You are on page 1of 6

15

International Journal of Algorithms, Computing and Mathematics


Volume 2, Number 3, August 2009
Eashwar Publications
Risk analysis of Decision makers attitude in Fuzzy PERT
G.Uthra
Research scholar, SCSVMV University, Department of Mathematics,
Saveetha Engineering College, Chennai, India.
Uthragopalsamy@yahoo.com
R.Sattanathan
P.G and Research Department of Mathematics, D.G.Vaishnav College, Chennai, India.
rsattanathan@gmail.com
Abstract
Among the problems in Operations Research, Scheduling problems are those with a lot of
applications. The duration in reality is often imprecise and the imprecision in the data is
critical for the scheduling procedures. Hence fuzzy approach draws active attention. In
this paper, we have discussed the risk analysis of Decision makers attitude in the FPERT
networks. The approach is illustrated by a numerical example.
Key words: FPERT, Triangular fuzzy numbers, Trapezoidal fuzzy numbers, Linguistic variable.
1. Introduction
The deterministic version of the PERT algorithms is well known for many years already.
Anyhow, the determinism of the model can be questioned in many practical situations.
The main goal of this paper is to generalize the well known model and include uncertain and
vague phenomena. This uncertainty is not based on statistical dispersion of possible values and,
consequently, application of probabilistic methods is not effective. Classical project scheduling
methods such as CPM, PERT etc., involve many drawbacks in estimating the duration of the
activities and they lack the ability of modeling practical projects. Fuzzy theory can be applied
effectively to deal with such uncertainties.
Fuzzy Set theory was introduced by L.A.Zadeh in 1965 and this opened promising new horizons
to different scientific areas. Since then it has been finding its way in almost all fields where
uncertainty prevails.
Classical PERT deals with problems of quantitative data and it does not support if the data is
qualitative. In other words, it fails if the data is expressed in linguistic terms like very fast,
extremely fast, etc. Hence the application of fuzzy theory is mandatory and the problem can be
solved by Fuzzy PERT (FPERT).
International Journal of Algorithms, Computing and Mathematics
16
In this paper, we consider a PERT situation in a fuzzy environment. We extend the method using
triangular fuzzy numbers dealt in [6] to trapezoidal fuzzy numbers. We then discuss the risk
analysis of Decision makers attitude in the FPERT networks. The method is illustrated by a
numerical example.
2. Definitions and Formulations
Definition 2.1 A Triangular fuzzy number can be defined by a triplet (a, b, c) whose membership
function is given by
0, x < a
(x a)/(b-a) b x a s s
) (x = (x c)/(b-c) c x b s s
0, x > c
Definition 2.2 A trapezoidal fuzzy number is a fuzzy number (a
1
, a
2
, a
3
, a
4
) and its membership
function is defined as

(x) = (x - a
1
) / (a
2
- a
1
) if a
1
x a
2
1 if a
2
x a
3
(x a
4
) / (a
3
a
4
) if a
3
x a
4

0 otherwise
From the definition of triangular and trapezoidal fuzzy numbers it is clear that the triangular
fuzzy number is special case of trapezoidal number.
Definition 2.3. A linguistic variable is variable whose values are linguistic terms.
The concept of a linguistic variable is very useful in dealing with situations which are too
complex to be reasonably described in conventional quantitative expressions. For example,
height is a linguistic variable its values can be very high, high, medium, low, very low etc. The
algorithm dealt in [6] for triangular fuzzy numbers is being extended for trapezoidal fuzzy
numbers.
The Step by step procedure of the algorithm is given as follows:
- Construct the membership function =
a b
a x

--------------(1)
where a is the lowest time estimate and b is the highest time estimate.
- Convert the linguistic expressions (qualitative data) in to quantitative data in
the form (A,B,C), where A,B,C are numerical values.
- Express the situation in the matrix form as below.
(
(
(
(

