Illustration of Gauss – Seidel Method Using MATLAB

V.V MEHTRE* SIDDHARTH ROY**

. * Assistant Professor, Department of Student, Department of Electrical
Electrical Engineering, Engineering
Bharati Vidyapeeth University,Pune, BharatiVidyapeeth University, Pune,
India India.
([email protected]) ([email protected])

Abstract :- Liebmann method and this method is similar

Considering the case of numerical algebra, to the jacobbi method. Numerical solution of
the answer to system of linear equation is system of linear equation using gauss seidel
an assignment of ideals of variables such method is given ahead.
that entire equations in the systems are Keywords:-
concurrently fulfilled. Gauss– Seidel method System of linear equations, Gauss-Seidel
is an iterative method to answer a system of Method, MATLAB solution, Theoretical
linear equations. It is also named as solution, Applications.
INTRODUCTION:-
The direct methods of solving linear the main memory of the computer in order
equations are recognized to have their to capably solve the linear system.
difficulty. For example the problem with The problem of continuous mathematics are
Gauss elimination approach lies in control of solved numerically with the help of
the accrual of rounding errors Turner. The algorithms which are created, analyzed and
Systems of linear equations ensue in bulky implemented by the numerical analysis
number of areas both directly in physical which comes under the area of mathematics
situations and ultimately in the numerical and computer science. These problems are
solutions of the other mathematics model. originated from real world applications (i.e.
These applications happen in nearly all areas algebra, geometry and calculus) are involved
(i.e. physical, biological and social science). of continuously varying variables. Direct
Linear systems are in the arithmetical methods or iterative methods are one of the
solution of many problems (Such as: - two categories by which the solution of
optimization system of non linear equations linear equations can be accomplished. In
and partial differential equations). The most general we should prefer a direct method for
ordinary type of problem is to explain a the solution of liner equations. But, when it
square linear system is [YX = Z] of comes to matrices with a large number of
moderate order with coefficient that are zero elements we should use iterative
generally non zero, such linear system of methods which presents these elements are
any order are called dense since the advantageous. Round off errors are
coefficient matrix A is generally stored in controlled by the Gauss seidel method. In
short gauss seidel method is the modification of Jacobi’s iteration method.

PROBLEM FORMULATION:- YX=Z –––––––––––––(2)

Let us deem a system of ‘n’ linear equations Y= y11 y12 y13 ––––––– y1n
with ‘n’ unknowns:- y21y22 y23 ––––––– y2n
Let us deem a system of ‘n’ linear equations In general,
with ‘n’ unknowns:-
yn1 yn2 yn3 ––––––– ynn
y11x1 + y12x2 + y13x3 …………… +
y1nxn = z1
X= x1 And, B= z1
y21x1 + y22x2 + y23x3 …………… +
y2nxn =z2 x2 z2
In general, In general,

