Search NASA⌕ Search

SEARCH · Search NASA

Results for “eigenvalue method”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 73 records · Page 4
All sourcesNASA NTRSDOE OSTI
DOE OSTI2018

An approach based on dwindling filter method for positive definite generalized eigenvalue problem

In this paper, we present a new spectral residual method for solving large-scale positive definite generalized eigenvalue problems. The proposed algorithm is equipped with a dwindling multidimensional nonmonotone filter, in which the envelope dwindles as the step length decreases. We have also employed a relaxed nonmonotone line search technique in the structure of the algorithm which allows it to enjoy the nonmonotonicity from scratch. Under some mild and standard assumptions, the global convergence property of the proposed algorithm is established. An implementation of the new algorithm on some test problems shows the efficiency and effectiveness of the proposed algorithm in practice.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

A Flexible Power Method for Solving Infinite Dimensional Tensor Eigenvalue Problems

We propose a flexible power method for computing the leftmost, i.e., algebraically smallest, eigenvalue of an infinite dimensional tensor eigenvalue problem, $H x = \lambda x$, where the infinite dimensional symmetric matrix $H$ exhibits a translational invariant structure. We assume the smallest eigenvalue of $H$ is simple and apply a power iteration of $e^{-H}$ with the eigenvector represented in a compact way as a translational invariant infinite Tensor Ring (iTR). Hence, the infinite dimensional eigenvector can be represented by a finite number of iTR cores of finite rank. In order to implement this power iteration, we use a small parameter $t$ so that the infinite matrix-vector operation $e^{-Ht}x$ can efficiently be approximated by the Lie product formula, also known as Suzuki--Trotter splitting, and we employ a low rank approximation through a truncated singular value decomposition on the iTR cores in order to keep the cost of subsequent power iterations bounded. We also use an efficient way for computing the iTR Rayleigh quotient and introduce a finite size iTR residual which is used to monitor the convergence of the Rayleigh quotient and to modify the timestep $t$. In this paper, we discuss 2 different implementations of the flexible power algorithm and illustrate the automatic timestep adaption approach for several numerical examples.

Beeumen, Roel Van↗
NASA NTRS1989

Elimination of spurious eigenvalues in the Chebyshev tau spectral method

Spectral methods have been used to great advantage in hydrodynamic stability calculations; the concepts are described in Orszag's seminal application of the Chebyshev tau method to the Orr-Sommerfeld equation for plane Poiseuille flow in 1971. Orszag discusses both the Chebyshev Galerkin and the Chebyshev tau methods, but presents results for the tau method, which is easier to implement than the Galerkin method. The tau method has the disadvantage that two unstable eigenvalues are produced that are artifacts of the discretization. An extremely simple modification to the Chebyshev tau method is presented which eliminates the spurious eigenvalues. First a simplified model of the Orr-Sommerfeld equation discussed by Gottlieb and Orszag was studied. Then the Chebyshev tau method is considered, which has two spurious eigenvalues, and then a modification which eliminates them is described. Finally, results for the Orr-Sommerfeld equation are considered where the modified tau method also eliminates the spurious eigenvalues. The simplicity of the modification makes it a convenient alternative to other approaches to the problem.

Mcfadden, G. B.↗
NASA NTRS1997

The Path Resistance Method for Bounding the Smallest Nontrivial Eigenvalue of a Laplacian

We introduce the path resistance method for lower bounds on the smallest nontrivial eigenvalue of the Laplacian matrix of a graph. The method is based on viewing the graph in terms of electrical circuits; it uses clique embeddings to produce lower bounds on lambda(sub 2) and star embeddings to produce lower bounds on the smallest Rayleigh quotient when there is a zero Dirichlet boundary condition. The method assigns priorities to the paths in the embedding; we show that, for an unweighted tree T, using uniform priorities for a clique embedding produces a lower bound on lambda(sub 2) that is off by at most an 0(log diameter(T)) factor. We show that the best bounds this method can produce for clique embeddings are the same as for a related method that uses clique embeddings and edge lengths to produce bounds.

Guattery, Stephen↗
NASA NTRS1980

Comments on numerical solution of boundary value problems of the Laplace equation and calculation of eigenvalues by the grid method

The mathematics involved in numerically solving for the plane boundary value of the Laplace equation by the grid method is developed. The approximate solution of a boundary value problem for the domain of the Laplace equation by the grid method consists of finding u at the grid corner which satisfies the equation at the internal corners (u=Du) and certain boundary value conditions at the boundary corners.

Lyusternik, L. A.↗
NASA NTRS1977

Complex eigenvalue extraction in NASTRAN by the tridiagonal reduction (FEER) method

An extension of the Tridiagonal Reduction (FEER) method to complex eigenvalue analysis in NASTRAN is described. As in the case of real eigenvalue analysis, the eigensolutions closest to a selected point in the eigenspectrum are extracted from a reduced, symmetric, tridiagonal eigenmatrix whose order is much lower than that of the full size problem. The reduction process is effected automatically, and thus avoids the arbitrary lumping of masses and other physical quantities at selected grid points. The statement of the algebraic eigenvalue problem admits mass, damping and stiffness matrices which are unrestricted in character, i.e., they may be real, complex, symmetric or unsymmetric, singular or non-singular.

