International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-9, September- 2016]

Solving Complex Fuzzy Linear System of Equations by using

QR-Decomposition Method
Ke Zhang1, Xiaobin Guo2*
1,2

College of Mathematics and Statistics, Northwest Normal University, Lanzhou 730070, China
The work is supported by the Natural Scientific Funds of PR China (no:61262022).
*Corresponding Author. Tel.: 0086 13919030082.
Email: [email protected] ; [email protected]

Abstract In this paper, QR-decomposition method for solving the complex fuzzy linear equation Cz w in which C is
a crisp complex matrix and w is an arbitrary complex fuzzy vector is considered. Some examples are given to illustrate the
proposed method.
Keywords Complex fuzzy numbers, Fuzzy linear systems, Fuzzy approximate solutions, QR-decomposition.

I.

INTRODUCTION

Many real-world engineering systems are too complex to be defined in precise terms, imprecision is often involved in any
engineering design process. Fuzzy systems have an essential role in this fuzzy modeling, which can formulate uncertainty in
actual environment. In many linear systems, some of the system parameters are vague or imprecise, and fuzzy mathematics is
a better tool than crisp mathematics for modeling these problems, and hence solving a fuzzy linear system [11] or a fuzzy
differential equation is becoming more important. The concept of fuzzy numbers and arithmetic operations with these
numbers were first introduced and investigated by Zadeh [22]. A different approach to fuzzy numbers and the structure of
fuzzy number spaces was given by Puri and Ralescu [16], Goetschell and Voxman [12] and Wu and Ma Ming [20].
Since Friedman et al. proposed a general model for solving an n n fuzzy linear systems whose coefficients matrix is crisp
and the right-hand side is an arbitrary fuzzy numbers vector by an embedding approach, many works have been done about
how to deal with some advanced fuzzy linear systems such as dual fuzzy linear systems (DFLS), general fuzzy linear systems
(GFLS), full fuzzy linear systems (FFLS), dual full fuzzy linear systems (DFFLS) and general dual fuzzy linear systems
(GDFLS). These works were performed mainly by Allahviranloo et al. [2,3,4,5], Abbasbandy et al. [1,7], Zheng et al. [1,9]
and Dehgham et al. [10] and so on. In traditional fuzzy linear systems, the uncertain elements were usually denoted by the
parametric form of fuzzy numbers. Based on arithmetic operations of the fuzzy number, the fuzzy linear systems could be
extended into crisp function linear systems. Therefore the solutions of the fuzzy linear systems can be obtained by solving
the model by means of ordinary analytical and numerical methods.
However, very few researchers have developed methods to solve fuzzy complex system of linear equations. The fuzzy
complex numbers was introduced by J.J.Buckley in 1989 [9]. The n n fuzzy complex linear systems have been studied
by M.A. Jahanigh [15]. Solution of fuzzy complex linear system of linear equations was described by Rahgooy et al. [18] and
applied to circuit analysis problem. In 2014, Behera and Chakraverty [8] discussed the fuzzy complex system of linear
equations by adding and subtracting the left and right bounds of the fuzzy complex unknowns and the right-hand side is
fuzzy complex vector.
In this paper complex fuzzy linear system is investigated. A numerical procedure for calculating the fuzzy solution is
designed. Finally, some examples are given to illustrate our method. The structure of this paper is organized as follows:
In Section 2, we recall the triangle fuzzy number and present the concept of the complex fuzzy linear systems. The
computing model to the complex fuzzy linear systems is proposed in detail and the fuzzy approximate solution of the
complex fuzzy linear systems is obtained by solving the crisp systems of linear equations using QR- decomposition method.
Some examples are given to illustrate our method in Section 4 and the conclusion is drawn in Section 5.

II.

PRELIMINARIES

There are several definitions for the concept of fuzzy numbers.

