Toeplitz方程组的近似计算

从通常所使用的求解Toeplitx方程组的Fourier变换方法出发，结合预条件共乾梯度法(Preconditioned Conjugate Gradient Method,记为PCGM)，在对Toeplitz矩阵系统中的系数矩阵作ω循环延拓后再对其进行求解。理论和实际数值计算表明，该方法化于传统的采用简单循环的普通Fourier变换方法，所得结果具有较好的精度。     

预条件共轭梯度法在地震数据重建方法中的应用

基于最小平方的Fourier地震数据重建方法最终转化为求解一个线性方程组, 其系数矩阵是Toeplitz矩阵,可以用共轭梯度法求解该线性方程组.共轭梯度法的迭代次数受系数矩阵病态程度的影响,地震数据的非规则采样程度越高,所形成的系数矩阵病态程度越高,就越难收敛和得到合理的计算结果.本文研究了基于Toeplitz矩阵的不同预条件的构造方法,以及对共轭梯度法收敛性的影响.通过预条件的使用,加快了共轭梯度法的迭代速度, 改进了共轭梯度算法的收敛性,提高了计算的效率.数值算例和实际地震数据重建试验证明了预条件共轭梯度法对计算效率有很大的提高.     

ω循环型边界条件

通过利用预条件共轭梯度法对对称正定Toeplitz 矩阵系统进行分析，重点介绍了一种新的嵌入式预条件矩阵构造方法，证明了以前的预条件矩阵构造方法大都是这种方法的特例.提出了ω循环型边界条件，并将其与普通循环型及螺旋型边界条件作了分析、比较后得到了一种新的边界条件即混合型边界条件.     

空间域位场延拓新方法研究

在空间域进行位场延拓,需要数值求解第一类Fredholm积分方程,由于所得方程组系数矩阵不是稀疏矩阵,求解该方程组需要的计算机内存大,计算量大,导致延拓算法在一般计算机上难以实现,阻碍了对空间域位场延拓方法的研究.在分析系数矩阵结构特征的基础上,本文证明了方程组系数矩阵是对称的分块Toeplitz型矩阵.利用系数矩阵的对称性和分块Toeplitz型矩阵与向量相乘的快速算法,解决了系数矩阵的存储和计算问题,使得空间域位场延拓成为可能,为研究新的位场延拓方法和分析延拓误差提供了一条新的途径.利用模型数据和实测资料,对空间域位场向上延拓、空间域积分迭代法向下延拓进行了检验,结果证实了空间域位场延拓的可行性和正确性.     

图像重建块迭代算法中松弛参数选取的研究

在图像重建中,Landweber迭代算法是图像重建算法中的重要方法.本文将针对Landweber分块迭代算法中松弛参数的选取进行研究.在重建过程中采取对投影矩阵按投影角度分块的方法,选取特定的松弛参数.通过数值实验得出结论:对于按角度分块的块迭代算法,松弛参数选取为λ乘以块矩阵与其共轭转置矩阵乘积的最大特征值分之一,当采集完全投影数据,且λ接近(1/6)～(1/7)时效果最好.另外,本文按角度分块的做法和松弛参数的选取方法对于有限角度图像重建问题也是可行的.     

预条件方程组及其应用

建立了与一般Toeplitz方程组ANX=B所对应的ω循环型预条件方程组PN［ω］Xω=B.通过采用不同的准则构造预条件矩阵,可以得到不同的预条件方程组,计算出合理的ω值.理论分析和实际计算证明了该方法所得到的近似计算结果优于普通Fourier变换方法(ω=1)的分析结果.     

模拟地震波传播的优化质量矩阵Legendre谱元法

波动方程的数值求解是地震波正反演的重要环节,而数值算法的计算精度直接关系到地震波的模拟结果和成像质量.当前,谱元法由于同时具备有限元法的网格灵活性与谱方法的高精度性已被成功应用于不同尺度模型中的地震波模拟.然而,常见的Legendre谱元法在求解地震波运动方程时采用Gauss-Lobatto-Legendre(GLL)数值积分计算质量矩阵所包含的积分项,由于GLL数值求积无法对积分项精确估计,从而造成谱元法精度损失.针对谱元法精度上的不足,本文提出一种优化算法用于提升其精度.首先构造关于GLL数值求积积分权与质量矩阵对角线元素精确值的最小二乘目标函数,然后利用共轭梯度法求解目标函数得到优化权系数,该权系数能减小质量矩阵的离散误差最终提高谱元法的计算精度.通过数值频散分析、数值算例证实了本文给出的优化算法用于提升谱元法数值模拟精度的可行性和有效性.     

