三维高密点地震信号数字组合技术研究
|
摘要
目前寻找石油或天然气的主要方法是人工地震勘探,将野外采集的地震信号进行处理后可以直观地显示地下的地质结构,进而推断油气田的位置。高密点地震勘探技术是新一代的人工地震勘探技术,该技术在野外进行单点采集,在室内进行组合。目前国内主要采用算术平均组合算法对地震数据进行组合,其组合目的是压制干扰信号,但是采用该算法压制随机干扰的同时会造成高频信号的损失,降低地震资料的分辨率。本文主要研究适用于高密点地震数据的组合算法以及组合软件设计。
本文论述了高密点地震勘探的组合原理与方法,研究了一种基于主成分分析(Principal Component Analysis, PCA)的数字加权组合算法。数字加权算法使用模式识别中的PCA算法对组合内地震波形进行一致性分析,根据一致性分析结果确定组合内各地震道的权系数,畸变大的地震道给予较小的权系数,畸变小的地震道给予较大的权系数,并使用权系数对地震道进行加权叠加处理。经实验验证,该算法不但可以有效地压制干扰信号,还能提高高频地震信号的保真度以及地震资料的分辨率。最后根据相应算法和组合原理,设计适用于三维高密点地震数据的组合软件,显示三维地震数据的二维剖面以及三维可视化图像,使用基于GPU的光线投射算法实现三维地震数据的可视化,并对结果进行对比分析。
As is known to us all, the main way to find oil or natural gas is artifical seismic exploration. The geologic structure underground will be directly shown after processing the seismic signal collecting in the field, and the location of the oil can be speculated. A new technique is the high-density seismic prospecting technology, which conducts point-receiver acquisition with digital geophones outdoors and carries out digital array-forming processing indoors. The popular method of digital array-forming at home is arithmetic mean algorithm, and its goal is to inhibit the undesired signals. But this may cause the loss of high frequency signals, and decrease the seismic resolution. In this paper we mainly discuss the forming algorithms fit for high-density seismic data and the corresponding software designing.
Firstly, we describe the forming theories and methods of high-density seismic exploration. Then a digital array-forming weighted algorithm based on PCA (Principal Component) Analysis is proposed. The digital weighted algorithm makes consistent analysis on the forming wave by using PCA in the pattern recognition, and each weight can be made based on the result of the consistent analysis. If the distortion of seismic waves is soft, smaller weights will be given. Otherwise, lager weights will be made. Then, superimpose the seismic traces using weights calculated previously. Experiments show that this algorithm can not only control the undesired signals effectively, but aslo increase the fidelity of the high frequency signals and the seismic resolution. Finally based on this algorithm and the forming theories, a software is designed for 3D high-density seismic data to represent the 2D profile and the 3D visiable image of the seismic data. GPU-based ray-casting algorithm for visualization of 3D seismic data has been achieved in this thesis. Also conclusions will be made at last.
本文论述了高密点地震勘探的组合原理与方法,研究了一种基于主成分分析(Principal Component Analysis, PCA)的数字加权组合算法。数字加权算法使用模式识别中的PCA算法对组合内地震波形进行一致性分析,根据一致性分析结果确定组合内各地震道的权系数,畸变大的地震道给予较小的权系数,畸变小的地震道给予较大的权系数,并使用权系数对地震道进行加权叠加处理。经实验验证,该算法不但可以有效地压制干扰信号,还能提高高频地震信号的保真度以及地震资料的分辨率。最后根据相应算法和组合原理,设计适用于三维高密点地震数据的组合软件,显示三维地震数据的二维剖面以及三维可视化图像,使用基于GPU的光线投射算法实现三维地震数据的可视化,并对结果进行对比分析。
As is known to us all, the main way to find oil or natural gas is artifical seismic exploration. The geologic structure underground will be directly shown after processing the seismic signal collecting in the field, and the location of the oil can be speculated. A new technique is the high-density seismic prospecting technology, which conducts point-receiver acquisition with digital geophones outdoors and carries out digital array-forming processing indoors. The popular method of digital array-forming at home is arithmetic mean algorithm, and its goal is to inhibit the undesired signals. But this may cause the loss of high frequency signals, and decrease the seismic resolution. In this paper we mainly discuss the forming algorithms fit for high-density seismic data and the corresponding software designing.