2.1

The fuzzy number

Page | 54

International Journal of Engineering Research & Science (IJOER)

Definition 2.1. A fuzzy number is a fuzzy set like

ISSN: [2395-6992]

u : R I 0,1 which satisfies:

(1)

is upper semi-continuous,

(2)

is fuzzy convex, i.e., u ( x 1 y ) min{u ( x), u ( y)} for all

(3)

u is normal, i.e., there exists x0 R such that u ( x0 ) 1 ,

(4)

sup pu {x R u ( x) 0} is the support of the

Let

[Vol-2, Issue-9, September- 2016]

x, y R, 0,1 ,

u , and its closure cl(supp u ) is compact.

E1 be the set of all fuzzy numbers on R .

Definition 2.2. A fuzzy number u in parametric form is a pair

u, u of functions u r , u r , 0 r 1 ;which satisfies

the requirements:
(1)

u r is a bounded monotonic increasing left continuous function,

(2)

u r is a bounded monotonic decreasing left continuous function,

(3)

u r u r , 0 r 1.

A crisp number x is simply represented by (u r , u r ) ( x, x ), 0 r 1 . By appropriate definitions the fuzzy number

space {u r , u r } becomes a convex cone

E1 which could be embedded isomorphically and isometrically into a Banach

space.
Definition 2.3 Let x

x r , x r , y y r , y r ,0 r 1, and k R . Then

(1)

x y iff x r y r , x r y r ,

(2)

x y x r y r , x r y r ,

(3)

x y x r y r , x r y r ,

(4)

k x, k x , k 0

kx
,
k
x
,
k
x
,
k

Definition

2.4

An

arbitrary

fuzzy

X p r , p r i q r , q r .
p p r , p r and q q r , q r

complex
for all

number

may

be

represented

as

X p iq ,

where

0 r 1 are two real fuzzy number. Hence one may have

Definition 2.5 For any two arbitrary complex fuzzy numbers

x p iq, y u iv where p, q, u, v are fuzzy numbers,

their arithmetic is as follows:

(1)

z1 z2 x1 x2 i y1 y2 ;

(2)

kx kp ikq, k R;
Page | 55

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

(3)

z1 z2 x1 x2 y1 y2 i x1 y2 y1 x2 .

2.2

Complex fuzzy linear systems

[Vol-2, Issue-9, September- 2016]

Definition 2.6. The linear system equation

c11 z1 c12 z2 ... c1n zn w1

c z c z ... c z w
21 1 21 2
2n n
2

...

cn1 z1 cn 2 z2 ... cnn zn wn

where

(2.1)

cij ,1 i, j n are complex numbers and wi ,1 i, j n are complex fuzzy numbers, is called a complex fuzzy

linear system(CFLS).
Using matrix notation, we have

Cz w

(2.2)

A complex fuzzy numbers vector

z z1 , z2 ,..., zn

is called a fuzzy solution of the complex fuzzy linear system (2.1) if

III.
3.1

z satisfies (2.2).

SOLVING COMPLEX FUZZY LINEAR SYSTEM OF EQUATION

Equivalent fuzzy linear system

Theorem3.1. The n n complex fuzzy linear system (2.1) is equivalent to a

2n 2n order fuzzy linear system

Gx b

(3.1 )

where

A B
p
u
G
, x ,b
B A
q
v
Proof. We denote C A iB, A, B R
unknown vector
Since

nn

(3.2)

and w u iv where u , v are fuzzy number vectors. We also suppose the

z p iq where p, q are two unknown fuzzy number vectors.

Cz w , we have
( A iB )( p iq ) u iv

i.e.

Ap Bq i( Aq Bp) u iv .
Comparing with the coefficient of i , we have

Ap Bq u

Aq Bp v
i.e.

A B p u ,

B A q v
Page | 56

International Journal of Engineering Research & Science (IJOER)

It admits a

ISSN: [2395-6992]

[Vol-2, Issue-9, September- 2016]

2n order fuzzy linear system.

