基于麦克风阵列的运动声源识别技术  

Research of moving noise source recognition based on microphone array

作　　者：周家检[1] 郝璇[1] 陈大斌[1] 付增良[1] 梁彬[1] 张旭[1] ZHOU Jiajian;HAO Xuan;CHEN Dabin;FU Zengliang;LIANG Bin;ZHANG Xu(China Academy of Aerospace Aerodynamics,Beijing 100074,China)

机构地区：[1]中国航天空气动力技术研究院,北京100074

出　　处：《空气动力学学报》2022年第5期61-70,共10页Acta Aerodynamica Sinica

摘　　要：麦克风相阵列测量技术是气动噪声研究的重要手段。针对利用麦克风相阵列技术识别运动声源的难点,对时域波束形成算法、解多普勒效应技术等核心技术开展了研究,建立了适用于运动声源的麦克风相阵列测量技术。通过运动声源仿真,验证了阵列数据处理算法,进而分析了航迹误差和数据长度对声源识别的影响。提出了一种考虑声传播相关性丢失问题的阵列优化设计方法,设计了满足无人机噪声飞行试验需求的阵列。在此基础上,基于某型无人机噪声飞行试验开展运动声源识别技术验证,实现了起飞和低空平飞状态螺旋桨和发动机噪声源的定位与识别。Microphone phased array technique is an important measure for aeroacoustics research. In accordance with the difficulties of moving noise sources recognition, the core technology such as time domain beamforming algorithm and de-dopplerization methods were studied and the measurement technique of microphone phased array used for moving object was developed. The data processing algorithm is validated with the simulated noise source, as well as the influence of trajectory error and data length. In considering the loss of coherence due to atmospheric turbulence during propagation from noise sources on the aircraft to microphones on the ground, a kind of array optimization method was proposed to design the array suitable for UAV noise test.A flight test of UAV noise was carried out and the noise source of propeller and engine at take-off and lowaltitude cruise were identified successfully.

关 键 词：相阵列 时域波束形成 运动声源 解多普勒效应 螺旋桨噪声 

分 类 号：V211.7[航空宇航科学与技术—航空宇航推进理论与工程]