时间域航空电磁2.5维非线性共轭梯度反演

对于时间域航空电磁法二维和三维反演来说,最大的困难在于有效的算法和大的计算量需求.本文利用非线性共轭梯度法实现了时间域航空电磁法2.5维反演方法,着重解决了迭代反演过程中灵敏度矩阵计算、最佳迭代步长计算、初始模型选取等问题.在正演计算中,我们采用有限元法求解拉式傅氏域中的电磁场偏微分方程,再通过逆拉氏和逆傅氏变换高精度数值算法得到时间域电磁响应.在灵敏度矩阵计算中,采用了基于拉式傅氏双变换的伴随方程法,时间消耗只需计算两次正演,从而节约了大量计算时间.对于最佳步长计算,二次插值向后追踪法能够保证反演迭代的稳定性.设计两个理论模型,检验反演算法的有效性,并讨论了选择不同初始模型对反演结果的影响.模型算例表明:非线性共轭梯度方法应用于时间域航空电磁2.5维反演中稳定可靠,反演结果能够有效地反映地下真实电性结构.当选择的初始模型电阻率值与真实背景电阻率值接近时,能得到较好的反演结果,当初始模型电阻率远大于或远小于真实背景电阻率值时反演效果就会变差.     

用于波场成像的谱法LU分解

地震波场模拟和偏移成像等有限差分隐格式算法中的重要环节，是实现亥姆霍兹算子表示矩阵H的快速求逆运算. 在螺旋边界条件下，H具有Toeplitz结构的正定厄密矩阵，其快速求逆可由谱法LU分解实现. 本文分析了谱法LU分解对提高计算速度的原理及特点，并着重讨论了在不同类型的介质模型中，采用谱法分解矩阵H时带来的数值误差、误差的分布及其对波场计算的影响. 研究结果表明，对均匀介质而言，矩阵H各列具有相同的非零元素分布，谱法LU分解的误差在吸收边界条件下，不影响波场模拟和成像计算；但对于非均匀介质模型，矩阵H各列具有不同的非零元素分布，谱法LU分解的误差随介质不均匀性程度的增大而增大，势必影响非均匀介质中波场计算. 在波场模拟和成像等有限差分隐格式算法中，采用谱法LU分解完成矩阵求逆时，必须考虑到并尽量减少该方法的误差对波场计算的影响.     

变阻尼共轭梯度算法及其性能分析

为了提高反演的分辨率和计算效率,本文在传统阻尼共轭梯度法的基础上,提出了变阻尼共轭梯度算法.由于在最小二乘线性反演过程中,多数情况下都要计算偏导数矩阵,而偏导数矩阵列向量的长度大小决定了解向量在对应分量方向上前进的快慢,变阻尼共轭梯度算法的提出正是利用了偏导数矩阵的这一特点.从理论上讲,它要优于传统的固定阻尼共轭梯度法.最后通过计算实例证明了该算法计算精度高,稳定性好,收敛速度快.因此采用变阻尼共轭梯度算法进行地球物理反演是切实可行的.     

Regional seismic slope failure probability matrices

A method for constructing seismic slope failure probability matrices is presented. The core of the method is a probabilistic sliding block model which allows for systematic incorporation of the uncertainties associated with both the ground excitation and the strength of the slope materials. The extent of damage to a slope is defined in terms of the magnitude of the earthquake-induced permanent displacement. The intensity of the ground shaking is characterized by a peak ground acceleration as well as an earthquake magnitude, and the possible scatter in the ground motion details is included through the use of an equivalent stationary motion model. After the effects of essential contributing factors are discussed, regional seismic slope failure probability matrices are presented for general applications.     

A note on classical damping matrices

An alternative damping matrix that leads to classical normal modes and that depends explicitly on a set of prescribed modal damping ratios is presented. The alternative damping matrix can be thought of as a factorized Caughey series that allows for a simple explicit solution for the coefficients of the series and thus avoids the need to solve a potentially ill‐conditioned system of algebraic equations. The relation between the proposed damping matrix and Rayleigh, Caughey and modal damping matrices is examined. Copyright © 2008 John Wiley & Sons, Ltd.     

Propagator matrices for some geothermal problems

Summary In problems of linear flow of heat in inhomogeneous media, the governing equation is a second order ordinary differential equation with variable coefficients. When transformed into a set of first order ordinary differential equations with variable coefficients, the problem becomes amenable to an elegant method of propagator matrices. In this paper the propagator matrices for some steady and unsteady heat conduction problems (including a case of heat generation by an irreversible first order reaction) having conductivity and heat generation functions as piecewise continuous, have been described.     

Dynamic stiffness matrices for two circular foundations

A systematic procedure is presented for generating dynamic stiffness matrices for two independent circular foundations on an elastic half-space medium. With the technique reported in References 1–3, the analytic solution of three-dimensional (3D) wave equations satisfying the prescribed traction due to the vibration of one circular foundation can be found. Since there are two analytic solutions for two prescribed tractions due to the vibrations of two circular foundations, the principle of superposition must be used to obtain the total solution. The interaction stresses (prescribed tractions) are assumed to be piecewise linear in the r-directions of both cylindrical co-ordinates for the two circular foundations. Then, the variational principle and the reciprocal theorem are employed to generate the dynamic stiffness matrices for the two foundations. In the process of employing the variational principle, a co-ordinate transformation matrix between two cylindrical co-ordinate systems is introduced. Some numerical results of dynamic stiffness matrices for the interaction of two identical rigid circular foundations are presented in order to show the effectiveness and efficiency of the present method, and some elaborations for its future extensions are also discussed.     

