 OTHER

ANALYTICAL
OTHER METHODS METHODS
OF SOLUTION  NUMERICAL
METHODS
 Preliminary
Some Other Analytical Methods
Numerical Methods: the
CONTENTS Checklists
Summary
  In this section we consider other method of
solutions to our modelled problems.
 We highlight what remains to be taught and
look at areas of concentration.
 Most important analytical methods as
highlighted in the course outlines have been
PRELIMINARY considered, the remaining ones are to be
reviewed by the students.
 Some numerical methods that have not been
taught in the previous courses are presented
here.
 Students are advised to review those ones that
have bee covered in the pre-requisite courses.
 MATRIX METHODS
This method has been covered extensively
in the previous course (CHE 305)
SERIES SOLUTIONS
OTHER  Bessel and Legendre Equations with
ANALYTICAL variable coeffiecients
METHODS – THE
 Sturm-Loiville Boundary value problems
CHECKLIST
 USE OF OPERATORS IN THE SOLUTION
OF PDEs AND LINEAR INTEGRAL
EQUATIONS
These methods were adequately covered
in CHE 306. Students should review them
 CHECKLISTS
Fixed-point – Covered in CHE 306
Bi-section method – Covered in CHE 306
Gauss-Seidel – To be reviewed here
NUMERICAL Newton-Raphsons – Covered in CHE 306
METHODS – THE
Difference Operator – Covered in CHE
CHECKLIST
306
Forward, Central and Backward difference
– Covered in CHE 306
 Focus
After this Section, students should be able
to:
 solve a set of equations using the Jacobi
THE JACOBI AND and Gauss-Seidel methods,
GAUSS-SEIDEL
 recognize the advantages and pitfalls of the
ITERATIVE
Jacobi and Gauss-Seidel methods, and
METHODS
 determine under what conditions the the
methods always converge.
 Motivation -Why do we need another method
to solve a set of simultaneous linear
equations?
In certain cases, such as when a system of equations
is large, iterative methods of solving equations are
more advantageous.
Elimination methods, such as Gaussian elimination,
JACOBI AND are prone to large round-off errors for a large set of
SEIDEL METHODS equations.
Iterative methods, such as the Jacobi and Gauss-
Seidel methods, give the user control of the round-off
error.
Also, if the physics of the problem are well known,
initial guesses needed in iterative methods can be
made more judiciously leading to faster convergence.
 Two assumptions made on Jacobi Method:
The system given by

THE JACOBI (252)

METHOD

has a unique solution

The coefficient matrix A has no zeros on its
main diagonal, namely,
are nonzeros
 To begin, solve the 1st equation for x 1, the
2nd equation for x2 and so on to obtain the
rewritten equations:

THE JACOBI (253)

METHOD:
Then make an initial guess of the solution
MAIN IDEA

Substitute these values into the right hand

side the of the rewritten equations to
obtain the first approximation,
 This accomplishes one iteration.
In the same way, the second
THE JACOBI approximation, is
METHOD: computed by substituting the first
approximation’s x-values into the RHS of
MAIN IDEA the rewritten equations (253)
By repeated iterations, we form a sequence
of approximations
 The Algorithm
For each generate the components
of
from by
THE JACOBI
(254)
METHOD:

MAIN IDEA
Example:
Apply the Jacobi method to solve
(255)
 Continue iterations until two successive
approximations are identical when
rounded to three significant digits.
Solution
THE JACOBI To begin, rewrite the system
METHOD:

MAIN IDEA

Choose the initial guess, x1 = 0, x2 = 0,

x3 = 0
 The first approximation is

THE JACOBI
METHOD: Continue iteration, we obtain
n k=0 k=1 k=2 k=3 … k=6
MAIN IDEA
 Consider to solve an n x n size system of
linear equations with

THE JACOBI
METHOD: We split A into

IN MATRIX FORM
  is transformed into (D-L-U)x = b
THE JACOBI
METHOD: Given

IN MATRIX FORM

Then
 The matrix form of Jacobi iterative
method is

THE JACOBI
METHOD:
Jacobi iteration method can also be written
IN MATRIX FORM as
 Input: A =[aij], b, X0 = x(0), tolerance TOL,
maximum number of iterations N

THE JACOBI
METHOD:

NUMERICAL
ALGORITHM
THE JACOBI
METHOD:

