基于经验模态分解的f-x域面波压制方法  

Ground roll suppression in f-x domain based on empirical mode decomposition

作　　者：董烈乾 李振春[1] 杨少春[1] 李凤磊 王娇 

机构地区：[1]中国石油大学(华东)地球科学与技术学院,山东青岛266555 [2]中国石油集团新疆油田分公司,新疆克拉玛依834000

出　　处：《石油地球物理勘探》2013年第1期42-48,5,共7页Oil Geophysical Prospecting

基　　金：国家自然科学基金项目(40974073);国家油气重大专项课题(2011ZX05006-002);国家"863"课题(2009AA06Z206);CNPC物探重点实验室华东研究室联合资助

摘　　要：为了克服f-k滤波规则空间采样的限制和由于滤波窗口选取不当造成的有效信号损失或面波压制不完全的缺点,本文提出了基于经验模态分解(EMD)的f-x域面波压制的方法。经验模态分解可以自适应地按照频率由高到低将信号分解为不同的固有模态(IMF)分量,在空间方向上可以看作是波数由高到低排列,且IMF1分量主要表征高波数的数据。根据面波主要分布在低频高波数域的特性,在f-x域沿空间方向提取低频段数据进行EMD,舍弃分解得到的IMF1分量,就可以有效地压制面波。EMD对数据是否规则采样要求并不严格,这种方法也可以看作是自适应高切波数f-k滤波。模型及实际资料处理结果表明,这种方法较f-k滤波更能有效地压制面波并且保留有效信号。In order to overcome limits of regular trace spatial sampling in f-k filtering and signal loss or incomplete ground roll suppression due to inaccurate filtering window selection,we introduce in this paper a new method based on empirical mode decomposition(EMD)for ground roll suppression.EMD decomposes adaptively from signal high frequency to low frequency into oscillating components called intrinsic mode functions(IMFs),which can be also regarded as wave number ranges from high to low in spatial direction.Moreover,IMF1represents mostly high wave number data.Since ground rolls are mostly distributed in low frequency and high wave number,we can remove effectively ground rolls by casting out IMF1of low frequency obtained by EMD in special sequence.In addition,EMD does not need regular trace spatial sampling.Therefore,this method can be regarded as an auto-adaptive f-k filter with a frequency-dependent,high-cut wave number filtering property.Applications to model data and real field data indicate that this method suppresses not only ground rolls,but also greatly maintains signals.

关 键 词：经验模态分解 固有模态分量 自适应f-k滤波 面波 

分 类 号：P631[天文地球—地质矿产勘探]