《全频带地震信息挖掘与油气识别》由会员分享，可在线阅读，更多相关《全频带地震信息挖掘与油气识别（87页珍藏版）》请在金锄头文库上搜索。
1、成都理工大学硕士学位论文全频带地震信息挖掘与油气识别姓名：边立恩申请学位级别：硕士专业：地球探测与信息技术指导教师：贺振华;黄德济20090501摘 要 I全频带地震信息挖掘与油气识别全频带地震信息挖掘与油气识别 作者简介：边立恩，男，1982年09月生，师从成都理工大学贺振华教授、黄德济教授，2009年06月毕业于成都理工大学地球探测与信息技术专业，获得工学硕士学位。 摘 要 摘 要 油气勘探目标的复杂化和隐蔽性对勘探精度和油气识别方法提出了更高的要求， 因此开展新的更为有效的油气识别方法研究就成为摆在地球物理工作者面前的一个亟待解决的难题。经过几十年的研究和发展，人们相继提出了一系列的油气
2、识别方法，在实际生产中取得了不少成功的勘探实例。但由于实际地下地质构造和岩性的复杂性，每种方法都不是放之四海皆有效的，都有一定的局限性，这也是使该项研究能够不断深入地开展下去的一个主要原因。 鉴于常规地震资料处理和解释中，人们往往只利用了通频带以内的地震信息，对截频和限频以外的信息很少利用的不足，本文通过自主编程从全频带的角度探讨了用于油气识别的各种方法。 其中， 对低频地震信息的利用分为两个方面：一是低频信息直接用于烃类检测， 其理论依据是地震波穿过含油气地层时低频信息衰减相对较弱，这就使得低频信息中保留了较为丰富的反映地下岩性的信息；二是基于低频伴影的油气识别方法，该方法的应用主要得益于含
3、流体（油、气、水）介质对地震波的吸收衰减特性：选频吸收和相位延迟，使得地震波场在含流体层下方出现了低频高能量区，即低频伴影。本文对高频地震信息的利用除了直接利用高频信息的高分辨率特性对含油气地质体进行精细的构造刻画外， 主要是利用了其高频衰减特性进行油气识别， 因为高频成分在穿过含油气层发生衰减的同时其吸收特性显著增大， 这为我们利用高频信息的吸收衰减特性进行油气预测提供了依据和思路。 本文首先从常规时频分析方法入手，在深入研究并探讨了短时傅里叶变换、小波变换、Wigner-Ville分布、S变换等各种方法在信号分析中的优缺点的基础上，提出了一种新的高精度时频分析方法-自适应广义S变换，该方法
4、的实现是通过在S变换中引入新的参数来根据不同信号成分本身的特点自适应地调节窗口宽度，进而达到最优的时频分布。为后续章节中基于时频分析的各种地震切片的提取、共频率剖面的抽取及吸收、衰减属性剖面的求取提供了一种更为有效的工具和手段。然后在对低频伴影的基本概念、产生原因及国外实验室物理模拟成果认识和分析的基础上，详细探讨了基于低频伴影的油气识别方法与原理，推导了二成都理工大学硕士学位论文 II维粘滞弥散型波动方程的解， 分析和归纳了含流体介质的地震波场随粘滞系数和弥散系数变化的特征和规律，并对合成理论模型进行了数值模拟，取得了与实验室物理模拟类似的结果， 从理论的角度证实了含流体层下方存在的地震低频
5、伴影现象， 也为基于低频伴影的油气识别方法的可行性提供了依据并展示了方法的应用前景。接着，探讨了基于高频衰减特性的油气识别方法，文中主要采用了能量吸收分析技术（EEA）和频率衰减梯度两种方法分析地震波的高频衰减特性，并直接用于油气检测，理论模型试算揭示了方法的可行性和有效性。最后一章给出了全频带地震信息在油气识别中的应用实例， 实际应用结果取得了与钻井资料的高吻合度，效果是显著的。 理论模型试算并结合实际资料的应用效果表明， 本文提出的方法和研究思路是可行的，对于复杂地区的油气识别和储层预测研究具有一定的指导意义。 关键词：时频分析 粘滞弥散型波动方程 低频伴影 高频衰减 油气识别 Abstr
6、act IIIWide-band seismic data mining and identification of oil and gas Introduction of the author： Bianlien， male， was born in September， 1982， whose tutor was Professor He Zhenhua.He graduated from Chengdu University of Technology in Earth Exploration and Information Technology major and was grante
7、d the Master Degree in June, 2009. Abstract Abstract The complexity and concealment feature of oil and gas exploration goals call for higher exploration precision and better identification methods of oil and gas,so it has become a urgent problem for geophysical workers to develop new and more effect
8、ive methods of oil and gas identification. After decades of research and development, people have proposed a number of methods of oil and gas identification which have made many successful exploration examples in practical application. However, each method is not size-fits-all effective and has some
9、 limitations because of the complexity of geological structure and lithology underground which is one of the reasons that make the research carried out step by step. In view of the shortcomings that people tend to use the seismic information within pass band and hardly use the information out of cut
10、-off and roll-off during conventional seismic data processing and interpretation, this paper explores a variety of methods to identify oil and gas from perspective of full-frequency band by self-programming. Among them, the use of low-frequency seismic information is divided into two sides: first, l
11、ow frequency information can be directly used for hydrocarbon detection whose theoretical foundation is that when seismic waves go through oil and gas formation, low-frequency information has weaker attenuation which makes sure that low frequency information retain richer lithology-related informati
12、on; second, identification method of oil and gas based on low frequency shadow which is mainly due to the attenuation characteristics when seismic waves propagate in medium filled with fluid: frequency-selective absorption and phase delay, which result in that low-frequency high-energy zone of seism
13、ic wave field appears in the bottom of layer filled with fluid,that is low frequency shadow. In this paper, high frequency information is mainly used to identify oil and gas by its high frequency attenuation characteristics besides the direct use in subtle structure characterization of oil and gas g
14、eological body by its high-resolution characteristic, because while high 成都理工大学硕士学位论文 IVfrequency components go through oil and gas layer and attenuate,its absorption coefficient significantly increase,which provides basis and ideas for us to predict oil and gas by absorption and attenuation charact
15、eristics of high frequency information. In this paper,firstly,we begin from conventional time-frequency analysis methods, propose a new high-precision time-frequency analysis method-Adaptive Generalized S transform based on deep research and exploration the advantages and disadvantages ofShort-timeF
16、ouriertransform,wavelettransform,Wigner-Ville distribution,S transform and other signal analysis methods, this method is achieved by adding a new parameter in S transform to adjust the width of window adaptively according to characteristics of different components,and finally optimizes the time-frequency distribution,which provides a more effective tool and means for us to extract various seismic slice,single frequency section and attenuation propertie section based on time-frequency ana