We express it in matrix form as follow:

Gx b
3.2

The model

When z , w of fuzzy linear equation (2.3) are denoted by the parametric form i.e.

w [ w1 , w2 ,..., wn ]T , w j u j iv j [u j r , u j r ] i[v j r , v j r ],

z [ z1 , z2 ,..., zn ]T , z j p j iq j [ p j r , p j r ] i[q j r , q j r ], j 1, 2,..., n, 0 r 1.

Some results for solving fuzzy linear equation (3.2) are obtained by the following analysis.
Theorem 3.2. The fuzzy linear equation (3.2) can be extended to a crisp function linear system as follow

SX r Y r

(3.3)

Where

G
S
G
in which the elements

x r
br
G
,
X
(
r
)

,
Y
r

x r
G

b r

(3.4)

gij of matrix G and gij of matrix G are determined by the following way: if

gij 0, gij gij else gij 0,1 i, j n; if gij 0, gij gij else gij 0,1 i, j n.
Proof. Let b [b r , b r ], 0 r 1 and

matrix G and

x [ x(r ), x(r )]. Suppose G G G in which the elements gij of

gij of matrix G are determined by the following way: if gij 0, gij gij else gij 0,1 i, j n; if

gij 0, gij gij else gij 0,1 i, j n. The fuzzy linear equation Gx b can be expressed as
(G G )[ x r , x r ] [b r , b r ].

k x j r , k x j r , k 0

kx j
(k x j r , k x j r ), k 0

Since

Gx r , Gx r , G 0

Gx
Gx r , Gx r , G 0.

We have

So the equation (3.4) be rewritten as

G [ x r , x r ] G [ x r , x r ] [G x r , G x r ] [G x r , G x r ],
[G x r G x r , G x r G x r ] [b r , b r ].
Thus we have

Page | 57

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-9, September- 2016]

G x r G x r b r

G x r G x r b r

G x r G ( x r ) b r

G ( x r ) G x r b r

or

Expressing in matrix form , we have

G

G

Cz w , then z z and z z are the solutions of complex fuzzy linear system, i.e.,

Remark 3.1 If

G x r b r

G x r b r

C z z w w and C z z w w .
Remark 3.2 If

Cz w , then z z is the solution of complex fuzzy linear system, i.e., C z z w w and

C z z w w .
In order to solve the fuzzy system of linear equation (3.2), we need to consider the systems of linear equations (3.4). It seems
that we have obtained the minimal solution of the fuzzy linear system (3.2) as

X r S Y r
(3.5)

x r G

x
r

i.e.,

where S is the Moore-Penrose generalized inverse of matrix

Definition 3.1. Let
vector Z [ p

br

b
r

(3.6)

S.

X r x j r , x j r ,1 j 2n denotes the minimal solution of (2.5). The fuzzy number

r , p j r ] i[q j r , q j r ],1

j n defined by

p j r min{x j r , x j r , x j 1 , x j 1},
p j r max{x j r , x j r , x j 1 , x j 1},

(3.7)

j 1, 2,..., n, 0 r 1,
q j r min{x j r , x j r , x j 1 , x j 1},
q j r max{x j r , x j r , x j 1 , x j 1},

(3.8)

j n 1, n 2,..., 2n, 0 r 1.
is called the fuzzy minimal solution of the fuzzy linear systems (2.3). If p
then z {[ p

r , p j r ] i[q j r , q j r ],1

complex fuzzy linear systems (2.1) . Otherwise,

r , q j r ,1

j 2n are all fuzzy numbers

j n} is called a strong complex fuzzy minimal solution of the

z is called weak complex fuzzy minimal solution.

Page | 58

International Journal of Engineering Research & Science (IJOER)

3.3

ISSN: [2395-6992]

[Vol-2, Issue-9, September- 2016]

QR-decomposition method

Theorem 3.3 If eq.(3.4)

S m n , then there exist an orthogonal Q m n and an upper triangular R m n so that

S QR .
Proof. Suppose

n 1 and that Q is Householder matrix so that if R QT S then R 2 : m 0 .It follow that S QR

is a QR factorization of

S . For general n we partition S ,

S [ S1 v]

Where

v S :, n .

