Research And Implementation Of Sound Source Localization System Based On Microphone Array

The sound source localization technology based on microphone arrays is widely used in fields such as noise control,monitoring and tracking,and aerodynamic acoustics.Sound source localization is the process of collecting sound source information through a microphone array,and then calculating the sound source location through data processing and sound source localization algorithms.There are many factors that affect the accuracy of sound source localization results,among which the selection of array formation and localization algorithm is the most important influencing factor.This article designs a sound source localization system through the study of microphone array formation and localization algorithms,and applies it in practical applications such as discharge detection and gas leakage.The main research content of this article is as follows:In terms of array formation optimization,the performance of common twodimensional arrays such as Archimedean spiral arrays,Dougherty multi helix arrays,and Underbridge multi helix arrays was compared,and the array structures with better performance were analyzed;We studied common array optimization algorithms such as particle swarm optimization and simulated annealing.Based on this,we studied a simulated annealing algorithm based on particle swarm optimization.MATLAB simulation tools were used to verify the improved algorithm’s better performance.Based on this,we designed an array structure with narrower 3d B bandwidth and smaller maximum sidelobe level.In terms of research on sound source localization algorithms,conventional beamforming algorithms,MUSIC algorithms,and DAMAS localization algorithms were studied,and the advantages,disadvantages,and applicable environments of the algorithms were analyzed;By studying the DAMAS localization algorithm,improvements are made to the DAMAS algorithm.The improved algorithm mainly uses conventional beamforming algorithms for coarse localization,reducing the number of grids divided,thereby reducing the computational time of the algorithm and improving the accuracy of localization.Through experimental verification,the improved DAMAS algorithm studied has better performance compared to traditional algorithms.Finally,this article designs a sound source localization system,including hardware design and software development,to achieve the effect of real-time display of sound source position on video images,achieving the purpose of acoustic imaging.Through experiments,verify the practicality and accuracy of the positioning of the system developed in this article,and demonstrate the feasibility of the system.