You are on page 1of 11

1 May 3-5 2011 /

www.irtimco.com
Fuzzy Programming approach for Supplier Selection
Yosuff Mohammed
Industrial Engineering Department, Sharif University of Technology
PO Box 11364 Tehran, Iran.
yosmoh1980@Gmail.com
Abstract
Supplier selection problems are one of the most important decision-making problems for many
firms, which involving multiple objectives; however, it is difficult to solve because objectives are
often conflicting in nature. This study integrates fuzzy approaches as decision with multi choice
goal programming (MCGP) and aids to help decision makers to choose better suppliers by
considering multiple aspiration levels and vague goal relations. With fuzzy multi choice Goal
programming (FMCGP) method, decision makers can obtain the order quantities for suitable
suppliers based on different organizations supply chain strategies. In this paper, a formulation of
FMCGP is presented. To demonstrate the usefulness of the proposed method, a hypothetical case
study is presented and obtains a better solution than other approaches.
Keywords: Goal Programming; Multi-choice; Supplier selection; Fuzzy Goal Programming


.
) MCGP (
.
) FMCGP (
. FMCGP .
.

2 May 3-5 2011 /
www.irtimco.com
1-Introduction
Supplier selection is one of the most critical
activities of purchasing management in a
supply chain, because of the key role of
suppliers performance on cost, quality,
delivery and service in achieving the
objectives of a supply chain. Supplier
selection is a multiple-criteria decision-
making (MCDM) problem that is affected by
several conflicting factors. Depending on the
purchasing situations, criteria have varying
importance and there is a need to weight
criteria. In practice, for supplier selection
problems, most of the input information is
not known precisely.
In recent years, determining the best supplier
in the supply chain has become a key
strategic consideration in marketing
researches. However, these decisions usually
involve multi-goal and multi-criteria or
qualitative and quantitative factors. Supplier
selection decision is an important component
of purchasing management for many
companies (Dobler et al., 1990; Willis et al.,
1993). Selecting the right suppliers and
determining the appropriate orders from
them can bring significant benefit in the
reduction in purchasing cost, decrease in
supplying risk and improved product quality.
Accordingly, international organizations
usually build long term contracts with main
global suppliers and determine the
appropriate orders from them based on their
characteristics (cost capability, product
quality, on-time delivery and service) (Min,
1994; Ghodsypour & OBrien, 1998; Chan &
Kumar, 2007). According to suppliers
different characteristics, organizations adopt
different strategies such as cost leadership
strategy (Porter, 1980), differentiation
strategy (Porter, 1980), quick response (Iyer
& Bergen, 1997) and Just-in-time strategy
(Schonberger, 1982) to cope with different
issues.
When a supplier selection decision needs to
be made, the buyer generally establishes a set
of evaluation criteria that can be used to
compare potential sources. The basic criteria
typically utilized for this purpose are pricing
structure, delivery (timeliness and costs),
product quality, and service (personnel,
facilities, research and development,
capability, etc). For global markets, the set of
criteria is expanded to take into account the
new variables and risks associated with
international business transactions.
Frequently, these evaluation criteria conflict
one another. In addition, the importance of
each criterion varies from one purchase to
the next. This situation can be complicated
further by the fact that some of the criteria
are primarily quantitative (price and perhaps

3 May 3-5 2011 /
www.irtimco.com
quality) and some are qualitative (flexibility,
cooperation, service). Thus, a technique is
needed that can adjust for the decision
maker's attitude toward the importance of
each criterion for each item, as well as
capture both the subjective and the objective
criteria.
Analytic hierarchy process (AHP) was first
proposed by Saaty,1980. It has been widely
applied in areas such as MCDM and has
become a popular application for
performance evaluation. AHP must satisfy
the characteristic of independence among the
criteria before it can proceed to decision
making. However, given the problems
encountered in reality, a dependent and
feedback relationship will usually be
generated among the evaluation criteria and
such an interdependent relationship usually
becomes more complex with the change in
scope and depth of the decision-making
problems. When considering the setting of
the priority of goals, a decision maker (DM)
usually assigns different weights to each goal
that they are concerned with. Traditionally, a
DM can use analytic hierarchy process
(AHP) to structure the problem and
determine the weights of each attribute (Liou
& Chuang, 2008; Fazlollahtabar, 2008).
However, a DM will suffer from the
complicated pair comparison process
between the criteria in AHP hierarchy. With
AHP, a DM is asked to estimate pair wise
comparison ratios with respect to strength of
preference between suppliers by using a long
questionnaire. It is too complicated for DMs
to conduct pair wise comparison because
many supplier selection criteria and potential
suppliers should be considered
simultaneously. Moreover, if the housing
criteria hierarchy is not carefully designed,
the result of the problem will be invalid and
biased. Compared with AHP FGP is much
more suitable for DMs to make decisions
because it only requires DMs to set their
housing goals and constraints. This can help
reduce a DMs evaluation time, and it can be
easily implemented with a computer
program.
Curry and Lazzari (2009) adopted fuzzy
concept as linguistic variables to deal
consumer preferences. For this, several
previous studies have discussed how to
approach the different importance of goals in
fuzzy goal programming (FGP). Chen and
Tsai (2001) formulated FGP incorporating
different importance by using an additive
model to maximize the sum of achievement
degrees from all fuzzy goals. Akz and
Petrovic (2007) proposed a modified FGP
method to handle imprecision of the relative
importance relations among goals. However,