yn1x1 + yn2 x2 + yn3x3 …………… + x3 zn

ynnxn= zn
The matrix can also be represented as:-
In this the dioagonal element Yij doesn’t So, the system of equation (1) can be
vanish. If so, the equation should be rewritten as:-
rearranged so as to satisfy the condition.
X1 = z1 – (y12x1 + y13x3 + ……….y1nxn) diagonally governing or symmetric positive
Y11 definite. Also: If the physics of the difficulty
are understood, a shut initial guess can be
X2 = z2 – (y21x1 + y23x3 + ……….y2nxn)
assumed, declining the number of iterations
Y22 required.
In general, MATLAB SOLUTION:-
X1 = zn – (yn1 x1 + yn2x3 + ……….ynn-1xn-1) Let us now find the roots of subsequent
Ynn simultaneous equations by the Gauss-Seidel
method.
Now, if there was an availability of all the
20x+y-2z=17 ----------(1)
unknown guesses i.e.
X 1,X 2,X 3……….X n. The value could be
0 0 0 0 3x+20y-z=-18----------(1)
substituted in the equation (3) and the 2x-3y+20z=25---------(1)
updated guess could be computed (i.e.
X01,X02,X03……….X0n). There are several Result:-
ways to accomplish this depending on the Let’s find the particular equation in the form
most recently computed guess. as:
ITERATIVE METHODS:- A = (1/20) * (17-y+2z) ---------- (2)
These methods are used to get more B = (1/20) * (-18-3x+z) --------- (2)
specified accuracy in finding of roots for the C = (1/20) * (25-2x+3y) -------- (2)
system of linear equations. This We create approximation by A0=B0=C0=0.
phenomenon of sequence construction is Substituting B=B0 and C=C0 in right hand
known as iteration. Iterative methods starts side of first of equation (2)
with an initial approximation and then goes A1 = 17/20=0.85
on to infinite sequence whose limit is the
In second of equation second put A=A1 and
exact solution. C=C0
GAUSS SEIDEL METHOD:- B1 = (1/20) * (-18-3*0.85) = -1.0275
Gauss seidel method is an iterative process Put A=A1, B=B1 in third of equation (2)
used to solve system of n linear equations C1 = (1/20) * (25-2*0.85+3*-1.0275) =
with x unknowns of a square matrix. The 1.011
Gauss-Seidel Method permit the addict to likewise we get,
manage round-off mistake. In fact, iterative A2=1.002 B2 =-0.9998 C2 =0.9998
methods can be used to recover the
A3 =1.0000 B3 =-1.0000 C3 =1.0000
explanation we get by the direct methods.
A4 =1.0000 B4 =-1.0000 C4 =1.0000
The majority of elimination methods are
accountable to bear on or after round-off Now it is ample. If we watch over two sets
errors. This method is alteration of the of roots, they are approximately same.
Jacobi’s iteration method. It is defined on Hence, roots of given instantaneous
equations using Gauss-Seidel Method are:
matrices with non-zero diagonals, but
meeting is only sure if the matrix is either A=1; B = -1; C=1
THEORETICAL SOLUTION the system of thermal-radiation transfer
Theory solution for: equations for radiating, gripping and
20x+y-2z=17 ----------(1) scattering media. computation of a analysis
3x+20y-z=-18----------(1) setback are preformed for the instance of
the alteration shape of the nonlinear variant
2x-3y+20z=25---------(1)
of the method for the finite-difference
Here, the equation is already arranged
WDD scheme in planar geometry. The
So, 1st iteration: (k=0) method is taken for thetransport of γ-rays
X11 = (1/20)*[17-(1*0)-(-2*0)] and neutrons.
=0.85
X2 = (1/20)*[-18-(3*X11)-(-1*0)]
1
CONCLUSION:
= -1.0275
Several methods can be used to solve system
X3 = (1/20)*[25-(2*X11)-(-3*X21)]
1
of linear equations some of them are direct
= 1.010875 methods and some are numerical method
nd
2 iteration: (k=1) which are also termed as iterative method. In
X12 = (1/20)*[17-(1*X21)-(-2* X31)] this review, two iterative methods for
= 1.0024625 answering system of linear equations have
X2 = (1/20)*[-18-(3*X12)-(-1* X31)]
2 been offered in which the Gauss-Seidel
method proves to be actually fast. By the
= -0.999825625
practical example, it was seen that answers
X3 = (1/20)*[25-(2*X12)-(-3* X22)]
2
were obtained within minute iteration.
= 0.999779906
ACKNOWLEDGEMENT:
Similarly,
We would like to express our special thanks
3rd X13= X23 = - X33 =
iteration 0.999969 1.000006 1.000002 of gratitude to Dr. D.S. Bankar, Head,
(k=2) 271 395 114 Department would also like to thank the
4rd X14= X24 = - X34 = faculty members of the department of
iteration( 1.000000 0.999999 0.999999 electrical engineering who helped us with
k=3) 531 974 95 extended support.
5rd X15= X25 = - X35 = REFERENCE:
iteration( 0.999999 1.000000 1.000000
[1] F.MITENKOV, E.SOLOVEY “Analysis
k=4) 993 001 001
of an open fuel cycle for BN–800”.
6rd X16= X26 = - X36 = At.Energ, 113,No.2 73–79.(2012).Cross ref
iteration( 1.000000 1.000000 1.000000
k=5) 00 00 00 google scholar.
[2] A.L.GENGAJE(2016). “Matlab”
Research in engineerfing and technology.
APPLICATIONS
The relevance of the Gauss–Seidel method
is examine to obtain an iterative result of
[3] Attaway, Stormy. Matlab – “Programing York, NY : Springer, 2011. ISBN 978-1-
and difficulty solving”. Canada Inc., 2009. 4757-7394-1.

[4] Gilat, Amos. “MATLAB An [7] Kaw, Autar K, Kalu, Egwalu E and
Introduction with Applications”. Columbia, Nguyen, Duc. “Numerical Methods with
Ohio : John Wiley & Sons, Inc., 2011. Applications: Abridged”. Florida :
University of South Florida, 2011.
[5] “Gauss-Seidel Method and Inaccuracy
Decrease Factor”. H Kaur, Khushpreet Kaur. [8] M.K. Jain, S.R.K. Iyengar, R.K. Jain.
“Numerical methods for Scientific and
[6] Saleri, Alfio Quarteroni Riccardo Sacco Engineering computaion".
Fausto. “Numerical Mathematics”. New