Computation of eigenpair partial derivatives by Rayleigh–Ritz procedure

Based on Rayleigh–Ritz procedure, a new method is proposed for a few eigenpair partial derivatives of large matrices. This method simultaneously computes the approximate eigenpairs and their partial derivatives. The linear systems of equations that are solved for eigenvector partial derivatives are greatly reduced from the original matrix size. And the left eigenvectors are not required. Moreover, errors of the computed eigenpairs and their partial derivatives are investigated. Hausdorff distance and containment gap are used to measure the accuracy of approximate eigenpair partial derivatives. Error bounds on the computed eigenpairs and their partial derivatives are derived. Finally numerical experiments are reported to show the efficiency of the proposed method.     

Iterative Computation of Derivatives of Eigenvalues and Eigenvectors

Some refinements are proposed to increase the efficiency andrange of validity of an iterative method of Rudisill & Chufor numerical computation of partial derivatives of eigenvaluesand eigenvectors of a matrix which depends on a number of parameters.Difficulties arising when the matrix is close to one with repeatedeigenvalues are considered.     

IRAM‐based method for eigenpairs and their derivatives of large matrix‐valued functions

Based on the implicitly restarted Arnoldi method for eigenpairs of large matrix, a new method is presented for the computation of a few eigenpairs and their derivatives of large matrix‐valued functions. Eigenpairs and their derivatives are calculated simultaneously. Equation systems that are solved for eigenvector derivatives are greatly reduced from the original matrix size. The left eigenvectors are not required. Hence, the computational cost is saved. The convergence theory of the proposed method is established. Finally, numerical experiments are given to illustrate the efficiency of the proposed method. Copyright © 2010 John Wiley & Sons, Ltd.     

阻尼系统重特征对导数的计算

提出了一种计算阻尼系统重特征值及其特征向量导数的方法．该方法利用n维空间的特征向量计算特征对的导数,避免了状态空间中特征向量的使用,从而节省了计算量,提高了计算效率．最后以一个5自由度的非比例阻尼系统对所提方法进行了数值试验,数值结果表明方法是有效的．     

A new iterative method for many eigenpair partial derivatives

A new iterative method is proposed for computing partial derivatives of many eigenpairs. This method simultaneously computes a few eigenpair partial derivatives. For each origin shift, partial derivatives of eigenpairs whose eigenvalues are closest to the origin shift can be computed. Hence, this method may be used for partial derivatives of all eigenpairs. Convergence of the proposed method is established. Finally numerical experiments are given to show the effectiveness of the proposed method.     

Iterative computation of derivatives of repeated eigenvalues and the corresponding eigenvectors

Recently the authors proposed a simultaneous iteration algorithm for the computation of the partial derivatives of repeated eigenvalues and the corresponding eigenvectors of matrices depending on several real variables. This paper analyses the properties of that algorithm and extends it in several ways. The previous requirement that the repeated eigenvalue be dominant is relaxed, and the new generalized algorithm given here allows the simultaneous treatment of simple and repeated eigenvalues. Methods for accelerating convergence are examined. Numerical results support our theoretical analysis. Copyright © 2000 John Wiley & Sons, Ltd.     

DERIVATIVES OF EIGENPAIRS OF SYMMETRIC QUADRATIC EIGENVALUE PROBLEM

Derivatives of eigenvalues and eigenvectors with respect to parameters in symmetric quadratic eigenvalue problem are studied. The first and second order derivatives of eigenpairs are given. The derivatives are calculated in terms of the eigenvalues and eigenvectors of the quadratic eigenvalue problem, and the use of state space representation is avoided, hence the cost of computation is greatly reduced. The efficiency of the presented method is demonstrated by considering a spring-mass-damper system.     

