频率、相位可控的反褶积技术及其应用
|
摘要
文中借助常规反褶积算法中信号自相关的Toeplitz矩阵,在求得反滤波因子的前提下,结合地震学中的相位处理及反褶积理论,推导出一种具有较强实用性的频率、相位可控的反褶积算法。处理过程大体可分为信号的相位处理及替换处理两步,通过给定任意频率和相位的理想地震子波构造矩阵,在进行信号反褶积过程中进行地震信号的频率、相位的替换,从而达到改变记录频率及相位特征的目的,在不增加额外噪声的情况下使信号纵向分辨率达到最理想状态。此外,文中方法在求取反滤波因子过程中结合了相位处理技术,由此打破了常规反褶积中最小相位的假设,使得反滤波因子的求取没有相位的限制,在真正意义上实现了任意信号的盲反褶积处理。模拟试验和实际资料处理结果表明,在替换过程中剖面的分辨率随替换子波频率升高而提高,但信噪比并没有降低,子波间波形差异造成的影响随替换子波频率的升高而减小,反褶积前、后资料的能量匹配较好,说明文中方法具有很好的保幅特性,有着较好的应用前景和开发潜力。
With the help of signal autocorrelation Toeplitz matrix in routine deconvolution algorithm,based on estimated inverse filtering factor and by integrating phase processing and deconvolution theory in seismology,a practical frequency and phase controllable deconvolution algorithm was derived in this paper. The processing process in the algorithm contains two parts,the signal phase processing and signal replacement processing. During signal deconvolution the signal's frequency and phase were replaced through defining ideal wavelet with arbitrary frequency and phase to construct matrix,as a result the characteristics of the frequency and phase of the seismic data were modified. In the processing above,no extra noise added but the vertical resolution of the signals was raised to maximum value. In addition,the phase processing was involved in estimating inverse filtering factor,which violates the minimum phase assumption in routine deconvolution but removes phase limitation during estimating inverse filtering factor,in this way blind Deconvolution for arbitrary signals was truly realized. Simulating test and field data processing results show that during replacement process the seismic section's resolution increases with the wavelet frequency 's increase,but the signal to noise ratio is not decreased,the effects caused by waveform difference alleviates with the replaced wavelet's frequency's increase,and there is no big energy change before and after deconvolution,all of those results mentioned above demonstrated that the technique has a excellent amplitude-preserving property,it will have a bright future in application and development.
With the help of signal autocorrelation Toeplitz matrix in routine deconvolution algorithm,based on estimated inverse filtering factor and by integrating phase processing and deconvolution theory in seismology,a practical frequency and phase controllable deconvolution algorithm was derived in this paper. The processing process in the algorithm contains two parts,the signal phase processing and signal replacement processing. During signal deconvolution the signal's frequency and phase were replaced through defining ideal wavelet with arbitrary frequency and phase to construct matrix,as a result the characteristics of the frequency and phase of the seismic data were modified. In the processing above,no extra noise added but the vertical resolution of the signals was raised to maximum value. In addition,the phase processing was involved in estimating inverse filtering factor,which violates the minimum phase assumption in routine deconvolution but removes phase limitation during estimating inverse filtering factor,in this way blind Deconvolution for arbitrary signals was truly realized. Simulating test and field data processing results show that during replacement process the seismic section's resolution increases with the wavelet frequency 's increase,but the signal to noise ratio is not decreased,the effects caused by waveform difference alleviates with the replaced wavelet's frequency's increase,and there is no big energy change before and after deconvolution,all of those results mentioned above demonstrated that the technique has a excellent amplitude-preserving property,it will have a bright future in application and development.
