实数编码遗传算法在大地电磁测深二维反演中的应用

作为全局非线性优化的新方法之一的遗传算法,近年来已从生物工程流行到大地电磁测深资料解释中.然而,大地电磁反演问题具有不适定性,解的非唯一性.通过结合求解不适定问题的Tikhonov正则化方法,本文采用实数编码遗传算法求解大地电磁二维反演问题.此算法在构建目标函数时引入正则化的思想,利用遗传算法求解最优化问题.常规的基于局部线性化的最优化反演方法易使解陷入局部极小值,而且严重的依赖初始模型的选择.与传统线性化的迭代反演方法相比,实数编码遗传算法能够克服传统方法的不足且能获得更好的反演结果.通过对大地电磁测深理论模型进行计算,结果表明:该算法具有收敛速度快、解的精度高和避免出现早熟等优点,可用于大地电磁资料解释.     

用于速度图象重建的层析成象法

本文提出了一种用于速度图象重建的层析成象法。特别注意到地震学上的ST与医学上的CT的不尽相同之处,并处理了由此引起的困难。与速度图象重建有关的正问题的分析表明,不同频率范围的资料其分辨能力是不同的,因此,区域地震与远震资料的解释应予注意。本文有关速度图象重建的讨论不仅涉及到与成象方法有关的反演理论和方法,还给出了进行数值计算的内存量和运算量的分析。关于重建图象的质量评价,还给出了反演解的可靠性分析,为分辨特性的描述提供了一种既简便又形象直观的方案。 本文的方法与现行主要方法的比较表明:1.本文的方法允许存在速度间断面,有利于揭示不同构造区域的地壳厚度的差异;2.在常用的方法中,作为正问题假定的常速度块同反演结果解释时非常速度块之间不协调。由于我们把给定网格点的速度值的插值函数作为速度的空间函数,这一困难已被解决;3.相对于Aki等的一般方法,本文的内存量节省约一个量级、运算量减少约一半;4.相对于ART类方法本身不能给出成象结果的可靠性分析,本文的方法则具有显著的优点。 作者在分析了天然地震资料的状况之后相信,对充分利用我国地震台网的现有观测资料,本文的方法是有效的。     

一种新的地球物理反演方法——模拟原子跃迁反演法

详细研究了一般地球物理反问题的迭代优化求解过程与物理学中原子跃迁过程的对应关系,建立了反演问题中模型空间、初始模型、局部极值模型、最优化模型等与原子的态空间、定态、激发态、基态等的对应关系. 在此基础上,模拟了物理学中原子从激发态向基态跃迁的物理过程,建立了一种与原子跃迁过程相对应的非线性随机跃迁数学模型和模型解跃迁搜索准则,导出了适用于一般地球物理资料的模拟原子跃迁的非线性反演算法. 用理论测试函数对这种新的反演方法进行了数值试验,结果表明该方法具有解不依赖于初始模型、收敛速度快等优点.     

Separation method for multi-source blended seismic data

Multi-source seismic technology is an efficient seismic acquisition method that requires a group of blended seismic data to be separated into single-source seismic data for subsequent processing. The separation of blended seismic data is a linear inverse problem. According to the relationship between the shooting number and the simultaneous source number of the acquisition system, this separation of blended seismic data is divided into an easily determined or overdetermined linear inverse problem and an underdetermined linear inverse problem that is difficult to solve. For the latter, this paper presents an optimization method that imposes the sparsity constraint on wavefields to construct the object function of inversion, and the problem is solved by using the iterative thresholding method. For the most extremely underdetermined separation problem with single-shooting and multiple sources, this paper presents a method of pseudo-deblending with random noise filtering. In this method, approximate common shot gathers are received through the pseudo-deblending process, and the random noises that appear when the approximate common shot gathers are sorted into common receiver gathers are eliminated through filtering methods. The separation methods proposed in this paper are applied to three types of numerical simulation data, including pure data without noise, data with random noise, and data with linear regular noise to obtain satisfactory results. The noise suppression effects of these methods are sufficient, particularly with single-shooting blended seismic data, which verifies the effectiveness of the proposed methods.     