Firstly, we describe the forming theories and methods of high-density seismic exploration. Then a digital array-forming weighted algorithm based on PCA (Principal Component) Analysis is proposed. The digital weighted algorithm makes consistent analysis on the forming wave by using PCA in the pattern recognition, and each weight can be made based on the result of the consistent analysis. If the distortion of seismic waves is soft, smaller weights will be given. Otherwise, lager weights will be made. Then, superimpose the seismic traces using weights calculated previously. Experiments show that this algorithm can not only control the undesired signals effectively, but aslo increase the fidelity of the high frequency signals and the seismic resolution. Finally based on this algorithm and the forming theories, a software is designed for 3D high-density seismic data to represent the 2D profile and the 3D visiable image of the seismic data. GPU-based ray-casting algorithm for visualization of 3D seismic data has been achieved in this thesis. Also conclusions will be made at last.
引文
[1]马在田.三维地震勘探方法[M].第1版.北京：石油工业出版社,1989.
[2]张爱敏.采区高分辨率三维地震勘探[M].第1版.徐州：中国矿业大学出版社,1997.
[3]陆基孟.地震勘探原理[M].第1版.东营：山东大学出版社,1993.
[4]曲寿利,袁谋,黄河福,杨力.中石化集团东部油田勘探技术发展战略研究[J].石油化工管理干部学院学报,2004,6(2)：1-6.
[5]Gerrit Blacquiere, Leo Ongkiehong. Single sensor recording:anti-alias filtering, perturbations and dynamic range[C]. Expanded Abstract of 70th Annual International SEG Meeting,2000,19:33～36.
[6]曲寿利.高密度三维地震技术一老油区二次勘探的关键技术之一[J].石油物探,2006,43(6)：557-562.
[7]Newman, P., Mahoney, J.T.. Patterns-With a pinch of salt[J]. Geophysics. Prospecting. 1973,21(2):197～219.
[8]Ongkiehong, L., Huizer, W.. Dynamic range of the seismic system[J]. First Break, 1987,5(12):435～439.
[9]Ongkiehong, L.. A changing philosophy in seismic data acquisition[J]. First Break. 1988,6(9):281～284.
[10]Patrick Burger, Robert Garotta, Pierre-Yves Granger. Improving resolution and seismic quality assurance through field pre-processing[C] Expanded Abstract of 67th Annual International SEG Meeting,1997:100～103.
[11]Phil Christie, David Nichols, et.al.. Rising the standards of seismic data quality[J]. Oilfield Review.2001,13(2):16～31.
[12]Malcolm Lansley, Sercel, Mike Laurin et.al.. Land 3D:groups or single sensors? cables or radio? geophysical and operational considerations[C]. SEG/San Antonio 2007 Annual Meeting.6～10.
[13]Malcolm, Lansley, Mike Laurin, ShukiRonen. Modern land recording systems:How do they weigh up?[J]. The Leading Edge,2008,888～894.
[14]何樵登.地震勘探原理和方法[M].第1版.地质出版社北京.1986.
[15]陆基孟,王永刚.地震勘探原理[M].第3版.东营：中国石油大学出版社,2011.
[16]R.E.谢里夫.勘探地震学[M].第2版.北京：石油工业出版社,1999.
[17]李庆忠,魏继东.高密度地震采集中组合效应对高频截止频率的影响[J].石油地 球物理勘探,2007,42(4)：363-369.
[18]曹务祥.单道接收地震资料的室内组合方法[J].石油地球物理勘探,2006,41(6)：615～618.
[19]孙即祥.现代模式识别[M].第2版.北京：国防科技大学出版社,2001.
[20]蔡元龙.模式识别[M].西安：西安电子科技大学出版社,1986.
[21]马奎斯德萨著；吴逸飞译.模式识别一原理、方法及应用[M].北京：清华大学出版社,2002.
[22]K.Pearson. Mathematical contributions to the theory of evolution-Ⅲ. Regression, heridity and panmixia. Philosophical Transaction of the Royal Society of London, Series A,187:253-318,1896.
[23]方开泰.实用多元统计分析[M].上海：华东师范大学出版社,1992.
