Applied Soft Computing 19 (2014) 171176

Contents lists available at ScienceDirect

Applied Soft Computing

journal homepage: www.elsevier.com/locate/asoc

A simplied new approach for solving fuzzy transportation problems

with generalized trapezoidal fuzzy numbers
Ali Ebrahimnejad
Department of Mathematics, Qaemshahr Branch, Islamic Azad University, Qaemshahr, Iran

a r t i c l e

i n f o

Article history:
Received 6 July 2013
Received in revised form 29 January 2014
Accepted 30 January 2014
Available online 22 February 2014
Keywords:
Fuzzy transportation problem
Ranking function
Generalized trapezoidal fuzzy number

a b s t r a c t
In a recent paper, Kaur and Kumar (2012) proposed a new method based on ranking function for solving
fuzzy transportation problem (FTP) by assuming that the values of transportation costs are represented
by generalized trapezoidal fuzzy numbers. Here it is shown that once the ranking function is chosen,
the FTP is converted into crisp one, which is easily solved by the standard transportation algorithms.
The main contribution here is the reduction of the computational complexity of the existing method. By
solving two application examples, it is shown that it is possible to nd a same optimal solution without
solving any FTP. Since the proposed approach is based on classical approach it is very easy to understand
and to apply on real life transportation problems for the decision makers.
2014 Elsevier B.V. All rights reserved.

1. Introduction
Transportation problem is an important network structured linear programming problem that arises in several contexts and has
deservedly received a great deal of attention in the literature. The
central concept in the problem is to nd the least total transportation cost of a commodity in order to satisfy demands at destinations
using available supplies at origins. Transportation problem can be
used for a wide variety of situations such as scheduling, production, investment, plant location, inventory control, employment
scheduling, and many others. In general, transportation problems
are solved with the assumptions that the transportation costs and
values of supplies and demands are specied in a precise way i.e., in
crisp environment. However, in many cases the decision maker has
no crisp information about the coefcients belonging to the transportation problem. In these cases, the corresponding coefcients
or elements dening the problem can be formulated by means of
fuzzy sets, and the fuzzy transportation problem (FTP) appears in a
natural way.
Since the transportation problem is essentially a linear programming problem, one straightforward approach is to apply the existing fuzzy linear programming techniques [2,11,14,27,31,32,34,35]
to the FTP. But, some of these techniques [2,32] only give crisp
solutions which represent a compromise solution in terms of fuzzy
data.

Tel.: +98 09112129663.

E-mail addresses: aemarzoun@gmail.com, a.ebrahimnejad@srbiau.ac.ir
http://dx.doi.org/10.1016/j.asoc.2014.01.041
1568-4946/ 2014 Elsevier B.V. All rights reserved.

Several researchers have carried out investigations on FTP

[1,4,5,12,13,1526,28,29,33]. Zimmermann [35] developed Zimmermanns fuzzy linear programming into several fuzzy optimization
methods for solving the transportation problems. Oheigeartaigh
[25] proposed an algorithm for solving transportation problems
where the supplies and demands are fuzzy sets with linear or
triangular membership functions. Chanas et al. [4] investigated
the transportation problem with fuzzy supplies and demands and
solved them via the parametric programming technique. Their
method provided solution which simultaneously satises the constraints and the goal to a maximal degree. In addition, Chanas and
Kuchta [5] discussed the type of transportation problems with fuzzy
cost coefcients and converted the problem into a bicriterial transportation problem with crisp objective function. Their method only
gives crisp solutions based on efcient solutions of the converted
problems. Jimenez and Verdegay [15,16] investigated the fuzzy
solid transportation problem in which supplies, demands and conveyance capacities are represented by trapezoidal fuzzy numbers
and applied a parametric approach for nding the fuzzy solution.
Liu and Kao [24] developed a procedure, based on extension principle to derive the fuzzy objective value of FTP in that the cost
coefcients and the supply and demand quantities are fuzzy numbers. Gani and Razak [12] presented a two stage cost minimizing
FTP in which supplies and demands are as trapezoidal fuzzy numbers and used a parametric approach for nding a fuzzy solution
with the aim of minimizing the sum of the transportation costs
in the two stages. Yang and Liu [33] investigated the xed charge
solid transportation problem under fuzzy environment, in which
the direct costs, the xed charges, the supplies, the demands and
the conveyance capacities are supposed to be fuzzy variables. Li