Block tensor conjugate gradient-type method for Rayleigh quotient minimization in two-dimensional case

A method for solving a partial algebraic eigenvalues problem is constructed. It exploits tensor structure of eigenvectors in two-dimensional case. For a symmetric matrix represented in tensor format, the method finds low-rank approximations to the eigenvectors corresponding to the smallest eigenvalues. For sparse matrices, execution time and required memory for the proposed method are proportional to the square root of miscellaneous overall number of unknowns, whereas this dependence is usually linear. To maintain tensor structure of vectors at each iteration step, low-rank approximations are performed, which introduces errors into the original method. Nevertheless, the new method was proved to converge. Convergence rate estimates are obtained for various tensor modifications of the abstract one-step method. It is shown how the convergence of a multistep method can be derived from the convergence of the corresponding one-step method. Several modifications of the method with an low-rank approximation techniques were implemented on the basis of the block conjugate gradient method. Their performance is compared on numerical examples.     

Fast sensitivity analysis of defective system

Sensitivity analysis of linear vibration system is of wide interest. In this paper, sensitivity analysis based on non-defective system and defective system is summarized in all cases. Specially, for the defective systems, a fast method for the perturbation problem of state vectors is constructed in terms of the theories of generalized eigenvectors and adjoint matrices. By this method, the state vector derivatives can be expressed by a linear combination of generalized eigenvectors. The expansion coefficients can be obtained without solving large-scale equations based on eigensolutions of original system. Numerical results demonstrate the effectiveness and the stability of the method.     

非对称阻尼系统特征对一阶导数与二阶导数的计算

提出了一种计算非对称阻尼系统特征对一阶、二阶导数的方法.该方法利用阻尼系统的特征向量计算特征对的导数,避免了状态空间中特征向量的使用,节省了计算量,且不要求系统所有特征值的互异性.最后以两个非对称阻尼系统进行数值试验,数值结果表明提出的方法是有效的.     

Inverse heat problem of determining time-dependent source parameter in reproducing kernel space

A new method by the reproducing kernel Hilbert space is applied to an inverse heat problem of determining a time-dependent source parameter. The problem is reduced to a system of linear equations. The exact and approximate solutions are both obtained in a reproducing kernel space. The approximate solution and its partial derivatives are proved to converge to the exact solution and its partial derivatives, respectively. The proposed method improves the existing method. Our numerical results show that the method is of high precision.     

Two-Dimensional Parabolic Inverse Source Problem with Final Overdetermination in Reproducing Kernel Space

A new method of the reproducing kernel Hilbert space is applied to a twodimensional parabolic inverse source problem with the final overdetermination. The exact and approximate solutions are both obtained in a reproducing kernel space. The approximate solution and its partial derivatives are proved to converge to the exact solution and its partial derivatives, respectively. A technique is proposed to improve some existing methods. Numerical results show that the method is of high precision, and confirm the robustness of our method for reconstructing source parameter.     

A novel method for nonlinear fractional partial differential equations: Combination of DTM and generalized Taylor's formula

In this article, a novel numerical method is proposed for nonlinear partial differential equations with space- and time-fractional derivatives. This method is based on the two-dimensional differential transform method (DTM) and generalized Taylor's formula. The fractional derivatives are considered in the Caputo sense. Several illustrative examples are given to demonstrate the effectiveness of the present method. Results obtained using the scheme presented here agree well with the analytical solutions and the numerical results presented elsewhere. Results also show that the numerical scheme is very effective and convenient for solving nonlinear partial differential equations of fractional order.     

Estimation of partial derivative functionals with application to human mortality data analysis

To better describe and understand the time dynamics in functional data analysis, it is often desirable to recover the partial derivatives of the random surface. A novel approach is proposed based on marginal functional principal component analysis to derive the representation for partial derivatives. To obtain the Karhunen-Lo`eve expansion of the partial derivatives, an adaptive estimation is explored. Asymptotic results of the proposed estimates are established. Simulation studies show that the proposed methods perform well in finite samples. Application to the human mortality data reveals informative time dynamics in mortality rates.     

多参数结构特征二阶灵敏度

提出了一种有效计算多参数结构特征值与特征向量二阶灵敏度矩阵--Hessian矩阵的方法.将特征值和特征向量二阶摄动法转变为多参数形式,推导出二阶摄动灵敏度矩阵,由此得到特征值和特征向量的二阶估计式.该法解决了无法用直接求导法计算特征值和特征向量二阶灵敏度矩阵的问题.数值算例说明了该算法的应用和计算精度.     

