Bonfring International Journal of Man Machine Interface, Vol. 2, No.

2, June 2012 1
ISSN 2277 - 5064 | 2012 Bonfring
Abstract--- System of simultaneous linear equations plays
a vital role in mathematics, Operations Research, Statistics,
Physics, Engineering and Social Sciences etc. In many
applications at least some of the systems parameters and
measurements are represented by fuzzy numbers rather than
crisp numbers. Therefore it is imperative to develop
mathematical models and numerical procedures to solve such
a fuzzy linear system. The general model of a fuzzy linear
system whose coefficient matrix is crisp and the right hand
side column is an arbitrary fuzzy vector. In the fully fuzzy
linear system all the parameters are considered to be fuzzy
numbers. Since triangular fuzzy numbers is a special case of
trapezoidal fuzzy numbers, hence in this paper we considered
fully fuzzy linear system with trapezoidal fuzzy numbers. LU
decomposition method for a crisp matrix is well known in
solving linear system of equations. We discuss LU
decomposition of the coefficient matrix of the fully fuzzy linear
system, in which the coefficients are trapezoidal fuzzy
numbers.

Keywords--- Fully Fuzzy Linear System, LU
Decomposition, Trapezoidal Numbers
I. INTRODUCTION
HE concept of fuzzy numbers and fuzzy arithmetic
operations were first introduced by Zadeh. M. Fridamn,
M. Ming , M. Kandel [1] introduced a general model for
solving a fuzzy n x n linear system whose coefficient matrix is
crisp and the right hand side column is a fuzzy vector of
positive fuzzy numbers. M. Dehghan, B. Hashemi, M. Ghatee,
[3 ] are solved n x n fully fuzzy linear system using direct
method, Cramers rule, Gauss Elimination, Doolittle & Crout
factorization methods and Linear programming approach.
Amit Kumar, Neetu, Abhinav Bansal [8] discussed
consistency of the fully fuzzy linear system and the nature of
solutions. S.H. Neseri, M. Sohrabi, E. Ardil [5] Proposed
a method for solving fully fuzzy linear systems by certain
decomposition (LU) of the coefficient matrix with triangular

S. Radhakrishnan, Research Scholar, University of madras, Assistant
Professor, Department of Mathematics, D.G .Vaishnav College, Chennai,
India. E-mail:[email protected]
R. Sattanathan, Associate Professor & Head, Department of
Mathematics, D.G .Vaishnav College, Chennai, India.
P. Gajivaradhan, Associate Professor, Department of Mathematics,
Pachaiyapps College, Chennai, India.
fuzzy numbers. M. Mosleh, M. Otadi and A. Khanmirzaie
[6] introduced ST decomposition for 2x2 fully fuzzy linear
systems with triangular fuzzy numbers. P. SenthilKumar, G
Rajendran, [9] have solved n x n fully fuzzy linear system
using Cholesky method. Amit kumar, Abhinav Bansal, Neetu
[10] are bring in a new method for finding the non negative
solution of fully fuzzy linear system without any restriction on
the coefficient matrix. kumar, Neetu, Abhinav Bansal [11] are
put in a new method for finding the non negative solution of
the m x n fully fuzzy linear system without any restriction on
the coefficient matrix using Linear programming problem
method. Nasseri et al. [7] proposed Grevilles method to find
the positive solution of fully fuzzy linear system. T.
Allahyiranloo [2] proposed solution of a fuzzy linear system
by iterative methods such as Jacobi and Gauss Seidel methods.
A.K. Shyamal and M. Pal [4] discussed Triangular Fuzzy
matrices.
The structure of this paper is organized as follows
In Section II, we present some basic concepts of fuzzy set
theory and define a fully fuzzy linear system of equations. In
Section III, A numerical method for computing the solution of
Fully fuzzy linear system is discussed. Section IV deals with
a Numerical example to illustrate the above method. Section V
ends this paper with conclusion and References.
II. PRELIMINARIES
Definition 2.1: A fuzzy subset of R is defined by
its membership function : R [0,1], which
assigns a real number in the interval [0, 1], to
each element x R, Where the value of at x
shows the grade of membership of x in
Definition 2.2: A fuzzy number = (m, n, , ) is
said to be a trapezoidal fuzzy number if its
membership function is given by

(x) =

Definition 2.3: A fuzzy number is called positive
(negative), denoted by > 0 ( < 0), if its
membership function (x) satisfies (x) = 0, x
0 ( x 0 ).
LU Decomposition Method for Solving Fully
Fuzzy Linear System with Trapezoidal Fuzzy
Numbers
S. Radhakrishnan, R. Sattanathan and P. Gajivaradhan
T
Bonfring International Journal of Man Machine Interface, Vol. 2, No. 2, June 2012 2
ISSN 2277 - 5064 | 2012 Bonfring
Definition 2.4: Two fuzzy numbers = (m, n, , )
and = (p, q, , ) are said to be equal if and only if
m = p, n = q, = , and = .
Definition 2.5: A Trapezoidal fuzzy number = (m,
n, , ) is said to be zero trapezoidal fuzzy number if
and only if m= 0, n =0, =0, =0.
Definition 2.6: Let and
be two m X n and n X p fuzzy matrices. We
define Which is the m X p
matrix where


Definition 2.7: Arithmetic operations on trapezoidal
numbers
Let and be two
trapezoidal fuzzy numbers then

(i) =
= (m+p, n+q, + , + )
(ii) - = - =

(iii) 0 and o then
=
(mp, nq, m + p , n + q )

Definition 2.8: A matrix is called a fuzzy
matrix, if each element of is a fuzzy number. A
fuzzy matrix will be positive and denoted by
, if each element of be positive. We may
represent n X n fuzzy matrix
nXn
such that
, with the new notation
, where A =
are four n X n crisp
matrices.
Definition 2.9: A square fuzzy matrix will
be an upper triangular fuzzy matrix, if
i > j, and a square fuzzy matrix
will be a lower triangular fuzzy matrix, if
i < j.
Definition 2.10: Consider the n X n fuzzy linear
system of equations




..