172

A. Ebrahimnejad / Applied Soft Computing 19 (2014) 171176

et al. [22] proposed a new method based on goal programming for

solving FTP with fuzzy costs. Lin [23] used genetic algorithm for
solving transportation problems with fuzzy coefcients. Dinagar
and Palanivel [10] investigated FTP, with the help of trapezoidal
fuzzy numbers and applied fuzzy modied distribution method to
obtain the optimal solution in terms of fuzzy numbers. Pandian
and Natarajan [26] introduced a new algorithm namely, fuzzy zero
point method for nding fuzzy optimal solution for such FTP in
which the transportation cost, supply and demand are represented
by trapezoidal fuzzy numbers. Kumar and Kaur [19] proposed a
new method based on fuzzy linear programming problem for nding the optimal solution of FTP. Chakraborty and Chakraborty [3]
proposed a method for the minimization of transportation cost
as well as time of transportation when the demand, supply and
transportation cost per unit of the quantities are fuzzy. Basirzadeh
[1] proposed a systematic procedure for solving all types of FTP
whether maximize or minimize objective function. Tao and Xu [30]
established a rough multiple objective programming model for a
solid transportation problem. Gupat and Kumar [13] proposed a
new method to nd solution of a linear multi-objective transportation problem by representing all parameters as interval-valued
fuzzy numbers. Shanmugasundari and Ganesan [28] developed
the fuzzy version of Vogels and MODI methods for obtaining the
fuzzy initial basic feasible solution and fuzzy optimal solution,
respectively.
Chen [6] proposed the concept of generalized fuzzy numbers
for situations that the membership function is not restricted to
the normal form. Since then, a high number of researchers [79]
have devoted their efforts to use generalized fuzzy numbers in
real life problems, but there are few papers in which generalized fuzzy numbers are used to solve FTP. Kaur and Kumar [17]
proposed a new method based on ranking function for solving
FTP by assuming that the parameters of transportation problem
are represented by generalized trapezoidal fuzzy numbers. In a
recent paper in this journal, Kaur and Kumar [18] studied a special type of FTP by assuming that a decision maker is uncertain
about the precise values of transportation cost only but there is
no uncertainty about the supply and demand of the product. In
that study, transportation costs were represented by generalized
trapezoidal fuzzy numbers. They modied some existing methods
for nding the initial basic feasible solution (IBFS) and fuzzy optimal solution using ranking function, in which transportation costs
are represented as generalized trapezoidal fuzzy numbers. In this
paper, we show that once the ranking function is chosen, the FTP
is converted into crisp one, which is easily solved by the standard
transportation algorithms. It is demonstrated that the method used
in this study is simpler and computationally more efcient than the
proposed method by Kaur and Kumar [18]. Since the proposed technique is based on classical approach it is very easy to understand
and to apply on real life transportation problems for the decision
makers.
This paper is organized as follows: In Section 2, formulation of
special type of the FTP in terms of generalized trapezoidal fuzzy
numbers and summary of the existing method are given. In Section
3, a simplied new method is proposed to nd optimal solution
of FTP. In Section 4, the application of the proposed method is
shown using two application examples and the obtained results
are discussed. Section 5 ends this paper with a brief conclusion and
suggestion for future directions.

2. Fuzzy transportation problem

The FTP, in which a decision maker is uncertain about the precise values of transportation cost from the ith source to the jth

destination, but sure about the supply and demand of the product,
can be formulated as follows [18]:
min

n
m



cij xij

i=1 j=1
n

s.t.

xij = ai ,

i = 1, 2, . . ., m,

j=1

(1)

xij = bj ,

j = 1, 2, . . ., n,

i=1

xij 0, i = 1, 2, . . ., m, j = 1, 2, . . ., n.
where ai is the total availability of the product at ith source; bj is the
total demand of the product at jth destination; cij is the approximate
cost for transporting one unit quantity of the product from the ith
source to the jth destination; xij is the number of units of the product
that should be transported
m from
n the ith source to the jth destination
or decision variables;
c x is total fuzzy transportation
i=1
j=1 ij ij
cost.
Let u i and v i be the fuzzy dual variables associated with ith row
and jth column constraints, respectively, then the fuzzy dual of the
FTP given in Eq. (1) will be as follows [18]:
max

m


ai u i

i=1

n


bj v j

j=1

cij
s.t.u i v j

(2)

i = 1, 2, . . ., m, j = 1, 2, . . ., n.

