Adaptive Grid Hierarchical Localization Algorithm Based On Spherical Array

Streered response power(SRP)algorithm is a microphone array-based sound source localization method with high accuracy and reliability,which has been widely used in scenarios such as array direction finding and noise localization.However,the SRP algorithm determines the incident direction of the sound source by searching the global grid,which leads to its large computational volume and low computational efficiency.In contrast,in array signal processing,compared with the commonly used array forms,rigid spherical microphone arrays can perform high-precision three-dimensional analysis of the sound field and perform spherical harmonic domain data processing under the condition of no spatial ambiguity.Therefore,this paper investigates the SRP(Spherical harmonics SRP,SH-SRP)source localization algorithm based on the spherical harmonic function,and proposes an adaptive grid hierarchical localization SRP algorithm(MUSIC and relative entropy joint localization SRP algorithm MRE-SRP)based on the rigid spherical microphone array.The performance of the MRE-SRP algorithm is verified through analysis,simulation and experiments,and the main research contents are as follows:(1)The theory and algorithm of spatial spectrum estimation in the array element domain and spherical harmonic domain are studied.The conventional beamforming algorithm and its expansion in the spherical harmonic domain are derived,and the spatial spectral estimation performance of different algorithms in different domains is analyzed through simulations to give the beam maps.(2)To address the current problem of low computational efficiency of SH-SRP,the MRESRP algorithm is proposed on the basis of the existing grid hierarchy algorithm in the spherical harmonic domain.Secondly,the spatial relative entropy model is introduced into the SH-SRP algorithm to extract the information gain before and after grid stratification by the entropy difference function,which provides the basic basis for grid selection.Finally,in order to ensure the positioning accuracy of SH-SRP algorithm and reduce the computational effort of the algorithm,the adaptive grid selection criterion is designed to realize the accurate subdivision of hierarchical multi-resolution grid.(3)Analysis and simulation of the theoretical and application performance of the MRESRP algorithm.Firstly,we analyze the features of MRE-SRP algorithm that are different from existing methods,the computational complexity and the factors that affect the localization performance of the algorithm.The simulation results show that the MRE-SRP algorithm is more robust to noise and can effectively reduce the computation time while ensuring the localization accuracy.(4)Experimental verification of MRE-SRP algorithm performance.The test system was designed and built in three different environments: semi-anechoic chamber,indoor and outdoor field,and the average positioning error and computational efficiency of the test results were analyzed.The experimental results show that the MRE-SRP algorithm can accurately select the hierarchical grid for single and dual sound sources localization in different environments,and significantly reduce the computational effort of the SH-SRP algorithm without degrading the localization accuracy.