST is proposed for single car whistling,which can locate the whistling car quickly and accurately,and the implementation complexity is low.