Kaur and Kumar [18] based on the ranking function given in

Appendix A, introduced three methods namely generalized fuzzy
northwest corner method (GFNCWM), generalized fuzzy least-cost
method (GFLCM) and generalized fuzzy Vogels approximation
method (GFVAM) to nd the initial basic feasible solution (IBFS)
of the FTP given in Eq. (1). These methods are the direct extension
of classical approaches. Then, they applied the generalized fuzzy
modied distribution method (GFMDM) to nd the fuzzy optimal
solution of FTP given in Eq. (1) with the help of IBFS. The FGMDM is
based on dual feasibility and complementary slackness theorem. In
their proposed method all arithmetic operations are performed on
the generalized trapezoidal fuzzy numbers, i.e., u i , v j and cij . In the
following section, we show that it is possible to nd the same solution of the FTP given in Eq. (1) with the help of crisp transportation
problem and so all arithmetic operations are done on real numbers
instead of generalized trapezoidal fuzzy numbers.
3. Main results
According to ranking function given in Appendix A, it is possible to dene a rank for each generalized trapezoidal fuzzy number.
i = (ai , bi , ci , di ; wi ), (i = 1, 2, .., k) be k generalIt follows that if A
fuzzy numbers, then
ized trapezoidal 
 (A i ) = w(ai + bi + ci + di )/4,
where w = min wi , i = 1, 2, . . ., k . This helps us to convert the
FTP given in Eq. (1) into an equivalent crisp transportation problem.
To do this, we substitute the rank of each generalized trapezoidal
fuzzy number instead of the corresponding generalized trapezoidal
fuzzy number in the FTP under consideration. This leads to an
equivalent crisp transportation problem which can be solved by
the standard transportation algorithms. Then, all arithmetic operations are done on the crisp numbers. As a result, the computational
effort is decreased signicantly in our proposed approach.
It is worth noting that the general steps of the Kaur and Kumars
method [18] to the FTP given in Eq. (1) are as follows:
Step 1. Find an IBFS of the FTP given in Eq. (1) using GFNCWM or
GFLCM or GFVAM.

A. Ebrahimnejad / Applied Soft Computing 19 (2014) 171176

173

Table 1
Tabular representation of the chosen FTP.

S1
S2
S3
Demand (bj )

D1

D2

D3

Availability (ai )

(1,4,9,19;0.5)
(8,9,12,26; 0.5)
(11,12,20,27;0.5)
15

(1,2,5,9;0.4)
(3,5,8,12;0.2)
(0,5,10,15;0.8)
14

(2,5,8,18;0.5)
(7,9,13,28;0.4)
(4,5,8,11;0.6)
10

10
14
15

Step 2. Compute d ij = cij (u i v j ) for each non-basic variable.

Stop or select an entering column.
Step 3. Determine an existing column.
Step 4. Obtain the new BFS and repeat Step 2.
Our main contribution here is the reduction of the computational complexity of the proposed method by Kaur and Kumar
[18]. In particular, it is shown that our proposed method need to
less number of elementary operations such as additions, multiplications, and comparisons as compared to the Kaur and Kumars
method [18].
According to the Kaur and Kumars method [18], to carry out
the Step 1 using GFCLM it is required to determine the smallest
fuzzy transportation cost in fuzzy transportation tableau (FTT) of
order m n until the FTT is reduced into a FTT of order 1 1. To do
this, it is required to compare the fuzzy transportation costs using
ranking function given in Appendix A until the FTT is reduced into a
FTT of order 1 1. While based on our proposed method the comparison of fuzzy costs is done once. Moreover, to nd IBFS using
GFVAM it is required to compute a fuzzy penalty for each row of
the FTT of order m n by subtracting the smallest entry from the
next smallest entry in each step of this method. In a similar way,
a fuzzy penalty is computed for each column of the FTT of order
m n in all iterations. After that, the highest fuzzy penalty is determined based on ranking function given in Appendix A. These steps
are repeated until the FTT is reduced into a FTT of order 1 1. As
we see to do Step 1 using GFVAM, it is required a lot of fuzzy arithmetic operations such as additions, subtractions and comparison
on generalized trapezoidal fuzzy numbers. While based on our proposed method the comparison of fuzzy costs is done once and all
arithmetic operations are done on real numbers.
In addition, to carry out the Step 2 using the GFMDM proposed
by Kaur and Kumar [18] it is required to solve the fuzzy system
u i v j = cij with m + n 1 fuzzy equations corresponding to basic
variables. After solving this fuzzy system, the fuzzy value d ij for each

non-basic variable is obtained based on d ij = cij (u i v j ). Finally,

the entering column is determined according to the most negative
rank of d ij . This step requires to a lot of fuzzy additions and subtractions on generalized trapezoidal fuzzy numbers. While based on our
proposed methods, the entering column is found without solving
any fuzzy system and without any fuzzy arithmetic operations.
These results conrm that the proposed method is simpler and
computationally more efcient than the proposed method by Kaur
and Kumar [18].

