Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
30 views

Lecture 6: Gauss Elimination: A Is A Diagonal Matrix

The document discusses the Gauss elimination method for solving systems of linear equations. It involves transforming the coefficient matrix A into an upper triangular matrix using row operations, then solving the system using back substitution. Examples are provided for diagonal, upper triangular, and general matrices. The method is applied to solve a system from a two-loop circuit problem. Homework problems are assigned to use Gauss elimination to solve a specific system and prove a proposition about lower triangular matrices.

Uploaded by

TukangSantet
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
30 views

Lecture 6: Gauss Elimination: A Is A Diagonal Matrix

The document discusses the Gauss elimination method for solving systems of linear equations. It involves transforming the coefficient matrix A into an upper triangular matrix using row operations, then solving the system using back substitution. Examples are provided for diagonal, upper triangular, and general matrices. The method is applied to solve a system from a two-loop circuit problem. Homework problems are assigned to use Gauss elimination to solve a specific system and prove a proposition about lower triangular matrices.

Uploaded by

TukangSantet
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Lecture 6: Gauss Elimination

The objective is to solve for n unknowns given n linear equations. In terms of


matrices, find the nx1 column vector x so that Ax = d where A is nxn and d is nx1. We
will do this using the method called Gauss elimination, which requires about n2 units of
memory and about n3/3 arithmetic operations. So, if n is not too large, this method is
very useful; otherwise, one must go to other more specialized methods. We begin by
considering several special matrices A.

A is a Diagonal Matrix.
Let n = 3.
a11 x1 d1
a22 x = d
2 2
a3 x3 d3
Or ,
a11 x1 = d1
a22 x2 = d 2
a33 x3 = d3
So, if each aii 0, then xi = di / aii .

A is an Upper Triangular Matrix.


Let n = 3.
a11 a12 a13 x1 d1
a22 a23 x2 = d 2

a33 x3 d3
Or ,
a11 x1 + a12 x2 + a13 x3 = d1
a22 x2 + a23 x3 = d 2
a33 x3 = d3
So, if a33 0, then
x3 = d3 / a33 .
Solve the middle equation, if a22 0,
a22 x2 = d 2 a23 x3
x2 = (d 2 a23 x3 ) / a22 .
Solve the first equation, if a11 0,
a11 x1 = d1 a12 x2 a13 x3
x1 = (d1 a12 x2 a13 x3 ) / a11.

Proposition 3. Let A be a nxn upper or lower triangular matrix,


If each diagonal component is non-zero, then Ax = d has a solution.

Consider the following system with three unknowns and three equations, whose
coefficient matrix is not upper or lower triangular.
2 x1 + 6 x2 + 8 x3 = 16
4 x1 + 15 x2 + 19 x3 = 38
2 x1 + 0 x2 + 3 x3 = 6
Or , Ax = d
2 6 8 x1 16
4 15 19 x = 38
2
2 0 3 x3 6
Or , as a 3 x 4 augmented matrix
2 6 8 16
[ A d ] = 4 15 19 38 .
2 0 3 6

We will try to transform this into an equivalent problem that has an upper triangular
coefficient matrix, which may be solved by a backward substitution. Often this can be
done by a combination of row (or equation) operations:
(i). interchange rows (or equations)
(ii). multiply rows (or equations) by a suitable constant and
(iii). add or subtract rows (or equations).
Row Operations on the Augmented Matrix:
We will try to transform the augmented matrix to an upper triangular matrix by
using row operations to make the left most column 0 below row 1.
row 2 2(row 1) or multiply [A d] by E21(-2) and then
row 3 1(row 1) or multiply E21(-2) [A d] by E31(-1) to get
2 6 8 16
E31 (1) E21 (2)[ A d ] = 0 3 3 6 .
0 6 5 10

Next, we must use row 2 to transform the 6 to 0.


row 3 + 2(row 2) or multiply E31(-1) E21(-2) [A d] by E32(2) to get
2 6 8 16
E32 (2) E31 (1) E21 (2)[ A d ] = 0 3 3 6 .
0 0 1 2
Backward Substitution on the Transformed Augmented Matrix:
This is equivalent to the upper triangular system, which can be solved by
backward substitution.
2 6 8 x1 16
0 3 3 x = 6
2
0 0 1 x3 2
Solve for x3 , 1x3 = 2 so that x3 = 2.
Solve for x2 , 3 x2 + 3 x3 = 6 so that x2 = 0.
Solve for x1 , 2 x1 + 6 x2 + 8 x3 = 16 so that x1 = 0.

Gauss Elimination Method. Consider Ax = d.


Step 1: Use row operations on the augmented matrix to transform Ax = d to an
upper triangular problem.
Step 2: Use backward substitution to solve the upper triangular problem.

See the Matlab demo gauss_el.m for additional examples. There are a number of
very good computer implementations of the Gauss elimination method. In Matlab the
simple command A\d can be used to compute the solution for most problems Ax = d.
Application to the Two-loop Circuit.
In lecture 4 we formulated the algebraic problem for the two-loop circuit problem:
1 1 1 i1 0
R 0 R3 i2 = E1 .
1
0 R2 R3 i3 E2
Let R1 = 1, R2 = 2, R3 = 3, E1 = 10 and E2 = 20.
The augemented matrix is
1 1 1 0
[ A d ] = 1 0 3 10 .
0 2 3 20

Step 1: Use row operations.


row 2 row 1 or multiply by E12(-1) to get
1 1 1 0
E21 (1)[ A d ] = 0 1 4 10 .
0 2 3 20

row 3 + 2(row2) or multiply by E32(2) to get


1 1 1 0
E32 (2) E21 (1)[ A d ] = 0 1 4 10 .

0 0 11 40

Step 2: Use backward substitution.


1 1 1 i1 0
0 1 4 i = 10
2
0 0 11 i3 40
Solve for i3 , 11x3 = 20 so that i3 = 40 /11.
Solve for i2 , 1i2 4i3 = 10 so that i2 = 50 /11.
Solve for i1 , 1i1 1i2 + 1i3 = 0 so that i1 = 10 /11.
Homework.
1. Use Gauss elimination to solve Ax = d when
1 3 5 1
A = 0 2 1 and d = 2 .

2 2 0 3

2. Prove Proposition 3 when A is lower triangular.

You might also like