iterative methods to solve equation system

Upload: tatodc7

Post on 29-May-2018

227 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/9/2019 Iterative Methods to solve equation system

    1/19

    System of linear equations

    Iterative Methods

  • 8/9/2019 Iterative Methods to solve equation system

    2/19

    TOPICS

    Jacobi Method

    Gauss-Seidel Method Gauss-Seidel Method amended

    Nonlinear equation systems

  • 8/9/2019 Iterative Methods to solve equation system

    3/19

    Description All the presentation is going to manage

    the follow notation to write the matrixes

    -

    !

    -

    y

    -

    m

    i

    n

    i

    mnmm

    inii

    n

    n

    b

    b

    b

    b

    x

    x

    x

    x

    aaa

    aaa

    aaa

    aaa

    2

    1

    2

    1

    21

    21

    22221

    11211

    .

    /

    .

    //

    .

    .

    A x b

    Ax=bAx=b

    Given a square

    system ofn

    linear equations

    with unknown x:

  • 8/9/2019 Iterative Methods to solve equation system

    4/19

    Jacobi method Jacobi method is an algorithm for determining the solutions of a

    system of linear equations with largest absolute values in each

    row and column dominated by the diagonal element.

    Each diagonal element is solved for, and an approximate value

    plugged in. The process is then iterated until it converges

    This algorithm is a stripped-down version of the Jacobi

    transformation method of matrix diagonalization.

  • 8/9/2019 Iterative Methods to solve equation system

    5/19

    Given a square system of n linear equations

    -

    !

    mnmm

    inii

    n

    n

    aaa

    aaa

    aaa

    aaa

    A

    .

    /

    .

    //

    .

    .

    21

    21

    22221

    11211

    Ax=bAx=b

    -

    !

    n

    i

    x

    x

    x

    x

    x

    2

    1

    -

    !

    m

    i

    b

    b

    b

    b

    b

    2

    1

    , ,

  • 8/9/2019 Iterative Methods to solve equation system

    6/19

    ThenA can be decomposed into a diagonal component D,

    and the remainder R:

    -

    mna

    aa

    D

    .

    /

    .

    //

    .

    .

    00

    000

    0000

    22

    11

    -

    0

    0

    0

    0

    21

    21

    221

    112

    .

    /

    //

    .

    .

    mm

    inii

    n

    n

    aa

    aaa

    aa

    aa

    R,

    A = D + RA = D + R

  • 8/9/2019 Iterative Methods to solve equation system

    7/19

    The system of linear equations may be rewritten as:

    (D + R) x = b(D + R) x = b DxDx + Rx = b+ Rx = b

    And finally

    DxDx = b= b -- RxRx

  • 8/9/2019 Iterative Methods to solve equation system

    8/19

    The Jacobi method is an iterative technique that solves

    the left hand side of this expression for x, using previous

    value for x on the right hand side. Analytically, this may

    be written as

    )( )(1)1( kk

    i RxbDx !

    The element-based formula is thus

    )(1 )(

    ,

    ,

    )1( k

    j

    ij

    jii

    ii

    k

    i xaba

    x {

    !

    ii = 1,2,= 1,2,

  • 8/9/2019 Iterative Methods to solve equation system

    9/19

    the computation of xi(k+1) requires each element in x(k)

    except itself. Unlike the GaussSeidel method, we can't

    overwrite xi(k) with xi

    (k+1), as that value will be needed

    by the rest of the computation.

    This is the most meaningful difference between theJacobi and GaussSeidel methods, and is the reason

    why the former can be implemented as a parallel

    algorithm, unlike the latter. The minimum amount of

    storage is two vectors of size n.

  • 8/9/2019 Iterative Methods to solve equation system

    10/19

    Gauss Seidel Method Also known as the Liebmann method or the method of

    successive displacement, is an iterative method used to solve a

    linear system of equations.

    Though it can be applied to any matrix with non-zero elements on

    the diagonals, convergence is only guaranteed if the matrix is

    either diagonally dominant, or symmetric and positive definite.

  • 8/9/2019 Iterative Methods to solve equation system

    11/19

    Given a square system of n linear equations

    -

    !

    mnmm

    inii

    n

    n

    aaa

    aaa

    aaa

    aaa

    A

    .

    /

    .

    //

    .

    .

