一种基于麦克风阵列用于分离单极子和偶极子声源的方法

A method for separation of monopole and dipole sources based on phased microphone array

基于单极子假设的波束成形算法作为一种声场可视化技术,在声源识别中得到了广泛应用。但是在实际工程应用中,声源的类型较为复杂,基于单一声源假设的波束成形无法有针对性地识别不同类型的声源。本文提出了一种混合反卷积方法来分离含有单极子和偶极子的组合声源,该方法构造了波束成形输出与实际声源之间的线性方程,通过求解该线性方程,可以将单极子和偶极子从组合声源中提取出来,并通过多组仿真和实验来检验该混合反卷积算法。结果表明此方法确实可以有效地分离组合声源,并且保证了声源强度的准确性,即使在包含多个声源时依然有效。此方法有望应用于航空发动机气动噪声识别,从高速喷流噪声中提取目标源,更好地研究喷流噪声的成分。

As an acoustic field visualization technology,beamforming based on the monopole assumption has been widely applied in identifying acoustic sources.However,in practical engineering applications,complex types of sound sources make it difficult for beamforming based on the single sound source assumption to identify different types of sound sources pertinently.This paper proposes a hybrid deconvolution method to separate the combined sources containing monopoles and dipoles.The approach constructs a linear equation between the beamforming output and the actual sound source distribution,and monopoles and dipoles can be extracted from the combined sources by solving this linear equation.Four simulation cases and three experimental cases are designed to check the hybrid deconvolution algorithm.The combined sources in the experiment are composed of a dipole formed by a cylindrical spoiler and a monopole caused by a speaker.The results indicate that this method can separate the combined sound sources effectively and ensure the accuracy of the sound source strength,despite the multipoles.This method is expected to be applied in aerodynamic noise recognition,extracting target sources from high-speed jet noise,and better studying the composition of jet noise.