基于传递矩阵修正的近场MUSIC算法研究

Research on near-field MUSIC algorithm based on transfer matrix correction

针对MUSIC谱估计算法在近场声源定位成像中出现的边沿放大问题,提出了一种阵列传递矩阵修正方法,该方法不考虑近场声压幅值衰减效应。实现了近场条件下的MUSIC算法声源定位成像的应用,并通过仿真与直接波束形成算法、去自谱波束形成算法进行了对比研究,检验了该算法的性能。仿真实验证明采用修正阵列传递矩阵后的MUSIC算法可以实现近场声源定位成像,可作为较近距离和较低频率条件下一种补充定位成像算法。

The margin-amplification phenomenon in near-field acoustic image is studied and amend method for array＇s manifold matrix is brought up. This method doesn＇t take into account the near-field sound pressure amplitude attenuation effect.By this way,the super-resolution MUSIC algorithm is applied in near-field sound source localization. The performance is testified and compared with direct beamforming algorithm and de-spectral beamforming algorithm by simulate model. The simulation results show that this method can realize near-field sound source localization imaging,and it can be an ancillary localization algorithm in closer distance and lower frequencies.