Optimal Order Quantity of an EOQ Model Using
Expected Value of a Fuzzy Function
P. K. De1 , Apurva Rawat2
1. Department of Mathematics, National Institute of Technology, Silchar, 788010, Assam, India.
2. AIM & ACT, Banasthali University, Banasthali, 304022, Rajasthan, India.
Emails : 1. pijusde@rediffmail.com 2. apurva rawat@rediffmail.com
Abstract—In this paper, the optimization of a fuzzy inventory
model under fuzzy demand is done. Due to uncertainty in
demand, we can express it using some fuzzy distribution which
depends on the inventory manager. The fuzzy demand leads to a
fuzzy total cost function of the inventory problem. The paper uses
three types of fuzzy numbers to represent the uncertain demand
and the expected value of the total cost function is minimized
so as to find the optimal value of order quantity using each.
The expected value of the total cost function is based on the
lambda pessimistic and optimistic values of the fuzzy variables.
The formulae for optimal order quantity are then derived. The
results are justified with the help of some numerical illustrations.
The methodology of the paper can also be applied to other
inventory problems under uncertain factors affecting the total
cost of the problem.
Keywords: Fuzzy Inventory·Order Quantity·Fuzzy
variable·L-R type fuzzy number·Fuzzy expectation·Optimal
order quantity·lambda pessimistic value·Lambda optimistic
value
I. I NTRODUCTION
When uncertainty arises in data due to randomness in the
data, probability theory is used to represent and analyze it but
when uncertainty is due to fuzziness in data, then fuzzy set
theory best suited. In various real life applications of inventory
theory, uncertainty arises in the factors affecting the model
such as demand, associated costs, lead time etc.. The problem
tries to find out the optimal order quantity under the given
uncertain data and given set of conditions, i.e., the quantity
which should be ordered at certain intervals of time so that
the total cost is minimum calculated for the whole time cycle.
By minimizing the total cost, a person can gain a maximum
profit in his business or industry.
The above kind of inventory problems exist in retail mar-
kets, small scale industries, storage of goods in godowns,
construction and so on, where the problem is that the storage
cost of much large amount of inventory causes increased
carrying cost and small amount of inventory causes increased
shortage cost. Frequent orders to set up the inventory again
also adds up to the increase in total cost of the business.
The use of fuzzy numbers in fuzzy inventory models has
been done since a long time. Fuzzy set theory is applied to
represent uncertainties arising in demand or the cost coef-
ficients. By using it, one can easily quantify the vagueness
and imprecision in data. An extensive work can be found in
[14] related to the application of fuzzy set theory in inventory
problems. Park[1] used arithmetic operations under function
principle to deal with fuzzy costs in their inventory problem.
Chen et al.[2] used function principle to deal with the EOQ
problem. Gen et al.[3] uses triangular fuzzy numbers to repre-
sent the uncertain data in his model and applies interval mean
method to find their means. To represent the fuzzy demand
Yao et al.[4] uses triangular fuzzy number in an economic
order quantity(EOQ) model without backordering. Yao and
Lee[5] uses trapezoidal fuzzy numbers in EOQ problems with
and without backordering. Chang[6] used extension principle
and centroid method to estimate the total cost in a production
inventory model. Chen and Li[7] show how to treat the
defuzzification of fuzzy numbers with exponential membership
function by graded mean integration representation method.
Yao and Wu[8] construct a new ranking system for fuzzy
numbers by using the decomposition principle and the crisp
ranking system on R. Hseih[9] uses trapezoidal fuzzy number
to represent the fuzzy demand and optimizes his model under
k-preference. Hseih[10] obtained the optimal order quantity
by using both function principle and graded mean integration
representation method for both computing and representing
fuzzy total annual inventory cost.
The aim of this paper is to illustrate the application of
various L-R types of fuzzy numbers in fuzzy inventory models
using their lambda pessimistic and optimistic values. To deter-
mine the optimal order quantity that minimizes the total cost
function the expected value of a function of fuzzy variable
defined by [11] is used.
In the second section of the paper, a brief review of the
fuzzy numbers and their expectation is done. The third section
computes the lambda pessimistic and optimistic values of L-R
type fuzzy numbers. Then the inventory model is formulated
in first part of section four. The other part comprises of the
fuzzification of the assumed inventory model. Results are
illustrated using numerical examples in section five and a
conclusion is derived in the last section.
II. P RELIMINARIES
The term ”fuzzy” was introduced in fuzzy set theory by
Lotfi A. Zadeh in 1965. It is a form of many-valued logic
based on reasoning that is not fixed or exact. In the field
of computation fuzzy logic variables may have a truth value
that ranges in degree between 0 and 1, i.e., fuzzy logic has
been extended to handle the concept of partial truth, where the
truth value may range between completely true and completely where, a, b, c ∈ R and is represented graphically as
false. Fuzzy logic has been applied to many fields, from
inventory theory, networking to artificial intelligence. It uses
an imprecise but very descriptive language to deal with input
data more like a human operator. It is a way of processing
data by allowing partial set membership rather than crisp.
Fuzzy sets are the sets whose elements have degrees of
membership in that set. The notion of fuzzy sets were intro-
duced as an extension of the classical sets. In classical set
theory, the membership of elements in a set is in the form of Fig 1. Membership function of triangular fuzzy number A
in binary terms- an element either belongs or does not belong 2) Trapezoidal Fuzzy number: A trapezoidal fuzzy number
to the set. By contrast, fuzzy set theory permits the gradual is denoted by A= (a, b, c, d) and defined by the following
assessment of the membership of elements in a set; this is membership function
described with the aid of a membership function valued in the
real unit interval [0,1]. Fuzzy sets generalize classical sets,