By induction, there exists an orthogonal

Q1 Rmm so that R1 Q1T S1 is upper triangular. Set w QT v and let

w n : m Q2 R2 be the QR factorization of w n : m .If

0
I
Q Q1 n 1

0 Q2
A Q R1

Then
is a QR factorization of

w 1: n 1
R2

S.

Lemma 3.1 If A is an m n matrix with full column rank, then A can be factored as A QR , where Q is an m n
matrix whose column vectors form an orthonormal basis for the column space of A and R is a n n invertible upper
triangular matrix.
Theorem 3.4. If Eq.(2.5)
the system for

S is a matrix with full column rank , and if S QR is a QR decomposition of A , then

SX Y can be expressed as
RX QT Y

(3.11)

X R 1QT Y

(3.12)

and the solution can be expressed as

Proof. According to Lemma 3.1., we have

S QR
Where Q is

4m 4n orthogonal matrix , R is a 4n 4n invertible upper triangular matrix . The fact that Q has

orthonramal column implies that Q Q I , so multiplying both side of S QR by on the left side
T

R QT S
Therefore, the system for

(3.13)

SX Y can be expressed as
RX QT Y .

In order to obtain the solution, due to R is invertible upper triangular matrix, so

X R 1QT Y .
Page | 59

International Journal of Engineering Research & Science (IJOER)

3.4

ISSN: [2395-6992]

[Vol-2, Issue-9, September- 2016]

A sufficient condition for strong fuzzy solution

The key point to make the solution matrix being a strong fuzzy solution is that

S Y r is fuzzy matrix , i.e., each element

in which is a triangular fuzzy number. By the analysis, it is equivalent to the condition S 0 .

Theorem 3.5. If

( G G (G G ) ) 0,( G G (G G ) ) 0,

the complex fuzzy linear equation(2.1) has a strong complex fuzzy minimal solution as follows:

z [ p j r , p j r ] i[q j r , q j r ], 0 r 1

(3.10)

p r
x r
Eb r Fb r
q r

p r

x
r

Fb r Eb r

q r

E 1 ((G G ) (G G ) )

1
F ((G G ) (G G ) )

Where

Proof. The proof is straight forward

IV.

NUMERICAL EXAMPLE

In this section, we will demonstrate the efficiency and superiority of the proposed method using numerical examples.
Example 4.1 Consider the following CFSLE:

~ ~
z1 z2 (r , 2 r ) i (1 r ,3 r )
~
~
z1 3 z2 (4 r , 7 2r ) i (r 4, 1 2r )
~

__

__

Let z1 p1 i q1 ( p1 , q1 ) i (q1 , q1 ),

0
3
1
0
0
0
0
0

0
0
1
1
0
0
0
0

(4.1)

z2 p2 i q2 ( p2 , q2 ) i (q2 , q2 ). The extend maxtrix is

1
0

0
S
0

0
0

1
0
0
3
0
0
0
0

0
0
0
0
1
1
0
0

0
0
0
0
0
3
1
0

0
0
0
0
0
0
1
1

0
0

0
1

0
0

And Rank(S)=8,so we obtain

Page | 60

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-9, September- 2016]