transporting one unit quantity of product from each source to each

destination is represented by generalized trapezoidal fuzzy number. Determine the fuzzy optimal transportation of products such
that the total fuzzy transportation cost is minimum.
We rst put the rank order of each fuzzy transportation cost
(given in Table 1) instead of corresponding fuzzy number to
obtain the classical transportation problem. The results are given
in Table 2.
The crisp transportation problem given in Table 2 can be solved
by the standard transportation algorithms. We rst apply three
famous methods namely, the north-west corner method, least cost
method and Vogels approximation method for nding the IBFS and
then use the modied distribution method on the IBFS to obtain the
optimal solution. Finally, we compare the obtained result with the
one found by Kaur and Kumar [18].
If we use the north-west corner method to obtain the IBFS, we
get the following solution which is matched with IBFS obtained
based on GFNWCM proposed by Kaur and Kumar [18]:
x11 = 10, x12 = 0, x13 = 0,
x21 = 5, x22 = 9, x23 = 0,

(3)

x31 = 0, x32 = 5, x33 = 10.

Now applying the modied distribution method on the IBFS
given in Eq. (3) shows that this initial solution is the optimal solution of the FTP given in Table 1, too. It needs to point out that Kaur
and Kumar [18] obtained the same optimal solution by using the
generalized fuzzy modied distribution method on the IBFS given
in Eq. (3). But, in their method all arithmetic operations are performed on the generalized trapezoidal fuzzy numbers, while in
the proposed method all arithmetic operations are done on real
numbers.
In addition, if we apply the least cost method to obtain the IBFS,
we get the following solution which is matched with IBFS obtained
based on GFLCM which proposed by Kaur and Kumar [18]:
x11 = 0, x12 = 10, x13 = 0,
x21 = 10, x22 = 4, x23 = 0,

(4)

x31 = 5, x32 = 0, x33 = 10.

In this case, using the modied distribution method on the
obtained IBFS given in Eq. (4) gives the following optimal solution
after three iterations:
x11 = 10, x12 = 0, x13 = 0,

4. Application

x21 = 5, x22 = 9, x23 = 0,

In this section, two FTP (adopted from [18]) are used in order
to show the applicability of the proposed method and the obtained
results are discussed.

x31 = 0, x32 = 5, x33 = 10.

Table 2
Tabular representation of classical transportation problem.

4.1. Examples
Example 4.1. Table 1 gives the availability (ai ) of the product available at three sources Si , i = 1, 2, 3 and their demand (bj )
at three destinations Dj , j = 1, 2, 3, and the approximate cost for

(5)

S1
S2
S3
Demand (bj )

D1

D2

D3

Availability (ai )

1.65
2.75
3.5
15

0.85
1.4
1.5
14

1.65
2.85
1.4
10

10
14
15

174

A. Ebrahimnejad / Applied Soft Computing 19 (2014) 171176

Table 3
Summary of the fuzzy transportation problem.
Colliery

S1
S2
S3
S4
bj

Washery

ai

D1

D2

D3

D4

D5

D6

(20,27,35,41;0.7)
(20,25,35,41;0.7)
(9,10,12,16;0.7)
(9,11,12,14;.6)
112

(9,11,12,14;0.6)
(9,11,12,16;0.8)
(65,70,74,76;0.8)
(10,15,21,24;0.7)
90

(10,15,18,20;0.7)
(9,11,12,14,0.6)
(20,25,35,41;0.8)
(20,25,35,41;0.6)
84

(15,20,22,24;0.7)
(10,15,21,23;0.8)
(12,15,22,24;0.7)
(10,15,18,20;0.6)
92

(10,15,18,20;0.8)
(10,15,18,23;0.6)
(10,12,14,18;0.6)
(8,10,12,14;0.8)
106

(9,12,15,18;0.7)
(9,12,15,18;0.8)
(15,20,26,28;0.8)
(15,20,25,28;07)
80

We note that Kaur and Kumar [18] obtained the same optimal solution by using the generalized fuzzy modied distribution
method on the IBFS given in Eq. (4).
Finally, if we use the Vogels approximation method to nd the
IBFS, we reach to the following solution which is matched with IBFS
obtained based on GFVAM proposed by Kaur and Kumar [18]:
x11 = 10, x12 = 0, x13 = 0,
x21 = 0, x22 = 14, x23 = 0,

(6)

x31 = 5, x32 = 0, x33 = 10.

