摘要
针对多分量地震数据具有信噪比低、波场比较复杂等特点,同时为了解决有效波和面波在低频带具有相关性且频带有重叠的部分.本文基于希尔伯特—黄变换(Hilbert—Huang transform,HHT)提出一种多分量地震数据去噪方法.HHT的核心部分是固有模态分解(Empirical mode composition method,EMD),它将地震信号自适应分解为各个尺度的固有模态函数(Intrinsic mode function,IMF);然后根据有效波和噪声分布在不同的分辨尺度范围内进行数据重构;对于一些尺度分布比较宽的,进行更加严格的EMD分解,以及通过瞬时参数阈值大小进行处理以达到更好的去噪效果.本文所做的工作主要针对多波资料中面波的处理,利用面波跟有效波的频率尺度和瞬时属性差异,从仿真数据、模型和实际地震数据进行面波的压制实验,取得了比较满意的压制效果,提高了地震资料的分辨率.也表明了该方法在多分量地震去噪处理中的可行性.
For multi-component seismic data with low signal-tonoise ratio、complex wave field,Moreover,the effective wave and surface wave have overlap parts in low frequency band,Between them are related.In order to solve these problems.We present a novel method in multi-component seismic data,namely Hilbert-Huang transform（HHT）to suppress noise.The core of the approach is inherently Mode Decomposition（EMD）,which decomposed signals into each scale intrinsic mode function（IMF component）.which are all singlecomponent signals;Then we reconstructed data according to the effective wave and noise distribution in different resolution scales,For some relatively wide scale distribution, more stringent EMD decomposition been carried on,and we can achieve better denoising effect through the instantaneous parameter thresholding.This article focused on the work done in multi-wave surface wave data processing.Between the effective wave and Rayleigh wave,There are differences on the Frequency scale and instantaneous attributes,which would been used. We do a lot of suppression of surface wave experiment from the simulation data,model and actual seismic data, achieve satisfactory denoising effect,improve the resolution of seismic data.Also Show that this method is feasible in the Seismic noise elimination processing.
出处
《地球物理学进展》
CSCD
北大核心
2015年第6期2846-2854,共9页
Progress in Geophysics
基金
国土资源部海洋油气资源与环境地质重点实验室基金项目(MRE201414)资助
关键词
希尔伯特—黄变换
EMD分解
IMF分量
多波多分量
面波
Hilbert-Huang transform
EMD decomposition
IMF component
multi-component multi-wave
surface wave