An interval uncertainty analysis method for structural response bounds using feedforward neural network differentiation

This paper proposes a new interval uncertainty analysis method for structural response bounds with uncertain‑but-bounded parameters by using feedforward neural network (FNN) differentiation. The information of partial derivative may be unavailable analytically for some complicated engineering problems. To overcome this drawback, the FNNs of real structural responses with respect to structure parameters are first constructed in this work. The first-order and second-order partial derivative formulas of FNN are derived via the backward chain rule of partial differentiation, thus the partial derivatives could be determined directly. Especially, the influences of structures of multilayer FNNs on the accuracy of the first-order and second-order partial derivatives are analyzed. A numerical example shows that an FNN with the appropriate structure parameters is capable of approximating the first-order and second-order partial derivatives of an arbitrary function. Based on the parameter perturbation method using these partial derivatives, the extrema of the FNN can be approximated without requiring much computational time. Moreover, the subinterval method is introduced to obtain more accurate and reliable results of structural response with relatively large interval uncertain parameters. Three specific examples, a cantilever tube, a Belleville spring, and a rigid-flexible coupling dynamic model, are employed to show the effectiveness and feasibility of the proposed interval uncertainty analysis method compared with other methods.     

Approximate eigensolution of Laplacian matrices for locally modified graph products

Laplacian matrices and their spectrum are of great importance in algebraic graph theory. There exist efficient formulations for eigensolutions of the Laplacian matrices associated with a special class of graphs called product graphs. In this paper, the problem of determining a few approximate smallest eigenvalues and eigenvectors of large scale product graphs modified through the addition or deletion of some nodes and/or members, is investigated. The eigenproblem associated with a modified graph model is reduced using the set of master eigenvectors and linear approximated slave eigenvectors from the original model. Implicitly restarted Lanczos method is employed to obtain the required eigenpairs of the reduced problem. Examples of large scale models are included to demonstrate the efficiency of the proposed method compared to the direct application of the IRL method.     

Measuring full static displacements of cable domes based only on limited tested locations

The existing cable domes are normally faced with the possible deterioration of the structural stiffness which is caused by random member damages and pretension deviations. The static testing method can be potentially used in monitoring the structural stiffness, in which the tested locations are normally limited. To improve the testing efficiency and decrease the testing cost, a method to expand the limited static displacements is proposed. Based on the eigen-decomposition of the structural stiffness, the unified expression of the static displacements under a specific load is derived, in which the static displacements are approximately described as a linear combination of only a few same eigenvectors (i.e. contribution eigenvectors) for both the ideal structure and a random real structure. Therefore, the problem of expanding static displacements is transformed to obtain the combinational coefficients of the contribution eigenvectors. An iterative strategy based on the Fisher information matrix comprised of contribution eigenvectors is then suggested to optimize the tested locations and to obtain the unbiased combinational coefficients. A Geiger cable dome is numerically analysed. The results show that the unified expression of the static displacements under a specific load is correct, and good expansion results are obtained by the proposed method.     

A New Direction Set Method for Unconstrained Minimization without Evaluating Derivatives

A new direction set method for unconstrained minimization withoutevaluating derivatives is presented. The algorithm can be regardedas an application to function minimization of Jacobi's methodfor determining the eigenvalues and eigenvectors of a real symmetricmatrix. Numerical results are presented, illustrating the performanceof the new algorithm on well-known test problems; a comparisonwith other methods is also given.     

Valuation of Two-Factor Interest Rate Contingent Claims Using Green's Theorem

Abstract

Over the years a number of two-factor interest rate models have been proposed that have formed the basis for the valuation of interest rate contingent claims. This valuation equation often takes the form of a partial differential equation that is solved using the finite difference approach. In the case of two-factor models this has resulted in solving two second-order partial derivatives leading to boundary errors, as well as numerous first-order derivatives. In this article we demonstrate that using Green's theorem, second-order derivatives can be reduced to first-order derivatives that can be easily discretized; consequently, two-factor partial differential equations are easier to discretize than one-factor partial differential equations. We illustrate our approach by applying it to value contingent claims based on the two-factor CIR model. We provide numerical examples that illustrate that our approach shows excellent agreement with analytical prices and the popular Crank–Nicolson method.