In this case, if we apply the modied distribution method on the
IBFS given in Eq. (6), we nd the following optimal solution after
two iterations:

124
120
150
170

and several other reasons the owner of the company is not certain about the transportation cost (in Rs) from different collieries
to different washeries. According to past experiences of the owner
the transportation cost is represented by generalized trapezoidal
fuzzy numbers. The owner of the company wants to determine the
fuzzy optimal transportation of products such that the total fuzzy
transportation cost is minimum.
We rst substitute the rank order of each fuzzy transportation
cost (given in Table 3) instead of corresponding fuzzy number. This
leads to the classical transportation problem given in Table 4.
In this case, the crisp transportation problem shown in Table 4
can be solved by the standard transportation algorithms. The
obtained optimal solution and minimum total fuzzy transportation
cost are as follows:
x11 = 0, x12 = 80, x13 = 0, x14 = 0, x15 = 0, x16 = 44,

x11 = 10, x12 = 0, x13 = 0,

x21 = 5, x22 = 9, x23 = 0,

(7)

x21 = 0, x22 = 0, x23 = 84, x24 = 0, x25 = 0, x26 = 0,

x31 = 112, x32 = 0, x33 = 0, x34 = 38, x35 = 0, x36 = 0,

x31 = 0, x32 = 5, x33 = 10.

Kaur and Kumar [18] obtained the same optimal solution using
the generalized fuzzy modied distribution method on the IBFS
given in Eq. (6).
The minimum fuzzy transportation cost is achieved as follows:

(8)

x41 = 0, x42 = 10, x43 = 0, x44 = 54, x45 = 106, x46 = 0.

3
3



cij xij = (5148, 6474, 7802, 9244; 0.6).

i=1 j=1
3
3



cij xij = 10(1, 4, 9, 19; 0.5) 5(8, 9, 12, 26; 0.5) 9(3, 5, 8, 12; 0.2) 5(0, 5, 10, 15; 0.8) 10(4, 5, 8, 11; 0.6)

i=1 j=1

= (10, 40, 90, 190; 0.5) (40, 45, 60, 130; 0.5) (27, 45, 72, 108; 0.2) (0, 25, 50, 75; 0.8) (40, 50, 80, 110; 0.6)
= (117, 205, 352, 613; 0.2)
As shown here, the fuzzy optimal solution of the classical
method proposed in this study is equivalent to the optimal solution
derived from Kaur and Kumars method [18]. However, the method
proposed is by far simpler and computationally more efcient than
the fuzzy method proposed by Kaur and Kumar [18].
Example 4.2. The data, collected from an owner of a regional coal
company (the data is provided with a legal agreement that the name
of the company will not be disclosed) situated in Jharia (Dhanbad,
India) is shown in Table 3.
The owner of the company is certain about the availabilities
and demands of the coal at different collieries and washeries,
respectively. These parameters are represented by real numbers
in Table 3. Because of frequently variation in the rates of diesel
Table 4
Summary of the classical transportation problem.
Colliery

S1
S2
S3
S4
bj

Washery

ai

D1

D2

D3

D4

D5

D6

18.45
18.15
7.05
6.9
112

6.9
7.2
42.75
10.5
90

9.45
6.9
18.15
16.35
84

12.15
10.35
10.95
9.45
92

9.45
9.9
8.1
6.6
106

8.1
8.1
13.35
13.2
80

It is worth to note that the same optimal solution can be found

based on using all the proposed methods by Kaur and Kumar [18].
These results conrm that if we want to solve a FTP based on ranking
function, then the proposed approach is more effective than the
existing method [18].
4.2. Results and discussions
In this section, the main advantages of the proposed method
over the existing methods are explored.
It is worth noting that, as mentioned in Remark 2, one can use
any other linear ranking function, and although the obtained solution may be different but the results are still valid for the new
solution. In fact, depending on the ranking index to be used by
the decision-maker, the algorithm can return a set of the optimal
solutions or a unique solution.
For example, assume that the following linear ranking func=
tion is used for the generalized trapezoidal fuzzy number A
(a, b, c, d; w):