Propagator matrices in elastic wave and vibration problems

The boundary value problems most frequently encountered in studies of elastic wave propagation in stratified media can be formulated in terms of a finite number of linear, first order and ordinary differential equations with variable coefficients. Volterra (1887) has shown that solutions to such a system of equations are conveniently represented by the product integral, or propagator, of the matrix of coefficients. In this paper we summarize some of the better known properties of propagators plus numerica methods for their computation. When the dispersion relation is somem ^th order minor of the integral matrix it is possible to deal withm ^th minor propagators so that the dispersion relation is a single element of them ^th minor integral matrix. In this way one of the major sources of loss of numerical accuracy in computing the dispersion relation is avoided. Propagator equations forSH and forP-SV waves are given for both isotropic and transversely isotropic media. In addition, the second minor propagator equations forP-SV waves are given. Matrix polynomial approximations to the propagators, obtained from the method of mean coefficients by the Cayley-Hamilton theorem and the Lagrange-Sylvester, interpolation formula, are derived.     

工程结构地震破坏概率矩阵分析

本文提出了一种计算工程结构地震坡坏概率矩阵的方法，建立了地震地面运动模型和结构分析模型，对结构进行了随机地震反应分析，并获得了结构随机分应的统计量，进而采用双参数的结构破坏模型，给出了教育处结构地震破坏概率的表达式，利用此方法计算了一座按8度要求设计的钢筋混凝土框架型，给出了计算结构地震坡坏概率的表达式，利用此方法计算了一座按8度要求的钢筋混凝土框架结构的地震破坏概率矩阵，本文提出了方法可以在确定抗震设防标准和进行震害预测时采用。     

On some modifications of Thomson-Haskell matrices for love waves

Summary In Haskell's formulation the dispersion equation is expressed by means of matrices, the elements of which are real or pure imaginary [2]. In the present paper different matrices have been found, which have only real elements.Dedicated to 90th Birthday of Professor Frantiek Fiala     

Impedance matrices for circular foundation embedded in layered medium

A numerical scheme is developed in the paper for calculating torsional, vertical, horizontal, coupling and rocking impedances in frequency domain for axial-symmetric foundations embedded in layered media. In the scheme, the whole soil domain is divided into interior and exterior domains. For the exterior domain, the analytic solutions with unknown coefficients are obtained by solving three-dimensional (3D) wave equations in cylindrical coordinates satisfying homogeneous boundary conditions. For the interior domain, the analytical solutions are also obtained by solving the same 3D wave equations satisfying the homogeneous boundary conditions and the prescribed boundary conditions. The prescribed conditions are the interaction tractions at the interfaces between embedded foundation and surrounding soil. The interaction tractions are assumed to be piecewise linear. The piecewise linear tractions at the bottom surface of foundation will be decomposed into a series of Bessel functions which can be easily fitted into the general solutions of wave equations in cylindrical coordinates. After all the analytic solutions with unknown coefficients for both interior and exterior domains are found, the variational principle is employed using the continuity conditions (both displacements and stresses) at the interfaces between interior and exterior domains, interior domain and foundation, and exterior domain and foundation to find impedance functions.     

Low-rank seismic denoising with optimal rank selection for hankel matrices

Based on the fact that the Hankel matrix representing clean seismic data is low rank, low-rank approximation methods have been widely utilized for removing noise from seismic data. A common strategy for real seismic data is to perform the low-rank approximations for small local windows where the events can be approximately viewed as linear. This raises a fundamental question of selecting an optimal rank that best captures the number of events for each local window. Gavish and Donoho proposed a method to select the rank when the noise is independent and identically distributed. Gaussian matrix by analysing the statistical performance of the singular values of the Gaussian matrices. However, such statistical performance is not available for noisy Hankel matrices. In this paper, we adopt the same strategy and propose a rule that computes the number of singular values exceed the median singular value by a multiplicative factor. We suggest a multiplicative factor of 3 based on simulations which mimic the theories underlying Gavish and Donoho in the independent and identically distributed Gaussian setting. The proposed optimal rank selection rule can be incorporated into the classical low-rank approximation method and many other recently developed methods such as those by shrinking the singular values. The low-rank approximation methods with optimally selected rank rule can automatically suppress most of the noise while preserving the main features of the seismic data in each window. Experiments on both synthetic and field seismic data demonstrate the superior performance of the proposed rank selection rule for seismic data denoising.     

Extraction of real modes and physical matrices from modal testing

A technique to extract real modes from the identified complex modes is presented in this paper, which enables the normalized real mode shapes, modal masses, and full or reduced mass and stiffness matrices to be obtained. The theoretical derivation of the method is provided in detail. An 11-DOF vibration system is used to validate the algorithm, and to analyze the effects of the number of modes utilized and measurement DOFs on the extraction results. Finally, the method is used to extract real modes from both experimental modal analysis and operational modal analysis.