.. .. ..
9 8 7
6 5 4
3 2 1
a a a
a a a
a a a

- Form the membership matrix
A
using (1).
- Find the normalized matrix.
- Find the minimum value in the matrix for each activity (i-j).
i-j
i-k
j-k
.
Risk analysis of Decision makers attitude in Fuzzy PERT
17
- Consider the various paths of the project and sum up the minimum values of
the corresponding paths. The path with maximum sum is the critical path.
The method mentioned above can be extended to trapezoidal fuzzy numbers since trapezoidal
fuzzy numbers are more stable than triangular fuzzy numbers. For example, consider the average
marks secured by a student in an examination. The linguistic term average marks can be
represented as (50, 60, 70) using triangular fuzzy numbers. From this we can say that average
mark secured by a student is 60. But the same can be represented as (50, 60, 65, 70) using
trapezoidal fuzzy numbers. From this we say that the average marks of secured by a student
ranges between 60 and 65. The advantage of using trapezoidal fuzzy numbers is that it attains
the maximum membership value 1 between two points (i.e.) in an interval whereas a triangular
fuzzy number attains the same at only one point.
The concept using trapezoidal fuzzy numbers as the activity duration presented by Dubois and
Prade[2] is applied to a FPERT network.
Example
Fig.1
The critical path cannot be found until the linguistic expressions are converted in to quantitative
values. For that we use the following linguistic table.
Table-1
Linguistic expressions Value
Absolutely fast (1 2 3)
Fast (4 5 6)
Medium (7 8 9)
Slow (10 11 12)
1
2
3
4
(5,10,10,15) (6,8,10,11)
(Fast, Medium, Slow,
Slow)
(Absolutely fast, Fast,
Fast, Medium)
(1,3,5)
(2,4,5,6)
International Journal of Algorithms, Computing and Mathematics
18
Absolutely slow (13 14 15)
In this problem we have used five as the maximum number of intervals to form the table.
Minimum time estimate is 2 and maximum time estimate is 15 and hence we construct the
membership function using (1) as = (x-2)/(15-2) = (x-2)/13
(1) Express the situation in matrix form as given below.
2 5 5 8
5 10 10 15
5 9 10 11
2 4 5 6
6 8 10 11
(2) Form the membership matrix
A
0 3/13 3/13 6/13
3/13 8/13 8/13 1
3/13 7/13 8/13 9/13
0 2/13 3/13 4/13
4/13 6/13 8/13 9/13

(3) Find the normalized matrix.
0 0.115 0.1 0.1463
0.3 0.308 0.267 0.3171
0.3 0.269 0.267 0.2195
0 0.077 0.1 0.0976
0.4 0.231 0.267 0.2195
(4) Take the minimum value in each row.
Table.2
Activity 1-2 1-3 2-3 2-4 3-4
Min.Value 0 0.267 0.2195 0 0.2195
(5) Find the sum of minimum values for each path.
Sum for the path 1-2-4 = 0 + 0 = 0
Sum for the path 1-3-4 = 0.267 + 0.2195 = 0.4865
Sum for the path 1-2-3-4 = 0 + 0.2195 + 0.2195 = 0.4390
Maximum Sum = 0.4865. Hence the critical path is 1-3-4.
3. Risk analysis
1-2
1-3
2-3
2-4
3-4
A =
1-2
1-3
2-3
2-4
3-4