= w [(1 )a + d] +

 (A)
124
120
150
170

1
[(1 )(b a) + (c d)]
2

(9)

where  [0, 1] is the optimism index reecting the optimism

degree of a decision maker. The larger  is, the more optimistic the
decision maker is. The two extreme cases are  = 0, the decision
maker is completely pessimistic; and  = 1, the decision maker is

A. Ebrahimnejad / Applied Soft Computing 19 (2014) 171176

175

completely optimistic. The case  = 1/2 reects a linear decision

attitude.
It can be veried that the linear ranking function given in
Appendix A is a particular case of linear ranking function given in
Eq. (9) when  = 1/2.
Solving the FTP of Example 4.1 based on ranking function given
in Eq. (9) for  = 0 by use of both the proposed algorithm and
the Kaur and Kumars algorithm [18] produce the following same
result:

Here, we shall point out that the FTP studied in this paper is
not in the form of a problem whose model involves interval-valued
trapezoidal fuzzy numbers. Therefore, further research on extending the proposed method to overcome these shortcomings is an
interesting stream of future research. We shall report the signicant
results of these ongoing projects in the near future.

x11 = 10, x12 = 0, x13 = 0,

The author would like to express their sincerest thanks to the

anonymous referees and honorable editor for their valuable comments which help to improve throughout of this paper. Moreover,
the author greatly appreciates the ofce of vice chancellor for
research of Islamic Azad University, Qaemshahr Branch, for nancially support.

x21 = 5, x22 = 0, x23 = 9,

(10)

x31 = 0, x32 = 14, x33 = 1.

This solution is obtained based on point of view of a completely
pessimistic decision maker. Although the obtained optimal solution
based on this ranking function is different with the optimal solution
given in Eq. (7) but the main result is valid for this solution yet. In
other words, if a unique ranking function is used for solving FTP
under consideration, then both of our proposed method and Kaur
and Kuamrs approach produce a same optimal solution.
Let us explore the main advantages of the proposed method.
(1) The proposed technique does not use the goal and parametric
approaches which are difcult to apply in real life situations.
(2) By applying the proposed approach for nding the fuzzy optimal solution, there is no need of much knowledge of fuzzy linear
programming technique, Zimmerman approach and crisp linear programming which are difcult to learn for a new decision
maker.
(3) The proposed method to solve FTP is based on traditional
transportation algorithms. Thus, the existing and easily available software can be used for the same. However, the existing
method [18] to solve FTP should be implemented into a programming language.
(4) To solve the FTP by using the existing method [18], there is
need to use arithmetic operations of generalized fuzzy numbers. While, if the proposed technique is used for the same then
there is need to use arithmetic operations of real numbers. This
proves that it is much easy to apply the proposed method as
compared to the existing method [18].
(5) Moreover, it is possible to assume a generic ranking index for
comparing the fuzzy numbers involved in the FTP problem, in
such a way that each time in which the decision maker wants
to solve the FTP problem under consideration (s)he can choose
(or propose) the ranking index that best suits the FTP problem.
5. Conclusions and future work
A large number of transportation problems with different levels of sophistication have been studied in the literature. However,
some of these problems have limited real-life applications because
the conventional transportation problems generally assume crisp
data for the transportation cost, the values of supplies and demands.
Contrary to the conventional transportation problems, we investigated imprecise data in the real-life transportation problems and
developed an alternative method that is simple and yet addresses
these shortfalls in the existing models in the literature. In the FTP
considered in this study, the values of transportation costs are represented by generalized trapezoidal fuzzy numbers and the values
of supply and demand of products are represented by real numbers.
Here we concluded that once the ranking function is chosen, the FTP
is converted into crisp one, which is easily solved by the standard
transportation algorithms. By solving two numerical examples, we
conclude that it is possible to nd a same optimal solution without
solving any FTP.

Acknowledgements

Appendix A. Generalized fuzzy numbers, arithmetic

operations and ranking
= (a, b, c, d; w) is said to be
Denition 1 ([8,21]). A fuzzy number A
a generalized trapezoidal fuzzy number if its membership function
is given by:

A (x) =

(x a)

w (b a) , a x b,

w,

b x c,

0,

otherwise.

