改进的F-X域EMD去噪技术及分布式并行实现
|
摘要
在地震数据处理领域,较高的计算效率及较好的处理效果一直是地球物理工作者追求的目标。介绍了改进的F-X域(频率-空间域)经验模态分解(Empirical Mode Decomposition,EMD)方法,首先将地震信号转换到F-X域后分解成一系列固有模态函数(Intrinsic Mode Functions,IMF),然后通过小波阈值滤波对固有模态函数除第一个分量之外的其他分量进行滤波处理来达到去噪目的。此外,为解决改进后的F-X域EMD去噪方法计算效率问题,提出了分布式并行算法,兼顾了运算效率与计算精度。数值模拟结果表明,与F-X域预测滤波技术相比,改进后的F-X域EMD去噪方法可以更加有效地衰减随机噪声和压制线性干扰;最后将该方法应用于实际地震数据处理,进一步验证了该方法的有效性和优越性。
For seismic data processing,improving calculation efficiency and seeking well treatment effect were the constant goals of geophysicists.The improved empirical mode decomposition(EMD) in F-X domain(F-X-EMD) was mainly introduced for noise suppression.Firstly,the seismic signals were decomposed into a series of intrinsic mode functions(IMF) after they were transferred into F-X domain removing the first component,other components of IMF were filtered by wavelet threshold filtering for noise attenuation.Furthermore,in order to perfect the calculating efficiency of improved F-X-EMD,the distributed parallel algorithm with higher computational efficiency and accuracy was presented.The numerical simulation results show that compared with prediction of filtering in F-X domain,the improved F-X-EMD is more effective in random noise attenuation and coherent noise suppression.Finally,the real seismic data results show that the new method has effectiveness and superiority in seismic data processing analysis.
For seismic data processing,improving calculation efficiency and seeking well treatment effect were the constant goals of geophysicists.The improved empirical mode decomposition(EMD) in F-X domain(F-X-EMD) was mainly introduced for noise suppression.Firstly,the seismic signals were decomposed into a series of intrinsic mode functions(IMF) after they were transferred into F-X domain removing the first component,other components of IMF were filtered by wavelet threshold filtering for noise attenuation.Furthermore,in order to perfect the calculating efficiency of improved F-X-EMD,the distributed parallel algorithm with higher computational efficiency and accuracy was presented.The numerical simulation results show that compared with prediction of filtering in F-X domain,the improved F-X-EMD is more effective in random noise attenuation and coherent noise suppression.Finally,the real seismic data results show that the new method has effectiveness and superiority in seismic data processing analysis.
引文
[1]Huang N E,Shen Z,Long S R,et al.The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationarytime series analysis[J].The Royal Society,1998,454(1971):903～995.
[2]Huang N E,Wu M C,Long S R,et al.A confidence limit for the empirical mode decomposition and hilbert spectral analysis[J].Pro-ceedings of the Royal Society of London,2003,459(2037):2317～2345.
[3]Abdel O B,Jean-Christophe C.EMD-Based Signal Filtering[J].IEEE Transactions on Instrumentation and Measurement,2007,56(6):2196～2202.
[4]Bekara M,Baan M.F-X Noise Attenuation by Empirical Mode Decomposition[A].70th EAGE Conference&Exhibition[C].Rome,2008-06-9～12.
[5]Bekara M,Baan M.Random and coherent noise attenuation by empirical mode decomposition[J].Geophysics,2009,74(5):89～98.
[6]Luszczek P.Enhancing multicore system performance using parallel computing with MATLAB[J].MATLAB Digest,2008,17(5):1～7.
[7]Sharma G,Martin J.MATLAB:A language for parallel computing[J].Int J of Parallel Prog,2009,37(1):3～36.
[2]Huang N E,Wu M C,Long S R,et al.A confidence limit for the empirical mode decomposition and hilbert spectral analysis[J].Pro-ceedings of the Royal Society of London,2003,459(2037):2317～2345.
[3]Abdel O B,Jean-Christophe C.EMD-Based Signal Filtering[J].IEEE Transactions on Instrumentation and Measurement,2007,56(6):2196～2202.
[4]Bekara M,Baan M.F-X Noise Attenuation by Empirical Mode Decomposition[A].70th EAGE Conference&Exhibition[C].Rome,2008-06-9～12.
[5]Bekara M,Baan M.Random and coherent noise attenuation by empirical mode decomposition[J].Geophysics,2009,74(5):89～98.
[6]Luszczek P.Enhancing multicore system performance using parallel computing with MATLAB[J].MATLAB Digest,2008,17(5):1～7.
[7]Sharma G,Martin J.MATLAB:A language for parallel computing[J].Int J of Parallel Prog,2009,37(1):3～36.