Fuzzy Arithmetic

Fuzzy Numbers:
To qualify as a fuzzy number, a fuzzy set A on R must
possess at least the following three properties:
 A must be a normal fuzzy set;
  A must be a closed interval for every   (0,1]
 The support of A, 0 A must be bounded.
The fuzzy set must be normal since our conception of a
set of “real numbers close to r” is fully satisfied by r
itself; hence, the membership grade of r in any fuzzy
set that attempts to capture this conception must be
equal to 1.
Fuzzy Numbers contd.
The bounded support of a fuzzy number and all its  -
cuts for   0 must be closed intervals to allow us to
define meaningful arithmetic operations on fuzzy
numbers in terms of standard arithmetic operations on
closed on closed intervals.
Since  -cuts of any fuzzy number are required to be
closed intervals for all   (0,1] , every fuzzy number is a
convex fuzzy set. The inverse, however, is not necessarily
true, since  -cuts of some convex fuzzy sets may be
open or half-open intervals.
Special cases of fuzzy numbers, ordinary real numbers
and intervals of real numbers are illustrated in Fig. 4.1
Fuzzy Numbers contd.
Using fuzzy numbers , we can define the concept of a
fuzzy cardinality for fuzzy sets that are defined on finite
universal set s. Given a fuzzy set A defined on a finite
universal set X, its fuzzy cardinality, A , is a fuzzy number
~

defined on by the formula

~
A ( A) 
for all   ( A) . For example, the fuzzy cardinality of
fuzzy set D2 defined in Table 1.2, whose  -cuts are
shown in Fig. 1.8a is
~
D2  .13 /19  .27 /17  .4 /15  .53 /13  .67 /11  .8 / 9  .93 / 7  1/ 5.
Arithmetic Operations on Intervals contd.
Fuzzy arithmetic is based on two properties of fuzzy numbers:
1) Each fuzzy set, and thus also each fuzzy number, can fully and
uniquely be represented by its  -cuts ;and
2)  -cuts of each fuzzy number are closed intervals of real numbers
for all   (0,1]
These properties enable us to define arithmetic operations on fuzzy
numbers in terms of arithmetic operations on their  -cuts(i.e.,
arithmetic operations on closed intervals).
Let  denote any of the four arithmetic operations on closed
intervals: addition +, subtraction -, multiplication ., and division /.
Then ,
[a, b]  [d , e]  { f  g | a  f  b, d  g  e} (22)
is a general property of all arithmetic operations on closed
intervals, except that [a, b] / [d , e] is not defined when 0 [d , e] .
That is, the result of an arithmetic operations on closed intervals
is again a closed interval.
Arithmetic Operations on Intervals contd.
The four arithmetic operations on closed intervals are defined as
follows:
[a, b]  [d , e]  [a  d , b  e], (23)
[a, b]  [d , e]  [a  e, b  d ], (24)
[a, b].[d , e]  [min(ad , ae, bd , be), max(ad , ae, bd , be)] (25)
and provided that 0 [d , e] .
[a, b] / [d , e]  [a, b].[1/ e.1/ d ]
 [min(a / d , a / e, b / d , b / e), max(a / d , a / e, b / d , b / e)] (26)
Note that a real number r may also be regarded as a special
(degenerated) interval [r, r]. When one of the intervals in
(23)-(26) is degenerated, we obtain special operations;
when both of them are degenerated, we obtain the
standard arithmetic of real numbers.
Arithmetic Operations on Intervals contd.
Example: [2,5]  [1,3]  [3,8], [0,1]  [6,5]  [6,6],
[2,5]  [1,3]  [ 1, 4], [0,1]  [6,5]  [5,7],
[1,1].[2, 0.5]  [2, 2] [3, 4].[2, 2]  [6,8]
[1,1] / [2, 0.5]  [2, 2] [4,10] / [1, 2]  [2,10]
Arithmetic operations on closed intervals satisfy some
useful properties. To overview them,
Let A  [a1 , a2 ], B  [b1 , b2 ], C  [c1 , c2 ],0  [0,0],1  [1,1].
Using these properties are formulated as follows:
 A  B  B  A,
A  B  B  A(commutavity ).
 ( A  B)  C  A  ( B  C ),
( A  B)  C  A  ( B  C )(associativity ).
Arithmetic Operations on Intervals contd.
 A  0 A  A0
 A  1 A  A 1(identity ).
 A  ( B  C )  A  B  A  C ( subdistributivity ).
 If b  c  0 for every b  B and c  C , then
A  ( B  C )  A  B  A  C (distributivity ).
Furthermore, if A  [a, a], then a  ( B  C )  a  B  a  C.
 0  A  A and 1 A / A.
 A  E and B  F , then
A  B  E  F.
A  B  E  F.
A  B  E  F.
A / B  E / F (inclusionmonotonicity ).
Arithmetic Operations on Intervals contd.
As an example, we prove only the less obvious properties
of subdistributivity and distribuity. First, we have
A  ( B  C )  {a  (b  c) | a  A, b  B, c  C}
 {a  b  a  c | a  A, b  B, c  C}
 {a  b  a '  c | a  a '  A, b  B, c  C}
 A  B  A  C.
Hence, A  ( B  C )  A  B  A  C.
Assume now, without any loss of generality, that b1  0 and
c1  0 . Then, we have to consider the following three
cases: 1. If a1  0 , then A  ( B  C )  [a1  (b1  c1 ), a2  (b2  c2 )]
 [a1  b1 , a2  b2 ]  [a1  c1  a2  c2 ]
 A  B  A  C.
Arithmetic Operations on Intervals contd.
2. If a1  0 and a2  0 , then a2  0,( A)  [a2 , a1 ], and
 A  ( B  C )  ( A)  B  ( A)  C.
Hence A  ( B  C )  A  B  A  C.
3. If a1  0 and a2  0 , then
A  ( B  C )  [a1  (b2  c2 ), a2  (b2  c2 )]
 [a1  b2 , a2  b2 ]  [a1  c2  a2  c2 ]
 A  B  A  C.
To show that distributivity does not hold in general, let
A  [0,1], B  [1,2], C  [2, 1].
Then, A  B  [0,2], A  C  [2,0], B  C  [1,1], and