(x d)

, c x d,
w

(c d)

1 = (a1 , b1 , c1 , d1 ; w1 ) and A
2 =
Denition 2 ([8,21]). Let A
(a2 , b2 , c2 , d2 ; w2 ) be two generalized trapezoidal fuzzy numbers
1 and A
2 are given
and  R. Then the arithmetic operations on A
by:
1 = (a1 , b1 , c1 , d1 ; w1 )
i) 0, A
1 = (d1 , c1 , b1 , a1 ; w1 )
ii)  < 0, A
1 A
2 = A
1 = (a1 + a2 , b1 + b2 , c1 + c2 , d1 + d2 ; w =
iii) A
min {w1 , w2 })
1 A
2 = (a1 d2 , b1 c2 , c1 b2 , d1 a2 ; w = min {w1 , w2 })
iv) A
Remark 1. Kaur and Kumar [18] used the ranking function  :
F(R) R, where F(R) is a set of fuzzy numbers dened on set of real
numbers, for comparing two generalized trapezoidal fuzzy num 2 = (a2 , b2 , c2 , d2 ; w2 ). This
1 = (a1 , b1 , c1 , d1 ; w1 ) and A
bers A
ranking function maps each generalized trapezoidal fuzzy number
into the real line, where a natural order exists, i.e.,
2,
1
A
A

if

and

w(a2 +b2 +c2 +d2 )

,w
4

1
2,
A
A

if

1
2,
A
=A

if

1) =
(A

w(a1 +b1 +c1 +d1 )

4

2) =
(A

only

1) =
(A

if

w(a1 +b1 +c1 +d1 )

2)
(A
4

= min {w1 , w2 },

and

w(a2 +b2 +c2 +d2 )

,w
4

if

= min {w1 , w2 },

and

w(a2 +b2 +c2 +d2 )

,w
4

only

only

if

1) =
(A

w(a1 +b1 +c1 +d1 )

4

2) =
= (A

= min {w1 , w2 }.

Remark 2. We realize that the results obtained here are independent of the choice of the linear ranking function. In other words,
we can use any other linear ranking function, and although the
solution obtained may be different but the results are still valid
for the new solution. As for the types of the fuzzy data in the
model and the assumption of fuzziness in the variables, the choice
and compatibility of the ranking function for FTP should be the
decision makers main concerns. For generalized trapezoidal fuzzy

176

A. Ebrahimnejad / Applied Soft Computing 19 (2014) 171176

numbers, the linear ranking function used here is deemed to be

appropriate.
References
[1] H. Basirzadeh, An approach for solving fuzzy transportation problem, Applied
Mathematical Sciences 5 (32) (2011) 15491566.
[2] J.J. Buckly, Possibilistic linear programming with triangular fuzzy numbers,
Fuzzy Sets and Systems 26 (1) (1988) 135138.
[3] A. Chakraborty, M. Chakraborty, Cost-time minimization in a transportation
problem with fuzzy parameters: a case study, Journal of Transportation Systems Engineering and Information Technology 10 (6) (2010) 5363.
[4] S. Chanas, W. Kolodziejczyk, A. Machaj, A fuzzy approach to the transportation
problem, Fuzzy Sets and Systems 13 (3) (1984) 211221.
[5] S. Chanas, D. Kuchta, A concept of the optimal solution of the transportation
problem with fuzzy cost coefcients, Fuzzy Sets and Systems 82 (2) (1996)
299305.
[6] S.H. Chen, Operations on fuzzy numbers with function principal, Tamkang Journal of Management Sciences 6 (1985) 1325.
[7] S.J. Chen, S.M. Chen, A new method for handling multicriteria fuzzy decision making problems using FN-IOWA operators, Cybernetics and Systems 34
(2003) 109137.
[8] S.J. Chen, S.M. Chen, Fuzzy risk analysis on the ranking of generalized trapezoidal fuzzy numbers, Applied Intelligence 26 (2007) 111.
[9] S.M. Chen, J.H. Chen, Fuzzy risk analysis based on the ranking generalized fuzzy
numbers with different heights and different spreads, Expert Systems with
Applications 36 (2009) 68336842.
[10] D.S. Dinagar, K. Palanivel, The transportation problem in fuzzy environment,
International Journal of Algorithms, Computing and Mathematics 2 (3) (2009)
6571.
[11] A. Ebrahimnejad, A duality approach for solving bounded linear programming
with fuzzy variables based on ranking functions, International Journal of Systems Science (2014), http://dx.doi.org/10.1080/00207721.2013.844285.
[12] A. Gani, K.A. Razak, Two stage fuzzy transportation problem, Journal of Physical
Sciences 10 (2006) 6369.
[13] A. Gupta, A. Kumar, A new method for solving linear multi-objective transportation problems with fuzzy parameters, Applied Mathematical Modelling
36 (2012) 14211430.
[14] A. Hatami-Marbini, M. Tavana, An extension of the linear programming method
with fuzzy parameters, International Journal of Mathematics in Operational
Research 3 (1) (2011) 4455.
[15] F. Jimenez, J.L. Verdegay, Uncertain solid transportation problem, Fuzzy Sets
and Systems 100 (13) (1998) 4557.
[16] F. Jimenez, J.L. Verdegay, Solving fuzzy solid transportation problems by an
evolutionary algorithm based parametric approach, European Journal of Operational Research 117 (3) (1999) 485510.
[17] A. Kaur, A. Kumar, A new method for solving fuzzy transportation problems using ranking function, Applied Mathematical Modelling 35 (12) (2011)
56525661.