基于反演的衰减补偿方法(英文)

提高地震资料分辨率的一个有效途径就是衰减补偿,通过对地震波的衰减和频散效应进行校正,提高地震资料的分辨率。常规衰减补偿方法都是基于波场延拓的反Q滤波方法。本文利用Futterman衰减模型,导出了一种衰减介质中合成地震记录的计算方法,在此基础上将衰减补偿问题归结为一个Fredholm积分方程反问题,利用反演方法来实现衰减补偿。针对衰减补偿问题的不稳定性,利用Tikhonov正则化方法提高反演过程的稳定性,数值模拟资料和实际资料处理结果验证了方法的有效性。     

Non‐linear stochastic inversion of gravity data via quantum‐behaved particle swarm optimisation: application to Eurasia–Arabia collision zone (Zagros,Iran)

Potential field data such as geoid and gravity anomalies are globally available and offer valuable information about the Earth's lithosphere especially in areas where seismic data coverage is sparse. For instance, non‐linear inversion of Bouguer anomalies could be used to estimate the crustal structures including variations of the crustal density and of the depth of the crust–mantle boundary, that is, Moho. However, due to non‐linearity of this inverse problem, classical inversion methods would fail whenever there is no reliable initial model. Swarm intelligence algorithms, such as particle swarm optimisation, are a promising alternative to classical inversion methods because the quality of their solutions does not depend on the initial model; they do not use the derivatives of the objective function, hence allowing the use of L₁ norm; and finally, they are global search methods, meaning, the problem could be non‐convex. In this paper, quantum‐behaved particle swarm, a probabilistic swarm intelligence‐like algorithm, is used to solve the non‐linear gravity inverse problem. The method is first successfully tested on a realistic synthetic crustal model with a linear vertical density gradient and lateral density and depth variations at the base of crust in the presence of white Gaussian noise. Then, it is applied to the EIGEN 6c4, a combined global gravity model, to estimate the depth to the base of the crust and the mean density contrast between the crust and the upper‐mantle lithosphere in the Eurasia–Arabia continental collision zone along a 400 km profile crossing the Zagros Mountains (Iran). The results agree well with previously published works including both seismic and potential field studies.     

Optimization of Cell Parameterizations for Tomographic Inverse Problems

—?We develop algorithms for the construction of irregular cell (block) models for parameterization of tomographic inverse problems. The forward problem is defined on a regular basic grid of non-overlapping cells. The basic cells are used as building blocks for construction of non-overlapping irregular cells. The construction algorithms are not computationally intensive and not particularly complex, and, in general, allow for grid optimization where cell size is determined from scalar functions, e.g., measures of model sampling or a priori estimates of model resolution. The link between a particular cell j in the regular basic grid and its host cell k in the irregular grid is provided by a pointer array which implicitly defines the irregular cell model. The complex geometrical aspects of irregular cell models are not needed in the forward or in the inverse problem. The matrix system of tomographic equations is computed once on the regular basic cell model. After grid construction, the basic matrix equation is mapped using the pointer array on a new matrix equation in which the model vector relates directly to cells in the irregular model. Next, the mapped system can be solved on the irregular grid. This approach avoids forward computation on the complex geometry of irregular grids. Generally, grid optimization can aim at reducing the number of model parameters in volumes poorly sampled by the data while elsewhere retaining the power to resolve the smallest scales warranted by the data. Unnecessary overparameterization of the model space can be avoided and grid construction can aim at improving the conditioning of the inverse problem. We present simple theory and optimization algorithms in the context of seismic tomography and apply the methods to Rayleigh-wave group velocity inversion and global travel-time tomography.     

Inversion of electromagnetic data: An overview of new techniques

This paper explores some of the newer techniques for acquiring and inverting electromagnetic data. Attention is confined primarily to the 2d magnetotelluric (MT) problem but the inverse methods are applicable to all areas of EM induction. The basis of the EMAP technique of Bostick is presented along with examples to illustrate the efficacy of that method in structural imaging and in overcoming the deleterious effects of near-surface distortions of the electric field. Reflectivity imaging methods and the application of seismic migration techniques to EM problems are also explored as imaging tools. Two new approaches to the solution of the inverse problem are presented. The AIM (Approximate Inverse Mapping) inversion of Oldenburg and Ellis uses a new way to estimate a perturbation in an iterative solution which does not involve linearization of the equations. The RRI (Rapid Relaxation Inverse) of Smith and Booker shows how approximate Fréchet derivatives and sequences of 1d inversions can be used to develop a practical inversion algorithm. The overview is structured to provide insight about the latest inversion techniques and also to touch upon most areas of the inverse problem that must be considered to carry out a practical inversion. These include model parameterization, methods of calculating first order sensitivities, and methods for setting up a linearized inversion.     

Application of Surface-Wave Methods for Seismic Site Characterization

Surface-wave dispersion analysis is widely used in geophysics to infer a shear wave velocity model of the subsoil for a wide variety of applications. A shear-wave velocity model is obtained from the solution of an inverse problem based on the surface wave dispersive propagation in vertically heterogeneous media. The analysis can be based either on active source measurements or on seismic noise recordings. This paper discusses the most typical choices for collection and interpretation of experimental data, providing a state of the art on the different steps involved in surface wave surveys. In particular, the different strategies for processing experimental data and to solve the inverse problem are presented, along with their advantages and disadvantages. Also, some issues related to the characteristics of passive surface wave data and their use in H/V spectral ratio technique are discussed as additional information to be used independently or in conjunction with dispersion analysis. Finally, some recommendations for the use of surface wave methods are presented, while also outlining future trends in the research of this topic.     