引文
[1]黄绪德.反褶积与地震道反演.北京:石油工业出版社,1992
[2]Saggaf M Mand Enders A R.A unified frame work for the deconvolution of traces of nowhite reflectivity.Geophysics,2000,65(5):1660~1676
[3]Levy S and Fullager P K.Reconstruction of a sparse spick train froma portion of its spectrumand applica-tion to high resolution deconvolution.Geophysics,1981,46:1235~1243
[4]Walden A T and Hosken J WJ.An investigation of the spectral properties of pri mary reflection coeffi-cients.Geophysical Prospecting,1985,33:400~435
[5]Todoeschuck J P and Jense O G.Scaling geology and seismic deconvolution.Pure Appl Geophysics,1989,131:273~288
[6]Warrant T W.Si multaneous deconvolution and wave-let inversion as a global opti mizition.Geophysics,1999,64(4):1108~1115
[7]王大兴,雷晓,潘存焕等.分形反褶积方法研究及应用.石油物探,1996,35(4):93~97,1~10
[8]夏洪瑞,陈德刚,周开明.地震资料处理中随机干扰消除方法分析.石油物探,2003,42(1):93~97
[9]汪军,何振亚.基于高阶谱的信号盲分离.东南大学学报,1996,26(5):75~78
[10]王卫华.提高地震资料信噪比和分辨率的一种新途径.石油地球物理勘探,1997,32(2):246~256
[11]何艳,李大卫.一种适应任意相位子波的反褶积方法.石油地球物理勘探,2008,43(3):265~269
[12]刘企英.高信噪比、高分辨率和高保真度技术的综合研究.石油地球物理勘探,1994,29(5):610~622
[13]Sen M K and Stoffa P L.Nonliner one-di mensional seismic waveform inversion using si mulated annea-ling.Geophysics,1991,56:1624~1638
[14]李大卫,田继东,成竹等.基于高阶统计的混合相位子波估计及其处理.勘探地球物理学进展,2005,28(2):117~121
[15]唐斌,尹成.基于高阶统计的非最小相位地震子波恢复.地球物理学报,2001,44(3):404~410
[2]Saggaf M Mand Enders A R.A unified frame work for the deconvolution of traces of nowhite reflectivity.Geophysics,2000,65(5):1660~1676
[3]Levy S and Fullager P K.Reconstruction of a sparse spick train froma portion of its spectrumand applica-tion to high resolution deconvolution.Geophysics,1981,46:1235~1243
[4]Walden A T and Hosken J WJ.An investigation of the spectral properties of pri mary reflection coeffi-cients.Geophysical Prospecting,1985,33:400~435
[5]Todoeschuck J P and Jense O G.Scaling geology and seismic deconvolution.Pure Appl Geophysics,1989,131:273~288
[6]Warrant T W.Si multaneous deconvolution and wave-let inversion as a global opti mizition.Geophysics,1999,64(4):1108~1115
[7]王大兴,雷晓,潘存焕等.分形反褶积方法研究及应用.石油物探,1996,35(4):93~97,1~10
[8]夏洪瑞,陈德刚,周开明.地震资料处理中随机干扰消除方法分析.石油物探,2003,42(1):93~97
[9]汪军,何振亚.基于高阶谱的信号盲分离.东南大学学报,1996,26(5):75~78
[10]王卫华.提高地震资料信噪比和分辨率的一种新途径.石油地球物理勘探,1997,32(2):246~256
[11]何艳,李大卫.一种适应任意相位子波的反褶积方法.石油地球物理勘探,2008,43(3):265~269
[12]刘企英.高信噪比、高分辨率和高保真度技术的综合研究.石油地球物理勘探,1994,29(5):610~622
[13]Sen M K and Stoffa P L.Nonliner one-di mensional seismic waveform inversion using si mulated annea-ling.Geophysics,1991,56:1624~1638
[14]李大卫,田继东,成竹等.基于高阶统计的混合相位子波估计及其处理.勘探地球物理学进展,2005,28(2):117~121
[15]唐斌,尹成.基于高阶统计的非最小相位地震子波恢复.地球物理学报,2001,44(3):404~410
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心