叠前地震资料高分辨率非线性AVA反演

发布时间：2018-10-07 20:33

【摘要】：AVA反演是以层状弹性介质散射矩阵为理论基础,利用叠前非叠加或角度部分叠加数据,提取地下岩石的弹性参数,从而实现储层预测及流体识别的目的。但在实际应用中发现依然存在着以下几个问题:(1)反演结果分辨地下地质体的能力不高,具体表现为垂向分辨率不高和不能清晰地反演出地层边界位置及地质体边缘,很难满足微小地质体的地震地质精细描述的需要;(2)在AVA反演过程中,需要对纵横波速度比做出某种近似假设,比如常将速度比取为0.5,从而线性化本质上为非线性的AVA反演问题。但是,在实际应用中经常会遇到不满足这一假设条件的问题;(3)目前的叠前地震反演都建立在地层弹性参数小跃变假设基础之上,采用层状弹性介质散射矩阵的一阶近似实现,仅适用于弱反射界面,当地层弹性参数变化较大时,基于一阶近似的反演难以得到合理的反演结果(适应弹性参数变化较大的情况)。本论文针对上述问题,在常规AVA反演方法的基础上做了深入研究。通过将隐含在地震数据中的AVF信息作为地震信号表示的过完备字典引入到反演中,充分利用地震数据中每一道的振幅随频率变化信息,特别是高频与低频信息,提高反演的分辨率。通过引入全变差正则化约束,清晰地反演出地层边界位置及地质体边缘。通过把纵横波速度比作为一个未知的模型参数,采用非线性反演方法与策略求解,解决实际工作中遇到的纵横波速度比不满足常数假设条件的问题。通过对散射矩阵进行高阶Taylor级数展开,得到散射矩阵的高阶近似表达式,突破常规叠前地震反演弹性参数小跃变假设的限制。
[Abstract]:AVA inversion is based on the theory of layered elastic media scattering matrix, using prestack non-superposition or angle partial superposition data to extract elastic parameters of underground rock, so as to achieve reservoir prediction and fluid identification. However, the following problems still exist in the practical application: (1) the ability of the inversion results to distinguish the underground geological bodies is not high, and the vertical resolution is not high and the boundary position and the edge of the geological body can not be clearly represented. It is difficult to meet the need of fine description of seismic geology of small geological bodies; (2) in the process of AVA inversion, some approximate assumptions should be made for the velocity ratio of P-S wave to shear wave, for example, the velocity ratio is usually taken as 0.5, thus linearization is essentially a nonlinear AVA inversion problem. However, the problem of not meeting this assumption is often encountered in practical application. (3) the present prestack seismic inversion is based on the assumption of small jump change of formation elastic parameters, and is realized by using the first order approximation of the scattering matrix of layered elastic media. When the local layer elastic parameters change greatly, the inversion based on the first order approximation is difficult to obtain reasonable inversion results (suitable for the large change of elastic parameters). In order to solve the above problems, this paper makes an in-depth study on the conventional AVA inversion method. By introducing the AVF information hidden in the seismic data as an overcomplete dictionary of seismic signal representation, the information of amplitude variation with frequency, especially the information of high frequency and low frequency, is fully utilized in seismic data. Improve the resolution of inversion. By introducing total variation regularization constraints, the formation boundary position and geological body edge are clearly inversed. By taking the velocity ratio of P-S wave as an unknown parameter of the model, the nonlinear inversion method and strategy are used to solve the problem that the ratio of P-S wave velocity to velocity does not satisfy the constant assumption. By using high-order Taylor series expansion of the scattering matrix, the high-order approximate expression of the scattering matrix is obtained, which breaks through the limitation of the small jump hypothesis of elastic parameters in the conventional pre-stack seismic inversion.
【学位授予单位】：中国石油大学(华东)
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.44

【参考文献】