大地电磁的多尺度反演

对于迭代方式的参数化反演方法,如何使反演结果稳定地收敛到整体极小仍是目前大地电磁（MT）反演中急需解决的问题．本文利用小波变换理论中的多尺度分析方法将大地电磁反问题分解为依赖于尺度变量的反问题序列,然后按尺度从大到小的次序依次求解,求解过程中前一个尺度反问题的解作为下一个尺度反问题的初始模型,直到来出对应于尺度为0的原反问题的解为止．该方法称为多尺度反演方法．数值试验和实际资料的反演结果表明,该方法可有效改善传统广义逆反演方法易陷入局部极小的弊端．     

Non‐linear prestack seismic inversion with global optimization using an edge‐preserving smoothing filter

Estimating elastic parameters from prestack seismic data remains a subject of interest for the exploration and development of hydrocarbon reservoirs. In geophysical inverse problems, data and models are in general non‐linearly related. Linearized inversion methods often have the disadvantage of strong dependence on the initial model. When the initial model is far from the global minimum, inversion iteration is likely to converge to the local minimum. This problem can be avoided by using global optimization methods. In this paper, we implemented and tested a prestack seismic inversion scheme based on a quantum‐behaved particle swarm optimization (QPSO) algorithm aided by an edge‐preserving smoothing ( EPS) operator. We applied the algorithm to estimate elastic parameters from prestack seismic data. Its performance on both synthetic data and real seismic data indicates that QPSO optimization with the EPS operator yields an accurate solution.     

Model Fusion and Joint Inversion

Inverse problems are inherently non-unique, and regularization is needed to obtain stable and reasonable solutions. The regularization adds information to the problem and determines which solution, out of the infinitely many, is obtained. In this paper, we review and discuss the case when a priori information exists in the form of either known structure or in the form of another inverse problem for a different property. The challenge is to include such information in the inversion process. To use existing known structure, we review the concept of model fusion, where we build a regularization functional that fuses the inverted model to a known one. The fusion is achieved by four different techniques. Joint inversion of two data sets is achieved by using iterative data fusion. The paper discusses four different methods for joint inversion. We discuss the use of correspondence maps or the petrophysics of the rocks, as well as structure. In particular, we suggest to further stabilize the well-known gradient cross product and suggest a new technique, Joint Total Variation, to solve the problem. The Joint Total Variation is a convex functional for joint inversion and, as such, has favorable optimization properties. We experiment with the techniques on the DC resistivity problem and the borehole tomography and show how model fusion and joint inversion can significantly improve over existing techniques.     

Pg回折波的上地壳三维层析成象

提出利用人工地震Ｐｇ回折波走时重建区域上地壳三维速度分布的方法。正问题利用经典最小二乘迭代求解垂向速度梯度模型的参数值及相应走时残差。反问题采用泛函空间的最小二乘准则来实现。对速度不需要预先进行参数化处理，可以计算反演区内任意点的速度值。数值计算结果表明了该方法的可靠性。处理了滦县地震区人工地震Ｐｇ走时资料，获得了该区上地壳顶部的三维速度分布。     

Velocity analysis based on data correlation

Several methods exist to automatically obtain a velocity model from seismic data via optimization. Migration velocity analysis relies on an imaging condition and seeks the velocity model that optimally focuses the migrated image. This approach has been proven to be very successful. However, most migration methods use simplified physics to make them computationally feasible and herein lies the restriction of migration velocity analysis. Waveform inversion methods use the full wave equation to model the observed data and more complicated physics can be incorporated. Unfortunately, due to the band‐limited nature of the data, the resulting inverse problem is highly nonlinear. Simply fitting the data in a least‐squares sense by using a gradient‐based optimization method is sometimes problematic. In this paper, we propose a novel method that measures the amount of focusing in the data domain rather than the image domain. As a first test of the method, we include some examples for 1D velocity models and the convolutional model.     