0
0
0
0
0.7071 0.6396 0.2023 0.2236

0
0
0
0
0.7071 0.6396 0.2023 0.2236
0
0.4264 0.6068 0.6708
0
0
0
0

0
0
0.7416 0.6708
0
0
0
0

Q
0
0
0
0
0.7071 0.6396 0.2023 0.2236

0
0
0
0.7071 0.6396 0.2023 0.2236
0
0
0
0
0
0
0.4264 0.6068 0.6708

0
0
0
0
0
0.7416 0.6708
0
0
0.7071
0
0
0
0
1.4142 2.1213

2.3452 0.4264 0.6396

0
0
0
0
0
0
0
1.3484 2.4271
0
0
0
0

0
0
0
1.7889
0
0
0
0

R
0
0
0
0
1.4142 2.1213
0
0.7071

0
0
0
0
2.3452 0.4264 0.6396
0
0
0
0
0
0
0
1.3484 2.4271

0
0
0
0
0
0
1.7889
0
Now by using (3.2),we have

p1
r 1.375 0.625r

p2
4 r 0.875 0.125r

1 r 0.125 0.625r
q1

q
2 R 1QT r 4 1.375 0.125r .
r 2 0.2875 0.875r
p
1

2
r

1.375

0.375
r

p2
r 3 1.625 0.875r

q1

2r 1 0.875 0.375r
q2
The solution may be written as

~
z1 (1.375 0.125r , 2.875 0.875r ) i(0.125 0.625r,1.625 0.875r )
~ (0.875 0.125r ,1.375 0.375r ) i( 1.375 0.125r, 0.875 0.375r )
z2
Example 4.2 Conside a simple RLC circuit with fuzzy current and fuzzy sourse in Figure 1.The CFSLE for the circuit is as
follow

Page | 61

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-9, September- 2016]

~
~

(10 7.5i) z1 (6 5i) z2 (4 r , 6 r ) i(1 r ,1 r )

~
~
(6 5i) z1 (16 3i) z2 (2 r , r ) i(3 r , 1 r )

Let

(4.2)

z1 p1 i q1 ( p1 , p1 ) i(q1 , q1 ), z2 p2 i q2 ( p2 , q2 ) i (q2 , q2 ). The extend matrix is

0
6
0
5
10 0 7.5 0

0
6
0
5
3
0 16 0
0
5 10 0 7.5 0
0
6

5
3
0 16 0
0
6
0

S
.
0
6
0
5 10 0 7.5 0

0
5
3
0 16 0
0
6
7.5 0
0
6
0
5 10 0

0
6
0
5
3
0 16
0

And we obtain

0.6785

0.3392
Q

0.4071
0.5088

14.7394

0
R

0.0383
0.8876
0.2774
0.1473
0.3382
0.0230
0.0287
0
1.0177
18.0268
0
0
0
0
0
0

0.2174
0.1659
0.7365
0.2181
0.0622
0.1699
0.2804
4730
7.1238
2.3715
12.0838
0
0
0
0
0

0.3857
0.3010
0.1883
0.7970
0.2472
0.0013
0.0160
0.1694
9.7019
3.7792
4.9730
13.8874
0
0
0
0

0.0239
0.2183
0.0879
0.3458
0.8283
0.1107
0.2873
0.2090

0.3954
0.0736
0.2058
0.1555
0.0528
0.8671
0.0628
0.1146

0
10.7340
6.2713
2.9250
7.3583
0
0
0

0.4249
0.0590
0.2931
0.1656
0.3188
0.1014
0.7580
0.1272

13.1281
0.7411
4.0395
1.9066
0.4347
11.5309
0
0

0.0629

0.1938
04573

0.0966
0.1839

0.1747
0.0169

0.8213

7.1238 3.3923

7.5305 4.1354
5.4090 12.5773

4.9705 1.0086
5.9185 2.2809

0.7967
1.1572
4.4908 1.8456

0
11.3242
Page | 62

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-9, September- 2016]

Now by using (3.2),we have

p1

p2

q1
q
2 R 1QT
p
1

p2

q1

q2

4 r 0.3164 0.0378r

2 r 0.0347 0.0307 r
1 r 0.0708 0.0378r

3 r 0.2380 0.0307 r .
r 6 0.3920 0.0378r

r 0.0961 0.0307 r
r 1 0.1464 0.0378r

r 1 0.1765 0.0307 r

The solution may be written as

~
z1 (0.3164 0.0378r , 0.3920 0.0378r ) i(0.0708 0.0378r , 0.1464 0.0378r )
~ (0.0347 0.0307r , 0.0961 0.0307r ) i (0.2380 0.0307r , 0.1765 0.0307r )
z2
V.

