基于频率-速度域多重信号分类的面波高分辨率频散成像方法

准确的相速度频散图像是主动源面波勘探方法反演近地表横波速度的基础。提出了一种基于频率-速度域多重信号分类(multiple signal classification in frequency-velocity domain)的面波高分辨率频散成像方法(简称fv-MUSIC方法)。该方法首先将传统的频率-波数域波束形成器改造成频率-速度域形式,然后引入多重信号分类算法将空间谱相关矩阵分解为信号子空间和噪声子空间两个部分,最后利用噪声子空间部分生成最终的面波频散图像。理论数据和实际数据应用结果表明,该方法能产生较高精度的相速度图像,并且使用方便,计算效率高,尤其当接收排列较短时,该方法依然能保持较高的相速度分辨率,有利于提高主动源面波方法的横向速度分辨能力。

Accurate phase velocity dispersion image is the basis of surface wave inversion for near-surface shear wave velocity.We present a high resolution dispersion imaging method of surface wave named multiple signal classification in frequency-velocity domain (fv-MUSIC) based on multiple signal classification in frequency velocity domain.Firstly,a beam-former is constructed in frequency velocity domain instead of the conventional frequency wavenumber domain.Secondly,the MUSIC algorithm is introduced to decompose the spatial-spectral cross-correlation matrix into signal and noise subspaces.At last,the ultimate surface wave dispersion image is produced using the noise subspace.Theoretical and real surface wave data processing cases indicate that it has the advantages of convenient operation,less calculation time and high resolution phase velocity images.Especially when the length of receiving array is shorter,the fv-MUSIC method still remains higher phase velocity resolution.It helps to improve the lateral shear wave velocity resolution of the active source surface wave method.

国家自然科学基金项目(41374123)资助。