A  ( B  C )  [1,1]  [2, 2]  A  B  A  C.
Arithmetic Operations on Fuzzy Numbers

Let A and B denote fuzzy numbers and let  denote any

of the four basic arithmetic operations. Then, we define
a fuzzy set on , A  B , by defining its  -cut,  ( A  B) , as

( A  B)   A   B (27)
for any   (0,1] . (when  =/, clearly, we have to require
that 0   B for all   (0,1] .)
By using Theorem 5(First Decomposition Theorem), A  B
can be expressed as
A B   ( A  B) (28)
 0,1
Since  ( A  B) is a closed interval for each   (0,1] and
A, B are fuzzy numbers, A  B is also a fuzzy number.
Arithmetic Operations on Fuzzy Numbers contd.
Example:
Consider two triangular-shape fuzzy numbers A and B
defined as follows:
 0 for x  1and x  3

A( x)   ( x  1) / 2 for 1  x  1
(3  x) / 2 for 1  x  3,

 0 for x  1and x  5

B ( x)   ( x  1) / 2 for 1 x  3
(5  x) / 2 for 3  x  5.

Their  -cuts are 
A  [2  1,3  2 ],

B  [2  1,5  2 ].
Arithmetic Operations on Fuzzy Numbers contd.
Using (23)-(27), we get

( A  B)  [4 ,8  4 ], for   (0,1]

( A  B)  [4  6, 2  4 ] for   (0,1] ,

 [ 4 2
 12   5, 4 2
 16  15] for   (0,.5]
( A  B)  
 [4  2
 1, 4 2
 16  15] for   (.5,1],
  (2  1) / (2  1),(3  2 ) / (2  1)] for   (0,.5]
( A / B)  
(2  1) / (5  2 ),(3  2 ) / (2  1)] for   (.5,1].
The resulting fuzzy numbers are then
 0 for x  0 and x  8

( A  B)( x)   x / 4 for 0 x4
(8  x) / 4 for 4  x  8,

Arithmetic operations on Fuzzy numbers
Arithmetic Operations on Fuzzy Numbers contd.
 0 for x  6 and x  2

( A  B )( x)  ( x  6) / 4 for 6  x  2
(2  x) / 4 for 2  x  2,

 0 for x  5 and x  15
[3  (4  x)1/2 ] / 2 5  x  0
 for
( A  B )( x)  
 (1  x )1/2
/2 for 0  x  3
[4  (1  x)1/2 ] / 2 for 3  x  15,
 0 for x  1 and x  3
 ( x  1) / (2  2 x) for 1  x  0

( A / B)( x)  
(5 x  1) / (2 x  2) for 0  x  1/ 3
 (3  x) / (2 x  2) for 1/ 3  x  3.
Arithmetic Operations on Fuzzy Numbers contd.
Let  denote any of the four basic arithmetic operations
and let A, B denote fuzzy numbers. Then, we define a
fuzzy set on , A  B, by the equation
( A  B)( z )  sup min[ A( x), B( y )] (29)
z  x y

for all z  . More specifically, we define for all z  ,

( A  B)( z )  sup min[ A( x), B( y )] (30)
z  x y

( A  B)( z )  sup min[ A( x), B( y )]

(31)
z  x y
(32)
( A  B)( z )  sup min[ A( x), B( y )]
z  x y (33)
( A / B)( z )  sup min[ A( x), B( y )]
z x/ y
Where A B defined by (29) is a fuzzy set on .
 Fuzzy Arithmetic contd.

Linguistic Variables:

The concept of a fuzzy number plays a fundamental role

in formulating quantitative fuzzy variables.

These are variables whose states are fuzzy numbers.

When, in addition, the fuzzy numbers represent

linguistic concepts, such as very small, small, medium,
and so on, as interpreted in a particular context, the
resulting constructs are usually called linguistic variables.
Fuzzy Arithmetic contd.

Linguistic Variables:

 Each linguistic variable the states of which are expressed by

linguistic terms interpreted as specific fuzzy numbers is
defined in terms of a base variable, the values of which are
real numbers within a specific range. is fully characterized

 A base variable is a variable in the classical sense,

exemplified by any physical variable(e.g., temp., press.,
speed, humidity, etc.) as well as any other numerical
variable(e.g., age, intr. rate, salary, blood count, probability,
reliability, etc.).
Linguistic Variables contd.
 Each linguistic variable is fully characterized by a quintuple
(v, T , X , g , m) in which v is the name of the variable, T is the set
of linguistic terms of v that refer to a base variable whose values
range over a universal set X , g is a synthetic rule (a grammar) for
generating linguistic terms, and m is a semantic rule that assigns to
each linguistic term t  T its meaning, m(t ) , which is a fuzzy set on X
(i.e., m : T  F  X  ).

 An example of a linguistic variable performance is shown Fig 4.4.

each of the basic linguistic terms is assigned one of five fuzzy
numbers by a semantic rule. The fuzzy numbers , whose
membership functions have the usual trapezodial shapes, are
defined on the interval [0,100], the range of the base variable.
 Each of them expresses a fuzzy restriction on this range.
Linguistic Variables contd.

An example of a linguistic variable performance