 0, x ≤ a
since the indicator functions of classical sets are special cases x−a

 b−a , a ≤ x ≤ b


of the membership functions of fuzzy sets, if the latter only µA (x) = 1, b ≤ x ≤ c (2)
take values 0 or 1. In fuzzy set theory, classical bivalent sets d−x
 d−c , c ≤ x ≤ d


are usually called crisp sets. The fuzzy set theory can be used

0, x ≥ d

in a wide range of domains in which information is incomplete
or imprecise. where, a, b, c ∈ R and is represented graphically as

A. Fuzzy number
A fuzzy number is a convex normalized fuzzy set whose
membership function is piecewise continuous. Suppose X̃ is a
generalized fuzzy number (known as L-R type fuzzy number),
whose membership function µX̃ (x) satisfies the following
conditions with 0 < w ≤ 1 and −∞ < l < m < n < u < ∞;
1) µX̃ (x) is a continuous mapping from <, to the closed
interval [0,1], Fig 2. Membership function of trapezoidal fuzzy number A

2) µX̃ (x) = 0, −∞ < x < l, 3) Exponential Fuzzy number: An exponential fuzzy num-
3) µX̃ (x) = L(x), is strictly increasing on [l, m], ber is denoted by A = (a, b, α, β; w); 0 < w ≤ 1 and defined
4) µX̃ (x) = w, m < x < n, by the following membership function
5) µX̃ (x) = R(x), is strictly decreasing on [n, u],  a−x
6) µX̃ (x) = 0, u < x < ∞.  we− α , x ≤ a

This type of generalized fuzzy number is denoted as µA (x) = w, a ≤ x ≤ b (3)
µX̃ (x) = (l, m, n, u; w)LR . When w = 1, it can be written as  we− x−b

β ,x ≥ b

µX̃ (x) = (l, m, n, u)LR .

where, a, b, c ∈ R; α, β are positive real numbers. A
B. Some L-R types of fuzzy numbers = (a, b, α, β)E denotes exponential fuzzy number when w=1,
In the most the real life applications fuzzy numbers are represented graphically as
one way to describe the data vagueness and imprecision.
They can be regarded as an extension of the real numbers.
In the literature of fuzzy sets there is some shortcoming in
the definition of fuzzy number. Frankly some authors do not
distinguish between positive and nonnegative fuzzy numbers.
Here, we focus on mainly three types of fuzzy numbers viz.
Fig 3. Membership function of exponential fuzzy number A
the triangular, trapezoidal and exponential fuzzy number.
1) Triangular Fuzzy number: A triangular fuzzy number
is denoted by A = (a,b,c) and defined by the following C. λ-pessimistic and λ-optimistic value of a fuzzy variable
membership function Let ξ be a fuzzy variable and λ ∈ (0, 1], then
ξλ0 = inf {x | µ(x) ≥ λ}

 0, x ≤ a
 x−a , a ≤ x ≤ b and ξλ00 = sup{x | µ(x) ≥ λ}

µA (x) = b−a (1)
c−x are called λ−pessimistic and λ−optimistic values of fuzzy
,b ≤ x ≤ c
 c−b