CONCLUSION

In this work we presented a model for solving complex fuzzy linear equation Cz w where C is a crisp complex matrix
and w is an arbitrary complex fuzzy vector, respectively. The complex fuzzy linear system is converted to a high order
linear system SX r Y r . We use the QR-decomposition of the coefficient matrix

S to obtain fuzzy solution of

complex fuzzy linear systems. In addition, numerical examples showed that our method is feasible to solve this type of
complex fuzzy linear systems.

REFERENCES
[1] S. Abbasbandy, R. Ezzati, A. Jafarian, LU decomposition method for solving fuzzy system of linear equations, Applied Mathematics
and Computation, 172 (2006) 633-643.
[2] T. Allahviranloo, A comment on fuzzy linear systems, Fuzzy Sets and Systems, 140 (2003) 559-560.
[3] T. Allahviranloo, Numerical methods for fuzzy system of linear equations, Applied Mathematics and Computation, 153 (2004) 493502.
[4] T. Allahviranloo, Successive over relaxation iterative method for fuzzy system of linear equations, Applied Mathematics and
Computation, 162 (2005) 189-196.
[5] T. Allahviranloo, The adomian decomposition method for fuzzy system of linear equations, Applied Mathematics and Computation,
163 (2005) 553-563.
[6] H. Anton, R.C.Busby, Comtemporary Linear Algebra, John Wiley, 2003.
[7] B. Asady, S. Abbasbandy, M. Alavi, Fuzzy general linear systems, Applied Mathematics and Computation, 169 (2005) 34-40.
[8] D. Behera, S. Chakraverty, Solving fuzzy complex system of linear equtions, Information Sciences, 277(2014) 154-162.
[9] J.J. Buckley, Fuzzy complex number, Fuzzy Sets Syst. 33 (1989) 333-345.
[10] M. Dehghan, B. Hashemi, Iterative solution of fuzzy linear systems, Applied Mathematics and Computation, 175 (2006) 645-674.
[11] M. Friedman, M. Ma, A. Kandel, Fuzzy linear systems, Fuzzy Sets and Systems, 96 (1998) 201-209.
[12] R. Goetschel, W. Voxman, Elementary calculus, Fuzzy Sets and Systems, 18 (1986) 31-43.
[13] X.B. Guo, Z.T. Gong, Block Gaussian elimination methods for fuzzy matrix equations, International Journal of Pure and Applied
Mathematics, 58 (2008) 157-168.
[14] M.A. Jahantigh, S. Khezerloo, M. Khezerloo, Complex fuzzy linear systems, Int.J. Indust. Math. 2 (2010) 21C28.
[15] M.L. Puri, D.A. Ralescu, Di_erentials for fuzzy functions, Journal of Mathematics Analyisis and Application, 91 (1983) 552-558.
[16] J. Qiu, C. Wu, F. Li, On the restudy of fuzzy complex analysis:Part I.The sequence and series of fuzzy complex numbers and their
convergences, Fuzzy Sets Syst. 115 (2000) 445-450.
[17] T. Rahgooy, H.S.Yazdi, R. Monse_, Fuzzy complex system of linear equations applied to ciccuit analysisInt.J.Comput.Electr.Eng.
1 (2009) 1793-8163.
[18] K. Wang, B. Zheng, Inconsistent fuzzy linear systems, Applied Mathematics and Computation, 181 (2006) 973-981.
[19] C.X. Wu, M. Ma, Embedding problem of fuzzy number space: Part I, Fuzzy Sets and Systems, 44 (1991) 33-38.
[20] L.A. Zadeh, Fuzzy sets, Information and Control, 8 (1965) 338-353.

Page | 63