NUMERICAL
ALGORITHM
 With the Jacobi method, the values of x i(k)
obtained in the kth iteration remain
unchanged until the entire (k+1)th iteration
has been calculated.
THE GAUSS-
With the Gauss-Seidel method, we use the
SEIDEL METHOD
new values of xi(k+1) as soon as they are
MAIN IDEA known.
For example, once we have computed of
x1(k+1) from the first equation, its value is
then used in the second equation to obtain
the new x2(k+1) and so on.
 Example. Derive iteration equations for
the Jacobi method and Gauss-Seidel
method to solve

THE GAUSS-
SEIDEL METHOD
  

MAIN IDEA
The Gauss-Seidel Method. For each k 0
generate the components of xi(k) from
xi(k-1) by
 Namely,

THE GAUSS-
SEIDEL METHOD

MAIN IDEA
 Gauss-Seidel method can be written as

THE GAUSS- Numerical Algorithm of Gauss-Seidel

SEIDEL METHOD
Method
NUMERICAL
ALGORITHM
THE GAUSS-
SEIDEL METHOD

NUMERICAL
ALGORITHM
 Let the iteration method be written as

THE JACOB AND

GAUSS-SEIDEL
METHODS If A is strictly diagonally dominant, then
for any choice of x(0), both the Jacobi and
CONVERGENCE Gauss-Seidel methods give sequences
THEOREMS OF that converges to the unique
THE ITERATION solution of .
METHODS
 The upward velocity of a rocket is given at
three different times in the following table
Table 1 Velocity vs. time data.
THE GAUSS-
SEIDEL METHODS

EXAMPLES

The velocity data is approximated by a

polynomial as
  Find the values of a1, a2 , and a3 using the
Gauss-Seidel method. Assume an initial guess
of the solution as

THE GAUSS-
SEIDEL METHODS

EXAMPLES
Solution
 The polynomial is going through three data
points (t1,v1), (t2,v2 ), and (t3,v3) where from the
above table
THE GAUSS- passes
SEIDEL METHODS through the three data points gives

EXAMPLES
 Substituting the data (t1,v1), (t2,v2 ), and
(t3,v3) gives

THE GAUSS-
SEIDEL METHODS
or
EXAMPLES
 The coefficients a1, a2, and a3 for the above
expression are given by

THE GAUSS-
SEIDEL METHODS Rewriting the equations gives

EXAMPLES
 Iteration #1
Given the initial guess of the solution
vector as

THE GAUSS-
SEIDEL METHODS

EXAMPLES we get

THE GAUSS-
SEIDEL METHODS

EXAMPLES The absolute relative approximate error for

each xi then is
THE GAUSS-
SEIDEL METHODS
At the end of the first iteration, the
EXAMPLES
estimate of the solution vector is
 and the maximum absolute relative
approximate error is 125.47%.
Iteration #2
The estimate of the solution vector at the
THE GAUSS- end of Iteration #1 is
SEIDEL METHODS

EXAMPLES
Now we get
THE GAUSS-
SEIDEL METHODS
The absolute relative approximate error for
EXAMPLES each xi then is
THE GAUSS-
SEIDEL METHODS

EXAMPLES At the end of the second iteration the

estimate of the solution vector is
 And the maximum absolute relative approximate
error is 85.695%.
Conducting more iterations gives the following
values for the solution vector and the
corresponding absolute relative approximate
THE GAUSS- errors.
SEIDEL METHODS

EXAMPLES

As seen in the above table, the solution estimates

are not converging to the true solution of
  The above system of equations does not seem to
THE GAUSS- converge. Why?
SEIDEL METHODS  Well, a pitfall of most iterative methods is that
they may or may not converge.
EXAMPLES  However, the solution to a certain classes of
systems of simultaneous equations does always
converge using the Gauss-Seidel method.
 This class of system of equations is where the
coefficient matrix [A] in [A][X ] = [C] is
diagonally dominant, that is
THE GAUSS-
If a system of equations has a coefficient matrix that
SEIDEL METHODS
is not diagonally dominant, it may or may not
converge.
EXAMPLES Fortunately, many physical systems that result in
simultaneous linear equations have a diagonally
dominant coefficient matrix, which then assures
convergence for iterative methods such as the
Gauss-Seidel method of solving simultaneous linear
equations.
 Class Work 7
 Find the solution to the following system of
equations using the Gauss-Seidel method

THE GAUSS-
SEIDEL METHODS

EXAMPLES Use

as the initial guess and conduct two iterations.