0, x ≥ c variable ξ.
D. Expectation of a function of a fuzzy variable 2) Trapezoidal Fuzzy Number: For a trapezoidal fuzzy
number ξ = (a, b, c, d) and λ ∈ (0, 1] as shown:
For the fuzzy variable ξ and a strictly increasing function
f : < → <, if the Lebesque integrals
Z 1 Z 1
f (ξλ0 ) dλ and f (ξλ00 ) dλ (4)
0 0

are finite, then the expectation of f (ξ), i.e.,

Z 1
1
E[f (ξ)] = [f (ξλ0 ) + f (ξλ00 )] dλ (5)
2 0
Fig 6. Graphical representation of L−1 (λ) and R−1 (λ) for trapezoidal
III. M ETHODOLOGY fuzzy number ξ

A. λ-pessimistic and λ-optimistic value of L-R type fuzzy

numbers L−1 (λ) = a + λ(b − a) (8)
−1
Let ξ be a generalized fuzzy variable (L-R type fuzzy & R (λ) = d − λ(d − c) (9)
number) and λ ∈ (0, 1] as shown:
3) Exponential Fuzzy Number: For an exponential fuzzy
number ξ = (a, b, α, β) and λ ∈ (0, 1] as shown:

Fig 7.Graphical representation of L−1 (λ) and R−1 (λ) for exponential
fuzzy number ξ

Fig 4. Membership function of generalized fuzzy number ξ L−1 (λ) = a + αlnλ (10)
Then ξλ0 = L−1 (λ) and ξλ00 = R−1 (λ) & R−1 (λ) = b − βlnλ (11)

B. Calculation of L−1 (λ) and R−1 (λ) for various fuzzy
numbers
1) Triangular Fuzzy Number: For a triangular fuzzy num-
ber ξ = (a, b, c) and λ ∈ (0, 1] as shown:
IV. M ODEL F ORMULATION
An inventory model under fuzzy demand without shortages
is considered under this section. The optimal order quantity
is computed initially under crisp environment and later on
in fuzzy environment by taking the fuzzy demand as various
types of fuzzy numbers.