A
=
1-2
1-3
2-3
2-4
3-4
Risk analysis of Decision makers attitude in Fuzzy PERT
19
The method presented in the previous section gives the critical path in a FPERT network. But
this does not explain the decision makers attitude towards risk and uncertainty. This can be
achieved by applying linguistic approach to those problems.
Risk analysis of Decision makers attitude in FPERT network uses the notion of extreme
optimism and pessimism. We consider the linguistic terms of risk attitudes for trapezoidal fuzzy
numbers adopted in [5]
Table -3
Linguistic terms Trapezoidal fuzzy numbers (a
1
, a
2
, a
3,
a
4
)
For benefit criteria For cost criteria
Absolutely optimistic (a
1
, a
3
, a
4
, a
4
) (a
1
, a
1
, a
2
, a
4
)
Very optimistic (a
1
, [a
2
+3a
3
] /4, [a
3
+3a
4
] /4, a
4
) (a
1
, [a
2
+3a
1
] /4, [a
3
+3a
2
] /4,a
4
)
Optimistic (a
1
, [a
2
+a
3
] /2, [a
3
+a
4
] /2, a
4
) (a
1
, [a
2
+a
1
] /2, [a
3
+a
2
] /2,a
4
)
Fairly optimistic (a
1
, [3a
2
+a
3
] /4, [3a
3
+a
4
] /4, a
4
) (a
1
, [3a
2
+a
1
] /4, [3a
3
+a
2
] /4,a
4
)
Neutral (a
1
, a
2
, a
3
, a
4
) (a
1
, a
2
, a
3
, a
4
)
Fairly pessimistic (a
1
, [3a
2
+a
1
] /4, [3a
3
+a
2
] /4, a
4
) (a
1
, [3a
2
+a
3
] /4, [3a
3
+a
4
] /4,a
4
)
Pessimistic (a
1
, [a
2
+a
1
] /2, [a
3
+a
2
] /2, a
4
) (a
1
, [a
2
+a
3
] /2, [a
3
+a
4
] /2,a
4
)
Very pessimistic (a
1
, [a
2
+3a
1
] /4, [a
3
+3a
2
] /4, a
4
) (a
1
, [a
2
+3a
3
] /4, [a
3
+3a
4
] /4,a
4
)
Absolutely pessimistic (a
1
, a
1
, a
2
, a
4
) (a
1
, a
3
, a
4
, a
4
)
Since critical path is the path of longest duration, it comes under benefit criteria. Using the
trapezoidal fuzzy numbers from the table-3 we get the following results.
Table-4
Decision makers attitude Maximum
Sum
Critical
path
Absolutely optimistic 0.4760 1-3-4
Very optimistic 0.4953 1-3-4
Optimistic 0.5052 1-3-4
Fairly optimistic 0.5019 1-3-4
Neutral 0.4865 1-3-4
Fairly pessimistic 0.4954 1-3-4
Pessimistic 0.5052 1-3-4
Very pessimistic 0.4954 1-3-4
Absolutely pessimistic 0.5195 1-3-4
From the table, any decision maker ranging from and extreme optimist to an extreme pessimist
will choose 1-3-4 as the critical path.
Conclusion
FPERT deals with the uncertainties prevailing in the PERT network. In this paper, a new
method based on linguistic expressions is presented for project scheduling in a fuzzy
environment. We have also expressed the risk analysis of the Decision makers attitude in the
context of FPERT networks. We hope this method will be useful to tackle problems of FPERT
networks involving linguistic Variables.
International Journal of Algorithms, Computing and Mathematics
20
References
[1]. Ahmad Soltani and Rasoul Haji, A project scheduling method based on Fuzzy theory, Journal of Industrial
and Systems Engineering, Vol.1, No.1.pp.70-80.
[2]. Dubois.D and Prade.H, 1985, An Approach to Computerised Processing of Uncertainty, Plenum Press, New
York.
[3]. George.J.Klir and Bo Yuan, 1995, Fuzzy sets and Fuzzy logic Prentice Hall of India, 2001.
[4]. Norman Fenton and Wei Wang, 2006, Risk analysis for multi criteria decision making, Knowledge based
systems, 19, pp. 430 437.
[5]. Uthra.G and Sattanathan.R, 2008, Risk analysis for Multi Criteria Decision Making using trapezoidal fuzzy
numbers, Proceedings of the National Conference on Fuzzy Mathematics and Graph theory, pp.168-174.
[6]. Uthra.G and Sattanathan.R, 2008, An analysis of Decision makers attitude in the Fuzzy PERT network,
Vol.2, pp. 289-293.