[24]张学工.模式识别[M].第3版.北京：清华大学出版社,2010
[25]JacksonJ E. A User's Guide to Principal Components [M]. NewYork John Wiley, 1991.
[26]何汉林.数值分析[M].第2版.北京：科学出版社,2011.
[27]Timothy Sauer著,吴兆金,王国英等译.数值分析[M].第1版.北京：人民邮电出版社,2010
[28]J.H.威尔金森.代数特征值问题[M].北京：科学出版社,2001.
[29]张润楚.多元统计分析[M].第1版.北京：科学出版社,2006.
[30]John Goerzen著,魏永明译Linux编程宝典[M].北京：电子工业出版社,2000.
[31]Kurt Wall等著,张辉译.GNU/Linux编程指南[M].第2版.北京：清华大学出版社,2002.
[32]Jasmin Blanchette, Mark Summerfield著,闫锋欣译C++GUI Qt 4编程[M].第2版.北京：电子工业出版社,2008.
[33]蔡志明.精通Qt4编程[M].北京：电子工业出版社,2008.
[34]马仁安,杨静宇,王洪元.适用于地震数据可视化的体绘制模型与算法.计算机工程与应用[J],2003,22：226～228
[35]马仁安.基于微机的三维地震数据可视化技术研究.南京理工大学博士学位论文,2004.
[36]Donald Hearn, M. Pauline Baker著,蔡士杰,宋继强,蔡敏译.计算机图形学[M].第3版.北京：电子工业出版社,2005.
[37]唐泽圣.三维数据场可视化.第1版.北京：清华大学出版社,1999.
[38]Levoy M. Display of Surface from Volume Data[J]. IEEE Computer Graphics and Application,1988,8(3):29～37.
[39]Westover L. Footprint Evaluation for Volume Rendering. Computer Graphics,1990, 24(4):367～376.
[40]Lacroute P, Levoy M. Fast Volume Rendering Using a Shear-Warp Factorization of the Viewing Transformation. Computer Graphics,1994,28(5):451～457.
[41]张浩,李利军,林岚.GPU的通用计算研究[J].计算机与数字工程,2005,33(12)：60～62.
[42]储璟骏,杨新,高艳.使用GPU编程的光线投射体绘制算法.计算机辅助设计与图形学学报,2007,19(2)：257～262.
[43]储璟骏.基于GPU的直接体绘制技术[D].上海交通大学硕士学位论文,2007.
[44]许寒,刘希顺,王博亮.光线投射算法中重采样的设计与实现[J].中国图像图形学报,2003,8(A)(12)：1427～1431.
[45]解梅,景孝凯.关于光影投射算法的OpenGL高效实现[J].电子科技大学学报,2008,37(4)：610～613.
[2]张爱敏.采区高分辨率三维地震勘探[M].第1版.徐州：中国矿业大学出版社,1997.
[3]陆基孟.地震勘探原理[M].第1版.东营：山东大学出版社,1993.
[4]曲寿利,袁谋,黄河福,杨力.中石化集团东部油田勘探技术发展战略研究[J].石油化工管理干部学院学报,2004,6(2)：1-6.
[5]Gerrit Blacquiere, Leo Ongkiehong. Single sensor recording:anti-alias filtering, perturbations and dynamic range[C]. Expanded Abstract of 70th Annual International SEG Meeting,2000,19:33～36.
[6]曲寿利.高密度三维地震技术一老油区二次勘探的关键技术之一[J].石油物探,2006,43(6)：557-562.
[7]Newman, P., Mahoney, J.T.. Patterns-With a pinch of salt[J]. Geophysics. Prospecting. 1973,21(2):197～219.
[8]Ongkiehong, L., Huizer, W.. Dynamic range of the seismic system[J]. First Break, 1987,5(12):435～439.
[9]Ongkiehong, L.. A changing philosophy in seismic data acquisition[J]. First Break. 1988,6(9):281～284.
[10]Patrick Burger, Robert Garotta, Pierre-Yves Granger. Improving resolution and seismic quality assurance through field pre-processing[C] Expanded Abstract of 67th Annual International SEG Meeting,1997:100～103.
[11]Phil Christie, David Nichols, et.al.. Rising the standards of seismic data quality[J]. Oilfield Review.2001,13(2):16～31.