Estimation of depth and magnetization of magnetic sources from magnetic data: The linearized continuous inverse problem for 21/2D structures

The estimation of the depth to the top and bottom of a magnetic source from magnetic data defines a nonlinear inverse problem, while the evaluation of the distribution of magnetization determines a linear inverse problem. In this paper, these interpretation problems are resolved in the continuous case of 2_1/2D magnetized bodies with lateral magnetization variations. A formulation of the magnetic problem accounting for different directions of remanent and total magnetization vectors and including a more general definition of apparent susceptibility is presented. Differences between 2D and 2_1/2D formulations are stressed, as regards the anomaly amplitude, shape and zero-level.In order to utilize well-known continuous linear inverse methods, Fréchet derivatives of the data functionals with respect to the depth of the source top and bottom, are analytically described. Thus, using the spectral expansion inverse method (Parker, 1977) and linearizing the problem at several steps of an iterative process, the source depth is obtained within a few iterations, although the starting model is distant from the final solution. The interpretation of an anomaly in the Italian region shows the usefulness of the method.     

A maximum likelihood approach to nonlinear inversion under constraints

Nonuniqueness in geophysical inverse problems is naturally resolved by incorporating prior information about unknown models into observed data. In practical estimation procedures, the prior information must be quantitatively expressed. We represent the prior information in the same form as observational equations, nonlinear equations with random errors in general, and treat as data. Then we may define a posterior probability density function of model parameters for given observed data and prior data, and use the maximum likelihood criterion to solve the problem. Supposing Gaussian errors both in observed data and prior data, we obtain a simple algorithm for iterative search to find the maximum likelihood estimates. We also obtain an asymptotic expression of covariance for estimation errors, which gives a good approximation to exact covariance when the estimated model is linearly close to a true model. We demonstrate that our approach is a general extension of various inverse methods dealing with Gaussian data. By way of example, we apply the new approach to a problem of inferring the final rupture state of the 1943 Tottori earthquake (M = 7.4) from coseismic geodetic data. The example shows that the use of sufficient prior information effectively suppresses both the nonuniqueness and the nonlinearity of the problem.     

An approximation method for solving the inverse MTS problem with the use of neural networks

The work develops the approximation approach to solving the inverse MTS problem with the use of neural networks. The inverse problem is considered in model classes of parametrized geoelectric structures, whose electric conductivity is controlled by a few hundreds of macroparameters (N ∼ 300). An approximate inverse operator of the problem is constructed for each model class as a neural network, whose coefficients are determined in the process of training on a representative sample of standard examples of forward problem solutions. The problem of determination of the model class of geolectric structures corresponding to the presented input MT data is solved with the use of the neural network classifier constructed for the available set of model classes of structures. Regularizing factors and errors of the neural network method are analyzed. The operation of the algorithm is illustrated by examples of the 2-D inversion of synthetic MT data.     

基于稀疏反演理论的自动叠加速度反演方法

叠加速度分析技术是常规地震资料处理中的重要环节,也是经典的时间域速度建模方法.叠加速度分析技术主要包括速度谱计算和拾取两个步骤.至今为止,多数研究工作通过提高速度谱的分辨率以及抗噪声能力,获得高质量的速度谱从而有利于拾取.本文的目标是将叠加速度分析技术转为一个全自动化的处理流程.从参数估计的角度出发,将叠加速度估计转化为稀疏反演框架下的模型参数估计问题,并通过稀疏反演算法自动反演叠加速度,进而提高叠加速度建模的效率.为实现这一目标,首先给出了正问题的定义,即层状介质中CMP道集的预测模型,利用叠加速度、垂向双程走时(t_0)以及反射子波以及CMP道集时距关系(如双曲时距关系)可以预测CMP道集.接着,速度分析反问题可以描述为已知观测的CMP道集,估计模型参数(叠加速度及t_0时间等).利用模型参数的稀疏性作为约束条件并用L_0范数作为模型稀疏性的度量准则,叠加速度分析可以转化为L_0范数约束下的稀疏反演问题.本文提出了一种基于预测校正思想的匹配追踪算法求解上述反问题,实现了自动叠加速度建模并为后续的高精度速度反演方法提供较好的初始模型.理论和实际资料的测试结果证明了本文方法的有效性.     

COMPARISON OF FIVE LEAST-SQUARES INVERSION TECHNIQUES IN RESISTIVITY SOUNDING*

A brief history of the development of the inverse problem in resistivity sounding is presented with the development of the equations governing the least-squares inverse. Five algorithms for finding the minimum of the least-square problem are described and their speed of convergence is compared on data from two planar earth models. Of the five algorithms studied, the ridge-regression algorithm required the fewest numbers of forward problem evaluations to reach a desired minimum. Solution space statistics, including (1) parameter-standard errors, (2) parameter correlation coefficients, (3) model parameter eigenvectors, and (4) data eigenvectors are discussed. The type of weighting applied to the data affects these statistical parameters. Weighting the data by taking log₁₀ of the observed and calculated values is comparable to weighting by the inverse of a constant data error. The most reliable parameter standard errors are obtained by weighting by the inverse of observed data errors. All other solution statistics, such as dataparameter eigenvector pairs, have more physical significance when inverse data error weighting is used.