Scribd
Upload
14 views12 pages

4.0 LU Decomposition

This document describes the LU decomposition method for solving systems of linear equations. It involves decomposing the coefficient matrix A into lower and upper triangular matrices L and U such that LU = A. The elements of L and U can then be determined through matrix multiplication and the system of equations can be solved using back substitution.

Uploaded by

kjptabarno
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
14 views12 pages

4.0 LU Decomposition

This document describes the LU decomposition method for solving systems of linear equations. It involves decomposing the coefficient matrix A into lower and upper triangular matrices L and U such that LU = A. The elements of L and U can then be determined through matrix multiplication and the system of equations can be solved using back substitution.

Uploaded by

kjptabarno
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 12

LU DECOMPOSITION

Civil Engineering Department

METHOD/ CROUT’S
METHOD/
College of Engineering

CHOLESKEY’S
METHOD
Contents
• Introduction
LU DECOMPOSITION METHOD Choleskey’s Method, also known as Crout’s Method or LU
Decomposition Method, is a method of matrix decomposition and
factorization.

Illustration:

Given is a set of simultaneous linear algebraic equations:


a11X1 + a12X2 + a13X3 = C1
a21X1 + a22X2 + a23X3 = C2 --------1
a31X1 + a32X2 + a33X3 = C3

This is represented by the matrix equation


a 11 a 12 a 13 X1 C1
a 21 a 22 a 23 X2 = C2
---------2
a 31 a 32 a 33 X3 C3
If we let A represent the coefficient matrix, X the
LU DECOMPOSITION METHOD column matrix of the unknowns, and C the column
matrix of the constants, equation 2 can be replaced by:

A X – C =0 ----3

If we could reduce the system of equations into an


equivalent system of the form

1 u 12 u 13 X1 d1
0 1 u 23 X2 = d2 -------4
0 0 1 X3 d3
Equation 4 can be easily solved by back substitution as
LU DECOMPOSITION METHOD discussed in Gaussian Elimination Method. The upper unit
triangular matrix of equation 4 will be represented by uij and
equation 4 can be written as
U X – D =0 ----5
Suppose there exists a lower triangular matrix
l11 0 0
l21 l22 0
l31 l32 l33

which will be represented by L and has the following


property:
“When we multiply the left side of equation 5 by L, it will
give us the left side of equation 3”.
In accordance with the given property of L, we have:
LU DECOMPOSITION METHOD
L U X - D = A X - C

Therefore:
LU = A -------6
and L D = C -------7

Equation 6 and 7 may be combined into a single matrix


equation

l11 0 0 1 u12 u13 l d1 a 11 a 12 a 13 l C1


l21 l22 0 0 1 u23 l d2 = a 21 a 22 a 23 l C2
l31 l32 l33 0 0 1 l d3 a 31 a 32 a 33 l C3
For the sake of convenience, the d’s will be represented by
ui4 and the C’s will be represented by ai4
LU DECOMPOSITION METHOD

l11 0 0 1 u12 u13 u14 a11 a12 a13 a14


l21 l22 0 0 1 u23 u24 = a21 a22 a23 a24
----8
l31 l32 l33 0 0 1 u34 a31 a32 a33 a34

The lij and uij elements may be determined from equation 8.


The augmented U matrix of equation 8 is the augmented
matrix of the equivalent triangular set given by equation 4.
Therefore, with the uij determined, the unknown X may be
determined by back substitution.
LU DECOMPOSITION METHOD MATRIX MULTIPLICATION:
1. Note that the first column of the matrix L is equal to the
first column of the augmented matrix A.

li1 = ai1

l11 (1) + 0 (0) + 0 (0) = a11 l11 = a11


l21 (1) + l22(0) + 0 (0) = a21 l21 = a21
l31 (1) + l32(0) + l33 (0) = a31 l31 = a31

We also note that:


l11 (u12) + 0 (1) + 0 (0) = a12 u12 = a12 / l11
l11 (u13) + 0 (u23) + 0 (1) = a13 u13 = a13 / l11
l11 (u14) + 0 (u24) + 0 (u34) = a14 u14 = a14 / l11
The first row of the U is equal to the first row of the
LU DECOMPOSITION METHOD augmented A matrix divided by l11 (or a11).
uij = aij / l11

2. After determining the 1st column of L and the 1st row of


U, we can now proceed in determining the 2nd column
of L, followed by the 2nd row of U.
l21 (u12) + l22 (1) + 0 (0) = a22
l31 (u12) + l32 (1) + l33 (0) = a32

For which we obtain


l22 = a22 - l21 (u12)
l32 = a32 - l31 (u12)
LU DECOMPOSITION METHOD Also,
l21 (u13) + l22 (u23) + 0 (1) = a23
l21 (u14) + l22 (u24) + 0 (u34) = a24
Or
 u23 = a23 - l21 (u13) / l22

 u24 = a24 - l21 (u14) / l22

3. Now we can determine the 3rd column of L and the 3rd


row of U.
l31 (u13) + l32 (u23) + l33 (1) = a33

 l33 = a33 - l31 (u13) + l32 (u23)


l31 (u14) + l32 (u24) + l33 (u34) = a34
LU DECOMPOSITION METHOD
 u34 = a34 - l31 (u14) + l32 (u24) / l33

In general, the equation for the elements of L and the


augmented U matrix for “n” number of equations in “n”
unknowns is:

1. lij = aij for i = 1, 2, . . . . n


j=1
1. A
2. uij = aij / lii for i=1
j = i + 1, i + 2, . . i + n
3. when j < i
Systems of Linear Equations
for j = 2, 3 . . . . n
i = j, i +1, . . . . . n (for each value of i)

4. when i < j
u
for i = 2, 3 . . . . n
j = i +1, i + 2,. . . i + n (for each value of j)

You might also like