[12]Malcolm Lansley, Sercel, Mike Laurin et.al.. Land 3D:groups or single sensors? cables or radio? geophysical and operational considerations[C]. SEG/San Antonio 2007 Annual Meeting.6～10.
[13]Malcolm, Lansley, Mike Laurin, ShukiRonen. Modern land recording systems:How do they weigh up?[J]. The Leading Edge,2008,888～894.
[14]何樵登.地震勘探原理和方法[M].第1版.地质出版社北京.1986.
[15]陆基孟,王永刚.地震勘探原理[M].第3版.东营：中国石油大学出版社,2011.
[16]R.E.谢里夫.勘探地震学[M].第2版.北京：石油工业出版社,1999.
[17]李庆忠,魏继东.高密度地震采集中组合效应对高频截止频率的影响[J].石油地 球物理勘探,2007,42(4)：363-369.
[18]曹务祥.单道接收地震资料的室内组合方法[J].石油地球物理勘探,2006,41(6)：615～618.
[19]孙即祥.现代模式识别[M].第2版.北京：国防科技大学出版社,2001.
[20]蔡元龙.模式识别[M].西安：西安电子科技大学出版社,1986.
[21]马奎斯德萨著；吴逸飞译.模式识别一原理、方法及应用[M].北京：清华大学出版社,2002.
[22]K.Pearson. Mathematical contributions to the theory of evolution-Ⅲ. Regression, heridity and panmixia. Philosophical Transaction of the Royal Society of London, Series A,187:253-318,1896.
[23]方开泰.实用多元统计分析[M].上海：华东师范大学出版社,1992.
[24]张学工.模式识别[M].第3版.北京：清华大学出版社,2010
[25]JacksonJ E. A User's Guide to Principal Components [M]. NewYork John Wiley, 1991.
[26]何汉林.数值分析[M].第2版.北京：科学出版社,2011.
[27]Timothy Sauer著,吴兆金,王国英等译.数值分析[M].第1版.北京：人民邮电出版社,2010
[28]J.H.威尔金森.代数特征值问题[M].北京：科学出版社,2001.
[29]张润楚.多元统计分析[M].第1版.北京：科学出版社,2006.
[30]John Goerzen著,魏永明译Linux编程宝典[M].北京：电子工业出版社,2000.
[31]Kurt Wall等著,张辉译.GNU/Linux编程指南[M].第2版.北京：清华大学出版社,2002.
[32]Jasmin Blanchette, Mark Summerfield著,闫锋欣译C++GUI Qt 4编程[M].第2版.北京：电子工业出版社,2008.
[33]蔡志明.精通Qt4编程[M].北京：电子工业出版社,2008.
[34]马仁安,杨静宇,王洪元.适用于地震数据可视化的体绘制模型与算法.计算机工程与应用[J],2003,22：226～228
[35]马仁安.基于微机的三维地震数据可视化技术研究.南京理工大学博士学位论文,2004.
[36]Donald Hearn, M. Pauline Baker著,蔡士杰,宋继强,蔡敏译.计算机图形学[M].第3版.北京：电子工业出版社,2005.
[37]唐泽圣.三维数据场可视化.第1版.北京：清华大学出版社,1999.
[38]Levoy M. Display of Surface from Volume Data[J]. IEEE Computer Graphics and Application,1988,8(3):29～37.
[39]Westover L. Footprint Evaluation for Volume Rendering. Computer Graphics,1990, 24(4):367～376.
[40]Lacroute P, Levoy M. Fast Volume Rendering Using a Shear-Warp Factorization of the Viewing Transformation. Computer Graphics,1994,28(5):451～457.
[41]张浩,李利军,林岚.GPU的通用计算研究[J].计算机与数字工程,2005,33(12)：60～62.
[42]储璟骏,杨新,高艳.使用GPU编程的光线投射体绘制算法.计算机辅助设计与图形学学报,2007,19(2)：257～262.
[43]储璟骏.基于GPU的直接体绘制技术[D].上海交通大学硕士学位论文,2007.
[44]许寒,刘希顺,王博亮.光线投射算法中重采样的设计与实现[J].中国图像图形学报,2003,8(A)(12)：1427～1431.
[45]解梅,景孝凯.关于光影投射算法的OpenGL高效实现[J].电子科技大学学报,2008,37(4)：610～613.