基于六元传声器阵列声源定位算法  

Sound Source Localization Algorithm Based on Six-Element Microphone Array

作　　者：刘彦峰 杨瑞峰 郭晨霞 LIU Yanfeng;YANG Ruifeng;GUO Chenxia(School of Instrumentation and Electronics,North University of China,Taiyuan 030051,China)

机构地区：[1]中北大学仪器与电子学院,山西太原030051

出　　处：《测控技术》2025年第3期18-24,共7页Measurement & Control Technology

基　　金：山西省中央引导地方科技发展自由探索类基础研究项目(YDZJSX20231A032)。

摘　　要：随着声源定位广泛应用于生产生活,传统的麦克风阵列广义互相关相位变换(Generalized Cross Correlation-Phase Transform,GCC-PHAT)加权算法的定位精度不足以支撑其在噪声和混响水平较高的情况下进行精确定位。为了提高在低信噪比环境下的定位精度,提出了一种改进定位方法。首先,在平面五元十字阵列的基础上,在原点上方z轴加上一个麦克风。然后,在使用二次相关的基础上,对加权函数和时延估计进行改进,进行以信噪比为参数的指数运算和虚假峰值消除。最后,通过几何定位,将其分为2组五元阵列,分别求解后求均值,并且通过角度将其分为2种,以防止极端角度的影响。在各种环境条件下进行的实验结果表明,改进算法相较于传统GCC-PHAT算法,在时延估计、距离定位和角度定位方面均有显著的性能提升。尤其是在低信噪比场景下,改进算法有着更高的定位精度和较强的抗干扰能力。As sound source localization is widely used in production and life,the positioning accuracy of the traditional microphone array generalized cross correlation-phase transform(GCC-PHAT)weighted algorithm can not support its precise positioning under high noise and reverberation levels.In order to improve the positioning accuracy in a low signal-to-noise ratio environment,an improved positioning method is proposed.Firstly,a microphone is added to the z axis above the origin based on the planar five-element cross array.Then,based on quadratic correlation,the weighted function and delay estimation are enhanced,and the signal-to-noise ratio is employed as the parameter for exponential calculation and false peak elimination.Finally,through geometric positioning,it is divided into two groups of five-element arrays,and the mean is calculated after solving them separately.It is also divided into two types by angle to prevent the influence of extreme angles.The experimental results conducted under various environmental conditions show that the improved algorithm has significant performance improvements in time delay estimation,distance localization,and angle localization compared to the traditional GCC-PHAT algorithm.Especially in low signal-to-noise ratio scenarios,the improved algorithm has higher positioning accuracy and stronger anti-interference ability.

关 键 词：声源定位 时延估计 二次相关 传声器阵列 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]