• 제목/요약/키워드: Lanczos method

검색결과 48건 처리시간 0.02초

A study on modified biorthogonalization method for decreasing a breakdown condition

  • Kim, Sung-Kyung
    • 한국산업정보학회논문지
    • /
    • 제7권5호
    • /
    • pp.59-66
    • /
    • 2002
  • 대규모 비대칭 행렬의 특정 고유치들이 많은 중요한 과학, 공학 문제들에서 요구된다. 그 문제를 해결할 수 있는 방법 중의 하나인 biorthogonal 란초스 알고리즘은 심각한 문제점이 있는데, 어떤 특이한 상황에서 알고리즘을 계속할 수 없는 경우가 발생할 수 있다는 것이다. 본 논문에서는 기본적인 biorhogonal 알고리즘이 만드는 축소된 삼중 대각 행렬에 대하여 동일한 고유치를 발견할 수 있는 향상된 biorhogonal 란초스 알고리즘을 소개한다. 이 새로운 알고리즘은 대규모 비대칭 행렬의 특정 고유치들을 구할 수 있으며 기본적인 biorthogonal 란초스 알고리즘에 비해서 안정적인 방법이라는 것을 Cray 컴퓨터를 이용한 실험을 통해서 보여준다.

  • PDF

PERFORMANCE ENHANCEMENT OF PARALLEL MULTIFRONTAL SOLVER ON BLOCK LANCZOS METHOD

  • Byun, Wan-Il;Kim, Seung-Jo
    • Journal of the Korean Society for Industrial and Applied Mathematics
    • /
    • 제13권1호
    • /
    • pp.13-20
    • /
    • 2009
  • The IPSAP which is a finite element analysis program has been developed for high parallel performance computing. This program consists of various analysis modules - stress, vibration and thermal analysis module, etc. The M orthogonal block Lanczos algorithm with shiftinvert transformation is used for solving eigenvalue problems in the vibration module. And the multifrontal algorithm which is one of the most efficient direct linear equation solvers is applied to factorization and triangular system solving phases in this block Lanczos iteration routine. In this study, the performance enhancement procedures of the IPSAP are composed of the following stages: 1) communication volume minimization of the factorization phase by modifying parallel matrix subroutines. 2) idling time minimization in triangular system solving phase by partial inverse of the frontal matrix and the LCM (least common multiple) concept.

  • PDF

Lanczos 알고리즘을 도입한 Ritz Vector법에 의한 구조물의 동적해석 (Structural Dynamic Analysis by Ritz Vector Method Modified with Lanczos Algorithm)

  • 심재수;황의승;박주경
    • 전산구조공학
    • /
    • 제8권4호
    • /
    • pp.181-187
    • /
    • 1995
  • 동적해석에 대한 최근의 연구는 구조물의 자유도보다 적은 모우드 형상들을 사용하여 구조물을 해석하는 효과적인 방법을 찾는데 있다. Ritz알고리즘과 모우드가속도법은 모우드중첩법을 개선하고자 개발되었는데, Ritz알고리즘은 하중의 공간적 특성을 포함하지만, 계산과정에서 유용한 직교성을 잃는 경향이 있으며, 모우드가속도법은 만족할 만한 해를 얻기 위해 많은 수의 모우드 형상들을 고려해야 하는 단점이 있다. 또한 앞의 두 방법을 조합한 방법이 개발되었으나 너무 많은 계산과정과 시간을 필요로 한다. 이 연구의 목적은 Lanczos알고리즘을 이용하여 Ritz알고리즘의 효율성과 정확성을 보완하고 이를 프로그램화하여 검증하는데 있다. 본 연구의 결과로부터 Modified Ritz알고리즘을 이용한 동적해석방법이 합리적임이 증명되었다.

  • PDF

Lanczos 방법에 의한 비비례 감쇠 시스템의 고유치 해석 (Solution of Eigenproblems for Non-proportional Damping Systems by Lanczos Method)

  • 김만철;정형조;오주원;이인원
    • 한국전산구조공학회:학술대회논문집
    • /
    • 한국전산구조공학회 1998년도 봄 학술발표회 논문집
    • /
    • pp.283-290
    • /
    • 1998
  • A solution method is presented to solve the eigenproblem arising in tile dynamic analysis of non-proportional damping systems with symmetric matrices. The method is based on tile use of Lanczos method to generate a Krylov subspace of trial vectors, witch is then used to reduce a large eigenvalue problem to a much smaller one. The method retains the η order quadratic eigenproblem, without the need to the method of matrix augmentation traditionally used to cast the problem as a linear eigenproblem of order 2n. In the process, the method preserves tile sparseness and symmetry of the system matrices and does not invoke complex arithmetics, therefore, making it very economical for use in solving large problems. Numerical results are presented to demonstrate the efficiency and accuracy of the method.

  • PDF

Efficient m-step Generalization of Iterative Methods

  • 김선경
    • 한국산업정보학회논문지
    • /
    • 제11권5호
    • /
    • pp.163-169
    • /
    • 2006
  • In order to use parallel computers in specific applications, algorithms need to be developed and mapped onto parallel computer architectures. Main memory access for shared memory system or global communication in message passing system deteriorate the computation speed. In this paper, it is found that the m-step generalization of the block Lanczos method enhances parallel properties by forming in simultaneous search direction vector blocks. QR factorization, which lowers the speed on parallel computers, is not necessary in the m-step block Lanczos method. The m-step method has the minimized synchronization points, which resulted in the minimized global communications and main memory access compared to the standard methods.