4 May 3-5 2011 /
www.irtimco.com
it is difficult to implement a situation when
the multiple aspiration levels appear in a goal
by using neither the proposed methods from
Chen and Tsai (2001) nor the work of Akz
and Petrovic (2007).
For the purchasing managers of
organizations, determining a specific target
value of each goal is not easy because only
partial information can be obtained in an
uncertain environment. This paper integrates
GP and fuzzy approaches in order to help
DMs to choose better suppliers by
considering multiple aspiration levels and
vague goal relations.
The aim of this study is to add the idea of
fuzzy relations to the GP method so as to
extend the possible solution space of supplier
selection problems. Sometimes, some
linguistic preferences among goals are not
easy to express by DMs, but different goals
relationships can affect the fitness of selected
suppliers for organizations. This study
proposes FGP for the problem of global
supplier selection in which linguistic
preferences among goals are allowable. With
the proposed method, DMs can set fuzzy
relations among goals according to their
supply chain strategy in a flexible and natural
way. In addition, a hypothetical case study is
presented to provide a better understanding
of the mechanism of the proposed FGP.
Considering different situations of decision
making, GP and FGP are both suitable for
individual DMs rational supplier evaluation.
FGP also can be adopted for rational multi-
DMs with the flexibility of adding supplier
selection goals and constraints. FGP can be
considered as a decision aid supporting DMs
supplier evaluation and reducing the
evaluation time that can be implemented
easily. AHP depends on decision makers
instinctive relative importance settings and it
can be implemented for individual or multi
persons housing decision making.
2-Multi choice goal programming
The MCGP method proposed by Chang
(2007) allows DMs to set multiple scalar
aspiration levels for each goal to avoid the
underestimation of goal setting in the initial
stage of decision problem. In order to solve
the above mentioned problem, the revised
MCGP approach has recently been
introduced by Chang (2008), which is
expressed as the following achievement
function.

5 May 3-5 2011 /
www.irtimco.com
(revised MCGP)
Min

+ +
=
+ + +
n
i
i i i i
n
i
i
e d d ) (e ) (
i
1

s.t.
n i g d d X f
i i i
,..., 2 , 1 , ) (
i
= = +
+
,
n i g
i i i
,..., 2 , 1 , g e e
max i,
= = +
+

n i g g g ,..., 2 , 1 ,
max i, i min i,
= s s
n i e d d
i i i
,..., 2 , 1 , 0 , e , ,
i
= >
+ +
F X e (F is a feasible set)
Where f
i
(x) is the linear function of x
1
,x
2
,
.,x
n
;d
i
+
,d
i
-
are over achievement and
under achievement of the i th goal; g
i
is the
interval aspiration level of the i th goal; e
i
+
,e
i
-
are the positive and negative deviations
attached to the i th goal;
min i, max i,
, g g are the
upper and lower bound of g
i
. The revised
MCGP represents a linear form of program
which can be easily solved using common
linear programming packages, not requiring
the use of integer programming packages.
3- Fuzzy Goal programming
Goal Programming is one of the most
powerful multi objective decision making
approaches. In a standard GP formulation,
goals and constraints are defined precisely.
In fact, very often it is a difficult job for a
decision maker to find out what attainments
are desired for each objective function.
Application of GP to the real life problems
may be faced with two important difficulties;
expressing the decision maker's vague goals
and/or constraints mathematically and
optimizing all goals simultaneously.
Goal programming (GP) is an analytical
multiple objectives decision making
approach designed to address decision-
marking problems in which targets have been
assigned to all attributes and where the
decision makers are interested in minimizing
the non-achievement of a particular goal
(Liao, 2009). The model can take into
account many simultaneous objectives as a
decision maker seeks the best solution from
among a set of feasible solutions. In
addition, GP is the most widely used
technique for solving multi criteria decision
problems as well as multi objective decision
problems in finding a set of satisfying
solutions. The main reason for its popularity
lies in its inherent flexibility that allows DMs
to formulate multi-objective decision
problems involving several criteria,
incomplete information, many decision
variables and resources constraints (Ura et
al., 2002).
The FGP can be expressed as follows.
A solution set X is found for FGP that
contains n fuzzy goals ) ( X f
i
:
i i
g X f > ) ( (or
i i
g X f s ) ( ),
i
=1,2,, n