The matrix form of the above equations is
Where the coefficient matrix , 1 i, j n is a n X n
fuzzy matrix and This system is called a fully
fuzzy linear system.
In this paper where = (A, B, M, N),
and = (b, g ,h, k) 0
We have
(A, B, M, N) (x, y, z, w) = (b, g, h, k)

Using 2.7 (iii) we have
(Ax, By, Az + Mx, Bw + Ny) = (b, g, h, k)

Using 2.4 we have
Ax = b
By = g
Az + Mx = h
Bw +Ny = k
By assuming that A and B are non-singular matrices we
have




III. A LU DECOMPOSITION METHOD FOR SOLVING FULLY
FUZZY LINEAR SYSTEMS FOR TRAPEZOIDAL FUZZY
NUMBERS
A. Theorem 3.1
Let A be an n X n matrix with all non-zero leading
principal minors. Then A has a unique factorization A = L U,
Where L is unit lower triangular and U is upper triangular.
Assume that , where A and B are the full
rank crisp matrices. Then if we let


Using 2.7 (iii) we have






Where matrices and are lower triangular crisp
matrices and matrices and are upper triangular crisp
matrices.
Consider the fully fuzzy linear systems.
We are going to construct a method for solving fully fuzzy
linear systems , Where ,
, = (b, g ,h, k), that is
(A, B, M, N) (x, y, z, w) = (b, g, h, k)


Using 2.7 (iii) we have

Bonfring International Journal of Man Machine Interface, Vol. 2, No. 2, June 2012 3
ISSN 2277 - 5064 | 2012 Bonfring


The current system by use of Definition2.4 can be rewrite
as follows





Therefore




B. Algorithm
(Fully Fuzzy Linear Systems Solver)
Step 1: Assume that Where A and B are
the full rank crisp matrices.
Compute LU- decomposition for crisp matrix
and
Step 2: Set
Step 3: Compute
Step 4: Compute
Step 5: Compute the solution of the fully fuzzy system
(A,B,M,N) (x,y,z,w) =(b,g,h,k) as follows




IV. NUMERICAL EXAMPLE
Consider the following fully fuzzy linear system
(3,6,2,2) (4,6,1,2) =
(27,66,26, 58)
(4,5,1,1) (5,8,1,2) =
(35,70,25,55)

Solution: The given fully fuzzy linear system can be
written as
=

, , ,
, , ,








and
V. CONCLUSION
In this paper, the solution of fully fuzzy linear system is
obtained by LU decomposition method in the form of
trapezoidal fuzzy number matrices. We assumed
because, it will reduce the computational work and our aim is
to find a lower triangular structure for Crisp matrices for A
and B not for M and N [3].
REFERENCES
[1] M. Fridamn, M. Ming , M. Kandel, Fuzzy linear systems, fuzzy sets and
systems 96, 1998 201-209.
[2] T. Allahviranloo, Numerical methods for fuzzy system of linear
equations, Applied mathematics and Computation, Vol,155, 2004, 493-
502.
[3] M. Dehghan, B. Hashemi, M. Ghatee, computational methods for
solving fully fuzzy linear systems, Applied Mathematics and
Computation, 179, 2006, 328-343.
[4] A.K. Shyamal and M. Pal, Triangular Fuzzy matrices, Iranian Journal of
Fuzzy systems.Vol.4, No.1,2007, 75-87.
[5] S.H. Nasseri, M. Sohrabi, E. Ardil, Solving fully fuzzy linear systems by
use of a Certain Decomposition of the Coefficient Matrix, World
Academy of Science, Engineering and Technology, 43, 2008, 787-789.
[6] M. Mosleh, M.Otadi, A. Khanmirzaie, Decomposition Method for
solving fully fuzzy linear systems, Iranian Journal of Optimization 1,
2009, 188-198.
[7] S.H. Nasseri, M. Matinfar, Z. Kheiri, Grevilles Method for the fully
fuzzy linear system of equations, Advances in fuzzy set and systems, 4
,2009, 301-311.
[8] Amit kumar, Neetu, Abhinav Bansal, A new method to solve fully fuzzy
linear system with trapezoidal fuzzy numbers, Canadian Journal on
Science and Engineering Mathematics Vol.1, No.3, August 2010, 45-56.
[9] P. SenthilKumar, G Rajendran, New approach to solve symmetric fully
fuzzy linear systems, Indian academy of sciences, Sadhana, Vol.36,
Part 6, December 2011, 933-940.
[10] Amit kumar, Abhinav Bansal, Neetu, Solution of fully fuzzy linear
system with arbitrary coefficients, International journal of Applied
Mathematics and Computation, Vol.3(3), 2011, 232-237.
[11] Amit kumar, Neetu, Abhinav Bansal, A new approach for solving fully
fuzzy linear systems.