[18] A. Kaur, A. Kumar, A new approach for solving fuzzy transportation problems
using generalized trapezoidal fuzzy numbers, Applied Soft Computing 12 (3)
(2012) 12011213.
[19] A. Kumar, A. Kaur, Application of linear programming for solving fuzzy transportation problems, Journal of Applied Mathematics and Informatics 29 (3-4)
(2010) 831846.
[20] A. Kumar, A. Gupta, A. Kaur, Two new methods for solving fuzzy assignment
problems, Journal of Advanced Research in Scientic Computing 2 (2) (2010)
2134.
[21] A. Kumar, A. Gupta, Methods for solving fuzzy assignment problems and fuzzy
travelling salesman problems with different membership functions, Fuzzy
Information and Engineering 1 (2011) 321.
[22] L. Li, Z. Huang, Q. Da, J. Hu, A new method based on goal programming for
solving transportation problem with fuzzy cost, in: International Symposiums
on Information Processing, 2008, pp. 38.
[23] F.T. Lin, Solving the transportation problem with fuzzy coefcients using
genetic algorithms, in: IEEE International Conference on Fuzzy Systems, 2009,
pp. 14681473.
[24] S.T. Liu, C. Kao, Solving fuzzy transportation problems based on extension principle, European Journal of Operational Research 153 (3) (2004) 661674.
[25] M. Oheigeartaigh, A fuzzy transportation algorithm, Fuzzy Sets and Systems 8
(3) (1982) 235243.
[26] P. Pandian, G. Natarajan, A new algorithm for nding a fuzzy optimal solution
for fuzzy transportation problems, Applied Mathematical Sciences 4 (2) (2010)
7990.
[27] S. Saati, A. Hatami-Marbini, M. Tavana, E. Hajiahkondi, A two-fold linear
programming model with fuzzy data, International Journal of Fuzzy System
Applications 2 (3) (2012) 112.
[28] M. Shanmugasundari, K. Ganesan, A novel approach for the fuzzy optimal
solution of fuzzy transportation problem, International Journal of Engineering
Research and Applications 3 (1) (2013) 14161421.
[29] C. Sudhagar, K. Ganesan, A fuzzy approach to transport optimization problem, Optimization Engineering (2013), http://dx.doi.org/10.1007/
s11081-012-9202-6.
[30] Z. Tao, J. Xu, A class of rough multiple objective programming and its application to solid transportation problem, Information Sciences 188 (1) (2012)
215235.
[31] P. Vasant, Solving fuzzy linear programming problems with modied S-curve
membership function, International Journal of Uncertainty, Fuzziness and
Knowledge-Based Systems 13 (1) (2005) 97109.
[32] J.L. Verdegay, A dual approach to solve the fuzzy linear programming problem,
Fuzzy Sets and Systems 14 (2) (1984) 131141.
[33] L. Yang, L. Liu, Fuzzy xed charge solid transportation problem and algorithm,
Applied Soft Computing 7 (3) (2007) 879889.
[34] G. Zhang, H. Zuo, Research on multi-objective linear programming with fuzzy
coefcients constraints, International Journal of Machine Learning and Cybernetics (2014), http://dx.doi.org/10.1007/s13042-013-0173-5.
[35] H.I. Zimmermann, Fuzzy programming and linear programming with several
objective functions, Fuzzy Sets and Systems 1 (1) (1978) 4555.