6 May 3-5 2011 /
www.irtimco.com
s.t , F X e ( F is a feasible set)
where
i i
g X f ) ( ) ( s > indicates the i th fuzzy
goal approximately greater than or equal to
(approximately less than or equal to) the
aspiration level
i
g .
The FGP has the advantage of allowing for
the vague aspirations of DMs, which can be
qualified using some natural language or
vague phenomena. To represent the
preference concept of DMs, the preference-
based membership functions are introduced
below.
i i
i i
i i i
i i
i i
i i
i
g X f
l X f if
g X f l if
l g
l X f
g X f if
X >

s
s s

>
= ) ( for
) ( , 0
) ( ,
) ) ( (
) ( , 1
) (

>
s s s

s
=
i i
i i i i i i
i i
i i
i i
i
u X f if
g X f u X f g if
g u
X f u
g X f if
X
) ( , 0
) ( for ) ( ,
)) ( (
) ( , 1
) (
where
i
l and
i
u are, respectively, lower and
upper tolerance limit for the
i
th goal.
With the MAXMIN approach of
Zimmermann (1978), the FGP can be
expressed as follows:
max
. to subject
n i X f
i i
,..., 3 , 2 , 1 )) ( ( = s
b AX s
0 > X
Where ,
+
i
d and

i
d are an additional
continuous variable, over and under
achievement of the i th goal respectively.
4-Fuzzy multi choice goal programming
The fuzzy multi choice goal programming
model proposed by Chang (2007, 2008) for
the supplier selection is obtained as follows:
Min

+ +
=
+ + +
n
i
i i i i
n
i
i
e d d ) (e ) (
i
1
(1)
Subject to
n i g d d X f
i i i
,..., 2 , 1 , ) (
i
= = +
+
(2)
For maximization
n i g
i i i
,..., 2 , 1 , g e e
max i,
= = +
+

(3)
For minimization
n i g
i i i
,..., 2 , 1 , g e e
min i,
= = +
+

(4)

7 May 3-5 2011 /
www.irtimco.com
n i g g g ,..., 2 , 1 ,
max i, i min i,
= s s
(5)
For minimization
min , max ,
max ,
) (
i i
i i
i
g g
x f g
y

=
(6)
For minimization
min , max ,
min ,
) (
i i
i i
i
g g
g x f
y

=
(7)
ij j i
K y y =
Means goal i is more important than
goal j by given value K
ij
.
(8)
n i e d d
i i i
,..., 2 , 1 , 0 , e , ,
i
= >
+ +
(9)
F X e (F is a feasible set) (10)
Equations (1) to (5) are adopted from revised
MCGP (Chang, 2008). Equation (6) the
achievement rate is calculated for
maximization goals, Equation (7) the
achievement rate is also calculated for
minimization of goals. In Equation (8), a DM
can determine the satisfaction level of fuzzy
relations between goal i and goal j , where
the preemptive priority level of goal i is
higher than the preemptive priority level of
goal j . Kij represents the deviation between
goal i and goal j .
5-Case study
In order to compare the proposed FMCGP
method and the previous methods, this study
uses the example in the literature Tiwari et
al., (1987), Chen and Tsai (2001) and Akz
and Petrovic (2007). The fuzzy goals and
constraints are as follows.

