压缩感知算法的低频声源定位研究  

Research on Low Frequency Sound Source Localization Based on Compressed Sensing Algorithm

作　　者：朱亚辉 章林柯[1] 刘畅[1] 国玉阔 李祥祥 ZHU Yahui;ZHANG Linke;LIU Chang;GUO Yukuo;LI Xiangxiang(School of Energy and Power Engineering,Wuhan University of Technology,Wuhan 430063,China)

机构地区：[1]武汉理工大学能源与动力工程学院,武汉430063

出　　处：《武汉理工大学学报（交通科学与工程版）》2021年第2期288-291,共4页Journal of Wuhan University of Technology（Transportation Science & Engineering）

摘　　要：针对在波束形成声源定位中小孔径传声器阵列低频声源定位会产生大的主瓣宽度导致空间分辨率差问题,基于波束形成压缩感知L1范数奇异值分解(L1-SVD)算法建立传声器阵列声源定位模型,通过仿真对200 Hz和500 Hz的低频声源定位研究,结果表明压缩感知L1-SVD算法对低频声源具有高的分辨率、旁瓣衰减和抗干扰能力,且500 Hz相干声源定位比200 Hz抑制噪声干扰能力强.To solve the problem of poor spatial resolution caused by large main lobe width in low-frequency sound source localization of small aperture microphone array in beamforming,a microphone array sound source localization model was established based on the L1 norm singular value decomposition(L1-SVD)algorithm of beamforming compressed sensing.The localization of low-frequency sound sources at 200 Hz and 500 Hz was studied by simulation.The results show that the compressed sensing L1-SVD algorithm has high resolution,sidelobe attenuation and anti-interference ability for low-frequency sound sources,and the coherent sound source localization at 500 Hz is better than that at 200 Hz in suppressing noise interference.

关 键 词：声源定位 低频 阵列 压缩感知 

分 类 号：U467.493[机械工程—车辆工程]