【作者】 邓小波；

【导师】 聂在平；

【作者基本信息】 电子科技大学 ， 电磁场与微波技术， 2005， 博士

【摘要】 本文主要研究了相位感应测井资料的非均匀介质反演方法,分别在一维和二维反演模型中利用两接收线圈的相位差Δφ和幅度比B曲线进行了反演计算。 首先,本文对电磁散射和逆散射的理论和方法进行了概述,着重介绍了高维逆散射问题的反演方法。 快速的测井资料的反演依赖于正演的计算速度,本文使用一维反演模型,提出了采用变分原理建立相位感应测井的反演方程,对相位感应测井的相位差Δφ测井资料进行快速反演,获得一维地层的电阻率分布,从而实现对相位感应测井资料的预处理。 然后,在二维非均匀介质反演模型中,着重阐述了轴对称二维非均匀介质中的数值模式匹配法(NMM),通过相位感应测井的相位差Δφ响应方程建立测井反演方程,并采用共轭梯度法(CG)进行反演计算。研究了反演方法的收敛速度,反演精度和抗噪声能力,分析了地层纵向边界位置的准确程度对地层电阻率反演精度的影响。通过相位差Δφ测井响应曲线确定出地层反演模型的初始值,对地层电阻率和地层纵向边界进行了整体反演计算。利用存在泥浆侵入的相位差Δφ响应曲线,反演计算地层的电阻率分布,对反演地层径向电阻率的能力进行了考察。 针对相位差Δφ和幅度比B测井曲线,采用共轭梯度法(CG)反演计算地层电阻率和介电常数分布。讨论了该反演方法的收敛速度,抗噪声能力,分析了相位感应测井反演过程中,地层纵向边界位置的准确程度影响电参数的反演精度,需要重视地层纵向边界位置的反演。利用相位差Δφ和幅度比B测井响应曲线,对地层电阻率、介电常数和地层纵向边界进行了整体反演计算,并由相位差Δφ和幅度比B测井响应曲线提取地层的纵向边界位置的初始值。为了提高反演计算的稳定性,对反演的步骤进行了研究。利用存在泥浆侵入的相位差Δφ和幅度比B响应曲线,反演计算地层径向的电阻率和介电常数分布,考察了反演地层径向电参数的能力。 最后,基于扩展玻恩近似的两步反演方法,并结合了多重网格技术和遗传算法(GA)对高对比度问题进行了反演。在迭代前期,采用遗传算法进行优化反演计算;在迭代后期保持格林函数不变进行迭代反演,增加了迭代反演过程中的抗噪声能力。并利用相位差Δφ测井响应曲线,采用遗传算法对二维地层模型的电阻率分布进行反演计算。更多还原

【Abstract】 The inhomogeneous inversion method on the phase induction logging has been studied in this dissertation. The inversion simulation is performed from the phase difference △φ and amplitude ratio B of the two receiver coils in one-dimensional and two-dimensional inverse modeling, respectively.Firstly, the theories and methods of electromagnetic scattering and inverse scattering are reviewed in this paper. The high-dimensional electromagnetic inverse scattering method is introduced mainly.The efficiency of the data inversion highly depends on the speed of the forward modeling. The subsurface structure is modeled as a one-dimension horizontally layered medium, the inversion equation of the phase induction logging is obtained by applying the variation principal. The algorithm is applied to pretreat phase difference △φ logging data, and the formation resistivity distribution of one-dimensional modeling is obtained.Following, in the two-dimensional inverse modeling, the numerical mode-matching (NMM) method as an efficient algorithm is used to model various multiregion vertically and cylindrically stratified inhomogeneous media. The inversion equation is obtained by applying the variation principal in a two-dimensional axisymmetrical inhomogeneous medium, which is solved by using the conjugate gradient (CG) method. The convergence speed, precision and tolerance to noise of the inversion method are discussed, it is found that the accuracy of formation boundary location influents the inversion precision of formation resistivity. The initial value of formation resistivity and boundary location are obtained from phase difference △φ logging response curve, then the formation resistivity and boundary location per bed are simultaneously reconstructed from the phase induction logging data. In order to investigate radial resolution, the formation resistivity is inverted from logging data with mud invasion.The formation resistivity and permittivity are inverted by using the conjugate gradient (CG) method from the phase difference △φ and amplitude ratio B logging data, the convergence speed and tolerance to noise of the inversion method are investigated. The accuracy of formation boundary location influents the inversion precision of formation resistivity and permittivity, the accuracy of boundary location should be paid much attention. The initial value of formation boundary location is obtained from phase difference △φ and amplitude ratio B logging response curves,更多还原