Find x
Goal 1
35 8 2 4
4 3 2 1
s + + + x x x x
Goal 2
100 2 6 7 4
4 3 2 1
> + + + x x x x
Goal 3
120 10 5 6
4 3 2 1
> + + x x x x
Goal 4 70 2 3 5
4 2 1
> + + x x x
Goal 5
40 4 4 4
4 3 2 1
> + + x x x
Subject to
98 2 3 5 7
4 3 2 1
s + + + x x x x
117 6 2 7
4 3 2 1
s + + + x x x x
130 6 2
4 3 2 1
s + + + x x x x
105 6 9
3 2 1
s + + x x x
4 ,... 2 . 1 , 0 = > i x
i
The tolerance limits for the five fuzzy goals
were as follows:
U1 = 55, L2 = 40, L3 = 70, L4 = 30, L5 =10.
In Tiwari et al. (1987) , Chen and Tsai
(2001), and Akz and Petrovic (2007), this

8 May 3-5 2011 /
www.irtimco.com
problem was solved where the following
preemptive priority levels were set:
Level 1: Goal 1 and Goal 3;
Level 2: Goal 2;
Level 3: Goal 4 and Goal 5.
The achievement rate are obtained in Tiwari
et al. (1987) , Chen and Tsai (2001), Akz
and Petrovic (2007), and FMCGP are
presented in Table 1. The resulting
achievement rate for the five goals are
(0.9,0.89,0.9,0.68,0.84) in FMCGP.
Compared with the result of Chen and Tsai,
the achievement rate of G3 is higher than G2
in FMCGP. The preemptive priority among
G3 and G2 is achieved better in FMCGP.
Compared with the result of Akz and
Petrovic (2007), the sum of achievement rate
of five goals is greater in FMCGP.
With FMCGP, DMs can set multiple goals
with fuzzy relationship, each of goals
mapping multiple aspiration levels. In the
computational process, goals or constraints
can be added/ deleted in any stages of the
solution process. Therefore, the proposed
FMCGP method is suitable for supplier
selection.
Table 1. Results obtained from the four methods
method x
1
x
2
x
3
x
4
y
1
y
2
y
3
y
4
y
5

i
i
y
Tiwari et al. (1987) 0.00 9.75 0.00 15.87 0.98 1.00 0.60 0.77 0.96 4.31
Chen and Tsai (2001) 0.00 8.26 0.30 16.12 1.00 0.86 0.86 0.68 0.81 4.21
Akz and Petrovic (2007) 0.00 7.48 0.47 16.25 1.00 0.80 1.00 0.62 0.73 4.15
FMCGP 0.00 8.22 0.56 16.13 0.90 0.89 0.90 0.68 0.84 4.21
6-Conclusion
This paper proposes a method called FMCGP
to aid DMs to set the satisfaction level of
fuzzy relationship among goals in describing
their ambiguous selection preferences.
Imprecise importance relations among goals
are modeled using fuzzy membership
functions. Moreover, the FMCGP
methodology allows DMs to set priority on
individual goals in supplier selection
decision with different supply chain
strategies. DMs can determine the low bound
satisfaction level on priority relations among
goals to obtain better suppliers according to
different supply chain strategies. The setting

9 May 3-5 2011 /
www.irtimco.com
of low bound satisfaction levels can ensure
that the satisfaction level achieves certain
level. Only minimum changes of coefficients
are needed in the proposed method, FMCGP,
while organizations change their supply
chain strategies to be suitable for different
supplier selection problems. Moreover,
organizations can simulate the supplier
selection process to test the average cost and
average delivery efficiency from different
suppliers in combination with FMCGP
before choosing a supply chain strategy in
supplier selection problems. This study
provides a hypostatical case study to
demonstrate how to select appropriate
suppliers and determine the optimal order
quantity from them with fuzzy importance
relations. The result from the fuzzy
importance relation setting in FMCGP
corresponds more closely with DMs
supplier selection goals. In this study, we
propose a simple practical method to aid
DMs to construct their preferences and
determine the weights on their decision
goals. The proposed model provides a
flexible method for DMs to assign suitable
fuzzy relationships among goals and
increases the probability of making the
appropriate decisions for their supplier
selection problems.
References
1. Akz, O., & Petrovic, D. (2007). A
fuzzy goal programming method with
imprecise goal hierarchy. European
Journal of Operational Research, 181(3),
14271433.
2. Chan, F. T. S., & Kumar, N. (2007).
Global supplier development
considering risk factors using fuzzy
extended AHP-based approach.
Omega, 35(4), 417431.
3. Chang, C.T. (2007). Multi-choice goal
programming. Omega 35(4), 389-396.
4. Chang, C.T. (2008). Revised multi-
choice goal programming. Applied
Mathematical Modelling, 32, 2587
2595.
5. Charnes, A. & Cooper, W.W. (1961).
Management model and industrial
application of linear programming.
New York: Wiley.
6. Chen, L.-H., & Tsai, F.-C. (2001).
Fuzzy goal programming with different
importance and priorities. European
Journal of Operational Research, 133(3),
548556.
7. Curry, B., & Lazzari, L. (2009). Fuzzy
consideration sets: a new approach
based on direct use of consumer
preferences. International Journal of
Applied Management Science, 1(4),
420436.