  • PDF

AN AFFINE SCALING INTERIOR ALGORITHM VIA CONJUGATE GRADIENT AND LANCZOS METHODS FOR BOUND-CONSTRAINED NONLINEAR OPTIMIZATION

  • Jia, Chunxia;Zhu, Detong
    • Journal of applied mathematics & informatics
    • /
    • 제29권1_2호
    • /
    • pp.173-190
    • /
    • 2011
  • In this paper, we construct a new approach of affine scaling interior algorithm using the affine scaling conjugate gradient and Lanczos methods for bound constrained nonlinear optimization. We get the iterative direction by solving quadratic model via affine scaling conjugate gradient and Lanczos methods. By using the line search backtracking technique, we will find an acceptable trial step length along this direction which makes the iterate point strictly feasible and the objective function nonmonotonically decreasing. Global convergence and local superlinear convergence rate of the proposed algorithm are established under some reasonable conditions. Finally, we present some numerical results to illustrate the effectiveness of the proposed algorithm.

A Study on the Efficient m-step Parallel Generalization

  • Kim, Sun-Kyung
    • 한국정보기술응용학회:학술대회논문집
    • /
    • 한국정보기술응용학회 2005년도 6th 2005 International Conference on Computers, Communications and System
    • /
    • pp.13-16
    • /
    • 2005
  • It would be desirable to have methods for specific problems, which have low communication costs compared to the computation costs, and in specific applications, algorithms need to be developed and mapped onto parallel computer architectures. Main memory access for shared memory system or global communication in message passing system deteriorate the computation speed. In this paper, it is found that the m-step generalization of the block Lanczos method enhances parallel properties by forming m simultaneous search direction vector blocks. QR factorization, which lowers the speed on parallel computers, is not necessary in the m-step block Lanczos method. The m-step method has the minimized synchronization points, which resulted in the minimized global communications compared to the standard methods.

  • PDF

Effect of cylinder diameter and boat tail angle on the free vibration characteristics of a typical payload fairing

  • Ramamurti, V.;Rajarajan, S.;Rao, G. Venkateswara
    • Structural Engineering and Mechanics
    • /
    • 제13권3호
    • /
    • pp.345-353
    • /
    • 2002
  • Three noded plate and shell finite element and 3D beam element in conjunction with Lanczos method are used for studying the effect of boat tail angle on the free vibration characteristics of a typical payload fairing for three different cylinder diameters with height to diameter ratio 1.5. Configurations without boat tail structural member are also studied. One half of the one side fairing structure is considered for the analysis and symmetric boundary conditions are used.

NUMERICAL IMPLEMENTATION OF THE QMR ALGORITHM BY USING DISCRETE STOCHASTIC ARITHMETIC

  • TOUTOUNIAN FAEZEH;KHOJASTEH SALKUYEH DAVOD;ASADI BAHRAM
    • Journal of applied mathematics & informatics
    • /
    • 제17권1_2_3호
    • /
    • pp.457-473
    • /
    • 2005
  • In each step of the quasi-minimal residual (QMR) method which uses a look-ahead variant of the nonsymmetric Lanczos process to generate basis vectors for the Krylov subspaces induced by A, it is necessary to decide whether to construct the Lanczos vectors $v_{n+l}\;and\;w{n+l}$ as regular or inner vectors. For a regular step it is necessary that $D_k\;=\;W^{T}_{k}V_{k}$ is nonsingular. Therefore, in the floating-point arithmetic, the smallest singular value of matrix $D_k$, ${\sigma}_min(D_k)$, is computed and an inner step is performed if $\sigma_{min}(D_k)<{\epsilon}$, where $\epsilon$ is a suitably chosen tolerance. In practice it is absolutely impossible to choose correctly the value of the tolerance $\epsilon$. The subject of this paper is to show how discrete stochastic arithmetic remedies the problem of this tolerance, as well as the problem of the other tolerances which are needed in the other checks of the QMR method with the estimation of the accuracy of some intermediate results. Numerical examples are used to show the good numerical properties.

중복근을 갖는 비비례 감쇠시스템의 고유치 해석 (Solution of Eigenvalue Problems for Nonclassically Damped Systems with Multiple Frequencies)

  • 김만철;정형조;오주원;이인원
    • 전산구조공학
    • /
    • 제11권1호
    • /
    • pp.205-216
    • /
    • 1998
  • 본 논문에서는 중복근을 갖는 비비례 감쇠시스템의 고유치 해석 방법을 제안하였다. 2차 고유치 문제의 행렬 조합을 통한 선형 방정식에 수정된 Newton-Raphson기법과 고유벡터의 직교성을 적용하여 제안방법의 알고리즘을 유도하였다. 벡터 반복법 또는 부분공간 반복법과 같은 기존의 반복법에서는 수렴성을 향상시키기 위해 변위법을 적용하였으며, 이 값이 시스템의 고유치에 근사하게 되면 행렬분해 과정에서 특이성이 발생한다. 그러나 제안방법은 구하고자 하는 고유치가 중복근이 아닐 경우에, 변위값이 시스템의 고유치 일지라도 항상 정칙성을 유지하며, 이것을 해석적으로 증명하였다. 제안방법은 수정된 Newton-Raphson기법을 이용하기 때문에 초기값을 필요로 한다. 제안방법의 초기값으로는 반복법의 중간결과나 근사법의 결과를 사용할 수 있다. 이들 방법중 Lanczon방법이 가장 효율적으로 좋은 초기값을 제공하기 때문에 Lanczon방법의 결과를 제안방법의 초기값으로 사용하였다. 제안방법의 효율성을 증명하기 위하여 두가지 예제 구조물에 대해 해석시간 및 수렴성을 가장 많이 사용하고 있는 부분공간 반복법과 Lanczon방법의 결과와 비교하였다.

  • PDF