A. Notations Used

C : carrying cost
O : ordering cost
D : total demand
T : total time
Fig 5. Graphical representation of L−1 (λ) and R−1 (λ) for triangular fuzzy
number ξ
tq : time of one order cycle
D̃ : fuzzy total demand
Q : order quantity
L−1 (λ) = a + λ(b − a) (6) Q∗ : optimal order quantity
−1 TC : total cost
& R (λ) = c − λ(c − b) (7)
B. Model 1: Inventory model under fuzzy demand without 1) For triangular Fuzzy Number:
shortages in Crisp Environment Z 1
dE[T C] C O
In this model, the economic lot size is obtained by the = − [a + λ(b − a) + c − λ(c − b)] dλ
Q 2 2Q2 0
following model equation Z 1
dE[T C] C O
r
2OD Q D = − [(a + c) + (b − a − c + b)] λ dλ
Q= ; where = Q 2 2Q2 0
CT tQ T dE[T C] C O
= − [(a + 2b + c)]
The total cost for the period [0; T] is given by Q 2 4Q2
r
total cost = F(q) = carrying cost + ordering cost O
⇒ Q∗ = (a + 2b + c) (15)
CT Q OD 2C
T C(C, O) = T C(Q) = + (12)
2 Q 2) For trapezoidal Fuzzy Number:
The crisp optimal solution is: Z 1
dE[T C] C O
optimal order quantity = − [a + λ(b − a) + d − λ(d − c)] dλ
r Q 2 2Q2 0
Z 1
2OD dE[T C] C O
Q∗ = (13) = − [(a + d) + (b − a − d + c)] λ dλ
CT Q 2 2Q2 0
and minimum total cost dE[T C] C O
= − [(a + b + c + d)]
√ Q 2 4Q2
T Cmin = 2ODCT (14) r
O
⇒ Q∗ = (a + b + c + d) (16)
Diagrammatic representation: 2C
3) For exponential Fuzzy Number:
Z 1
dE[T C] C O
= − [a + αlnλ + b − βlnλ] dλ
Q 2 2Q2 0
Z 1
dE[T C] C O
= − [(a + b) + (α − β) lnλ dλ
Q 2 2Q2 0
dE[T C] C O
Fig 8. Variation of stocked quantity w.r.t timeξ = − [(a + b) − (α − β)]
Q 2 2Q2
Example 1: Let O = 20; C = 12; D = 500; T = 1. Then r
O
Q=40.8 ≈ 41; TC=489.8 ≈ 490. ⇒ Q∗ = (a + b + β − α) (17)
C
C. Model 2: Inventory model under fuzzy demand without V. N UMERICAL I LLUSTRATION
shortages in Fuzzy Environment
Let D̃ be a generalized fuzzy number. When the demand Consider the problem in example 1 taking the demand D as
is uncertain, the total cost function also becomes a fuzzy. fuzzy number.
Assuming T as unity TC in (12) is transformed according to 1) Let D = (400, 500, 800) be a triangular fuzzy number.
(5) as shown: Then from (15), we get Q* = 42.8 ≈ 43.
2) Let D = (400, 500, 600, 800) be a trapezoidal fuzzy
CQ number. Then from (16), we get Q* = 43.78 ≈ 44.
TC = ⊕ O ⊗ (D̃  Q)
2 3) Let D = (500, 600, 25, 5) be an exponential fuzzy
1 1 CQ L−1 (λ) R−1number.
Z
CQ (λ)
E[T C] = [( +O× )+( +O× )] dλThen from (17), we get Q* = 43.2 ≈ 43.
2 0 2 Q 2 Q
Z 1
CQ O
= + [L−1 (λ) + R−1 (λ)] dλ
2 2Q 0 VI. C ONCLUSION
Z 1
dE[T C] C O
= − [L−1 (λ) + R−1 (λ)] dλ An inventory problem deals with finding the optimal order
Q 2 2Q2 0
quantity to minimize the total cost. With this aim, the expected
The problem is to determine the optimal order quantity (Q*) value of a fuzzy function is used. The methodology of the
that minimizes the fuzzy total cost. Hence, in the preceding paper used for optimization can be applied to various real life
subsection the optimal values of Q and TC are derived by problems where the factors affecting the cost of the business
substituting the values of L−1 (λ) and R−1 (λ) in above are vague or uncertain. The model can also be extended and
equation, for various types of fuzzy numbers viz. triangular, can be applied for further research and development in the
trapezoidal and exponential fuzzy number. areas of inventory management.
 R EFERENCES
[1] K. S. Park, “Fuzzy set theoretic interpretation of economic order quan-
tity”, IEEE Transactions on Systems, Man and Cybernetics, vol. 17, pp.
1082–1084, 1987.
[2] S. H. Chen, C. C. Wang and A. Ramer, “Backorder fuzzy inventory model
under function principle”, Information Sciences, An International Journal,
vol. 95, pp. 71–79, 1996.
[3] M. Gen, Y. Tsujimura and D. Zheng, “An application of fuzzy set theory
to inventory control models”, Computer and Industrial Engineering, vol.
33, no. 3-4, pp. 553–556, 1997.
[4] J. S. Yao, S. C. Chang and J. S. Su, “Fuzzy inventory without backorder
for fuzzy order quantity and fuzzy total demand quantity”. Computer and
Operations Research, vol. 27, no.10, pp. 935–962, 1999.
[5] J. S. Yao and H. M. Lee, “Fuzzy inventory with or without backorder
for fuzzy order quantity with trapezoidal fuzzy number”, Computer and
Operations Research, vol. 27, no. 10, pp. 935–962, 1999.
[6] S. C. Chang, “Fuzzy production inventory for fuzzy product quantity with
triangular fuzzy number”, Fuzzy Sets and Systems, vol 107, pp. 37–57,
1999.
[7] S. H. Chen and G. C. Li, “Representation, Ranking, and Distance of Fuzzy
Number with Exponential Membership Function Using Graded mean
Integration method”, Tamsui Oxford Journal of Management Sciences,
vol. 16, no. 2, pp. 123–131, 2000.
[8] J. S. Yao and K. M. Wu, “Ranking fuzzy numbers based on decomposition
principle and signed-distance”, Fuzzy Sets and Systems, vol. 116, pp. 275–
288, 2000.
[9] C. H. Hseih, “Optimization of Fuzzy Inventory Models under k-
preference”, Tamsui Oxford Journal of Mathematical Sciences, vol. 19,
no. 1, pp. 79–97, 2003.
[10] C. H. Hseih, “Optimization of fuzzy inventory models under fuzzy
demand and fuzzy lead time”, Tamsui Oxford journal of management
sciences, vol. 20, no. 2, pp. 21–36, 2004.
[11] D. H. Hong, “Note on the expected value of a function of fuzzy
variable”, Journal of Applied Mathematics and Informatics, vol. 27, no
3, pp. 773–778, 2009.
[12] D. Dubois and H. Prade, Fuzzy Sets and Systems. New York, Academic
Press, 1980.
[13] H. J. Zimmermann, Fuzzy Set Theory and its Applications. 2nd Ed.,
Boston/Dorgrecht/London, Kluwer Academic Publishers, 1991.
[14] A. L. Guiffrida, Inventory Management: Non-classical Views(Chapter
8), M. Y. Jaber(Ed.), Boca Raton, FL: CRC, 2010.