10 May 3-5 2011 /
www.irtimco.com
8. Dobler, D. W., Lee, L., & Burt, N.
(1990). Purchasing and materials
management: text and cases. New
York: McGraw-Hill.
9. Fazlollahtabar, H. (2008). Applying
Multiple-Criteria Decision Making
methods for developing IT industry.
Int. J. Information and Decision
Sciences, 1(1), 115131.
10. Ghodsypour, S. H., & OBrien, C.
(1998). A decision support system for
supplier selection using an integrated
analytic hierarchy process and linear
programming. International Journal of
Production Economics, 56-57(20), 199
212.
11. Ignizio, J.P. (1976). Goal programming
and extensions. Lexington Books,
Lexington, MA.
12. Iyer, A. V., & Bergen, M. E. (1997).
Quick Response in Manufacturer-
Retailer Channels. Management
Science, 43(4), 559570.
13. Liao, C.N. (2009). Formulating the
multi-segment goal programming.
Computers & Industrial Engineering 56,
138-141.
14. Lin, C.-C. (2004). A weighted max
min model for fuzzy goal
programming. Fuzzy Sets and Systems,
142(3), 407420.
15. Liou, J. J. H., & Chuang, M.-L. (2008).
A hybrid MCDM model for evaluating
the corporate image of the airline
industry. Int. J. Applied Management
Science, 1(1), 4154.
16. Min, H. (1994). International supplier
selection: A multi-attribute utility
approach. International Journal of
Physical Distribution & Logistics
Management, 24(5), 2433.
17. Porter, M. E. (1998). Competitive
strategy: techniques for analyzing
industries and competitors. New York:
The Free Press
18. R.N. Tiwari, S. Dharmar and J.R. Rao.
(1987), Fuzzy goal programming - an
additive model. Fuzzy Sets Systems
24, 27 -34.
19. Romero, C. (2001). Extended
lexicographic goal programming: a
unifying approach. Omega 29(1), 63-
71.
20. Schonberger, R. J. (1982). Japanese
Manufacturing Techniques: Nine hidden
lessons in simplicity. New York: The
Free Press
21. Tamiz, M., Jones, D. & Romero, C.
(1998). Goal programming for decision
marking: an overview of the current
state-of-the-art. European Journal of
Operational Research 111(3), 567-581.
22. Ura, M. VR, Caballero, R, Ruiz, F &
Romero, C.(2002). Meta-goal
programming. European Journal of
Operational Research 136, 422-429
23. Willis, H. T., Huston, R. C., &
Pohlkamp, F. (1993). Evaluation

11 May 3-5 2011 /
www.irtimco.com
measures of just in time supplier
performance. Production and Inventory
Management Journal, 34(2), 15.
24. Zimmermann, H.J. (1978). Fuzzy
programming and linear programming
with several objective functions. Fuzzy
Sets and Systems 1, 45-55.
Abbreviations
n number of goals
d
i
+
over-achievements of the i
th
goal
d
i
-
under-achievements of the i
th
goal
e
i
+
over-achievement of the highest
(lowest) possible value of
membership function
e
i
-
under-achievement of the highest
(lowest) possible value of
membership function

i
the positive weights attached to
deviations d
i
+
in the objective
function

i
the positive weights attached to
deviations d
i
-
in the objective
function
) (X
i

a membership function of i
th
goal,
i = 1,2,..,n
l
i
lower limits for the i
th
goal
u
i
upper limits for the i
th
goal
t
i
the target aspiration level of the i
th
goal in FGP
K
ij
the deviation between goal i and
goal j
l
ij
lower limits of the fuzzy
relationship among goal i and goal j
t
ij
the target aspiration level of the
fuzzy relationship among goal i and
goal j
g
i
the aspiration level of the i
th
goal
in FMCGP
g
i,min
the lower bound aspiration for the i
th
goal
g
i,max
the upper bound aspiration for the i
th
goal
y
i
the ratio of the achievement in the i
th
goal
y
j
the ratio of the achievement in the j
th
goal
an additional continuous variable in
FGP

You might also like