    21

    21

    22221

    11211

    Ax=bAx=b

    -

    !

    n

    i

    x

    x

    x

    x

    x

    2

    1

    -

    !

    m

    i

    b

    b

    b

    b

    b

    2

    1

    , ,

  • 8/9/2019 Iterative Methods to solve equation system

    12/19

    A can be decomposed into a lower triangular component L

    and a strictly upper triangular component U

    -

    !

    000

    000

    000

    2

    112

    .

    /

    //

    .

    .

    in

    n

    n

    a

    aaa

    U

    -

    !

    mnmm

    ii

    aaa

    aa

    aa

    a

    L

    .

    /

    //

    .

    .

    21

    21

    2221

    11

    0...

    0

    00

    ,

    A = L + UA = L + U Lx = bLx = b -- UxUx

  • 8/9/2019 Iterative Methods to solve equation system

    13/19

    The GaussSeidel method is an iterative technique that

    solves the left hand side of this expression for x, using

    previous value for x on the right hand side. Analytically,

    this may be written as

    )( )(1)1( kk

    i UxbLx !

    However, by taking advantage of the triangular form ofL*, the

    elements ofx(k+1) can be computed sequentially using forward

    substitution

    )(1 )1(

    ,

    )(

    ,

    ,

    )1(

    "

    !k

    j

    ij

    ji

    k

    j

    ij

    jii

    ii

    k

    i xaxaba

    x

    Similar to

    Jacobi

    Method

    ii = 1,2,= 1,2,

  • 8/9/2019 Iterative Methods to solve equation system

    14/19

    The computation of xi(k+1) requires each element in x(k)

    except itself. Unlike the GaussSeidel method, we can't

    overwrite xi(k) with xi

    (k+1), as that value will be needed

    by the rest of the computation.

  • 8/9/2019 Iterative Methods to solve equation system

    15/19

    Convergence criterion to Gauss- Seidel method

    To ensure the convergence for the method, the

    diagonal coefficient of each equation must behigher than the sum of the absolute value from the

    others coefficients of the equation

    {!

    "

    n

    ijj

    jiij aa1

    , ||||

    The last criterion it is enough but not necessary to theconvergence.

    The convergence is ensure when the restriction is

    satisfied. Systems that meet the restriction are known

    as diagonally dominant

  • 8/9/2019 Iterative Methods to solve equation system

    16/19

    Gauss-Seidel Method amended

    (relaxation)

    Relaxation is an improvement made tothe Gauss-Seidel Method to achieve

    faster the convergence.

  • 8/9/2019 Iterative Methods to solve equation system

    17/19

    Improvement of convergence with relaxation

    The relaxation represents a soft modification to Gauss-

    Seidel method to improve the convergence. After all theprocess and the calculation of each x, that value is modify

    through an average of the results of each iteration made

    before an the actual one.

    last

    i

    new

    i

    new

    i xxx )1( PP !

    Where is a weighted factor that has a value between 0 and

    2. If has a value between 0 and 1, the result is a weightedaverage. These type of modifications are known as sub-

    relaxation.

    To values of between 1 and 2, is given an extra value to the

    actual one. This modification are called over-relaxation.

  • 8/9/2019 Iterative Methods to solve equation system

    18/19

    NON-LINEAR EQUATIONS

    Generally, nonlinear algebraic problems are often exactly solvable, and if notthey usually can be thoroughly understood through qualitative and numerical

    analysis. As an example, the equation

    012 ! xx

    Could be written as

    cxf !)( Where xxxf ! 2)( And C = 1C = 1

    and is nonlinear because f(x) satisfies neither additively nor

    homogeneity (the nonlinearity is due to the x2). Though nonlinear,

    this simple example may be solved exactly (via the quadratic

    formula) and is very well understood

  • 8/9/2019 Iterative Methods to solve equation system

    19/19

    Bibliography

    CHAPRA, Steven C. y CANALE, Raymond P.: Mtodos

    Numricos para Ingenieros. McGraw Hill 2002.

    Black, Noel and Moore, Shirley, "Gauss-Seidel Method"

    from MathWorld

    http://www.slideshare.net/nestorbalcazar/mtodos-

    numricos-06

    Khalil, Hassan K. (2001). Nonlinear Systems. Prentice

    Hall. ISBN 0-13-067389-7.