Newman, M.↗
NASA NTRS1973

Eigenvalue routine by Sturm sequence method

Computer program has been generated for efficient solution of certain broad classes of eigenvalue problems. Procedure fully exploits banded nature of associated matrices and further enables user to compute either all roots or any specific ones desired. Special storage options enable storing only nonzero elements of associated main matrix of eigenvalue problem.

Gupta, K. K.↗
NASA NTRS1976

Alternative methods for the design of jet engine control systems

Various alternatives to linear quadratic design methods for jet engine control systems are discussed. The main alternatives are classified into two broad categories: nonlinear global mathematical programming methods and linear local multivariable frequency domain methods. Specific studies within these categories include model reduction, the eigenvalue locus method, the inverse Nyquist method, polynomial design, dynamic programming, and conjugate gradient approaches.

Sain, M. K.↗
NASA NTRS1985

A control system design approach for flexible spacecraft

A control system design approach for flexible spacecraft is presented. The control system design is carried out in two steps. The first step consists of determining the ideal control system in terms of a desirable dynamic performance. The second step consists of designing a control system using a limited number of actuators that possess a dynamic performance that is close to the ideal dynamic performance. The effects of using a limited number of actuators is that the actual closed-loop eigenvalues differ from the ideal closed-loop eigenvalues. A method is presented to approximate the actual closed-loop eigenvalues so that the calculation of the actual closed-loop eigenvalues can be avoided. Depending on the application, it also may be desirable to apply the control forces as impulses. The effect of digitizing the control to produce the appropriate impulses is also examined.

Silverberg, L. M.↗
DOE OSTI2019

On the All-Fiber Optical Methods of the Generation and Recognition of Soliton States

The state of a soliton is characterized by the eigenvalues of the Zakharov–Shabat problem. The data transfer rate in the fiber-optic communication lines that use eigenvalues for coding signals can be significantly increased only by developing optical eigenvalue control methods. We propose to use optical fibers with a sinusoidally changing dispersion to generate given sets of complex eigenvalues and to detect soliton states. Under the action of a periodic change in dispersion, multisoliton pulses divide into several solitons moving at different group velocities. This effect can be used to prepare fixed sets of eigenvalues. A soliton signal can be recognized by analyzing pulses and their spectra at the exit from a fiber with a periodically changing dispersion. Using the solitons specified by sets of four eigenvalues, we show that the field at the exit from a fiber corresponds to a unique combination of the spectrum and the number of pulses determined by initial eigenvalues.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS1991

A comparison of numerical methods for the Rayleigh equation in unbounded domains

A second-order finite difference and two spectral methods, including a Chebyshev tau and a Chebyshev collocation method were implemented to determine the linear hydrodynamic stability of an unbounded shear flow. The velocity profile of the basic flow in the stability analysis mimicks that of a one-stream free mixing layer. Local and global eigenvalue solution methods are used to determine individual eigenvalues and the eigenvalue spectrum, respectively. The calculated eigenvalue spectrum includes a discrete mode, a continuous spectrum associated with the equation singularity and a continuous spectrum associated with the domain unboundedness. The efficiency and the accuracy of these discretization methods in the prediction of the eigensolutions of the discrete mode were evaluated by comparison with a conventional shooting procedure. Their capabilities in mapping out the continuous eigenvalue spectra are also discussed.

Liou, W. W.↗
NASA NTRS1977

Determination of pole sensitivities by Danilevskii's method

In control theory, a synonymous term for pole sensitivity is eigenvalue sensitivity. Existing methods of calculating eigenvalues are cumbersome, and cannot be trusted for systems roughly greater than tenth order. The method proposed in the present paper is applicable to high-order system. (It has been routinely used to generate eigenvalue sensitivities for systems up to 26th order, using a UNIVAC 1106.) Danilevskii's method is shown to be suitable for performing the necessary evaluations. The result is a rational function that can be used to evaluate the sensitivities for all distinct poles.

Nail, J. B.↗
NASA NTRS1993

Implicit treatment of diffusion terms in lower-upper algorithms

A method is presented which allows diffusion terms to be treated implicitly in the lower-upper (LU) algorithm (which is a commonly used method for solving 'compressible' Euler and Navier-Stokes equations) so that the algorithm's good stability properties will not be impaired. The new method generalizes the concept of LU factorization from that associated with the sign of eigenvalues to that associated with backward- and forward-difference operators without regard to eigenvalues. The method is verified in a turbulent boundary layer study.

Shih, T. I.-P.↗
DOE OSTI2018

Recursive integral method with Cayley transformation

Summary The recently developed RIM (recursive integral method) finds eigenvalues in a region of the complex plane. It computes an indicator to test if the region contains eigenvalues using an approximate spectral projection. If the indicator shows that the region contains eigenvalues, it is subdivided into smaller regions, which are then recursively tested and subdivided until any eigenvalues are isolated to a specified precision. We propose an enhancement to RIM that uses Cayley transformations and Arnoldi's method to greatly decrease the number of factorizations required to solve the linear systems that arise from a discretized contour integral to compute the indicator, which substantially reduces the computational cost. We demonstrate the efficacy of our method with numerical examples and compare with the implicitly restarted Arnoldi method, as implemented in eigs in MATLAB.

Huang, R.↗