International Conference on Electrical, Electronics, and Optimization Techniques (ICEEOT) - 2016

Contingent Demand Fulfilment Decision Problem from Supplier’s

Perspective

Akhilesh Kumar Anjana Gupta Aparna Mehra

Department of Applied Mathematics Department of Applied Mathematics Department of Mathematics
Delhi Technological University Delhi Technological University Indian Institute of Technology, Delhi
Delhi, India Delhi, India Delhi, India
[email protected] [email protected] [email protected]

Abstract— The planning of production and other operations received by these firms or a high fluctuation in their product
in a firm is done based on demand forecasts. But production demand for a short duration.
decisions get affected by actual realization of the demand. These
decisions are taken keeping into consideration production When this sudden or contingent demand arises on
capacities. Sometimes sudden and unforeseen demands from customers’ side, they place a demand for corresponding spare
customers arise. This kind of situation generally arises when parts to the vendor. The supplier’s decision to satisfy this
customers of a supplier firm get some new contracts or there is a demand, fully or partially, is restricted by his own available
short term but a high fluctuation in their demands. When this resources since the long term strategic actions like production
sudden or contingent demand arises on customers’ side, they capacity expansion cannot be carried out in such a short
create a demand for corresponding spare parts to their suppliers. response time. In this kind of situation, a supplier firm finds
In this situation, the vendor’s decision to satisfy this demand fully itself in a fix to decide on the demand fulfilment of multiple
or partially is restricted by its own available resources, as customers for various products under consideration. The
processes like production capacity expansion cannot be carried reason behind this problem is that the firm needs to consider
out in such a short response time. The supplier has to decide on not just a single objective of obtaining maximum profit from
the extent of fulfilment of demands of various customers. The the demand fulfilment but also some other important aspects
decision making in this situation is difficult, as here the supplier like present business relations with its customers and future
has to consider not just the profit from the deal but also some business prospects with them. As these aspects are qualitative
other criteria of business sustainability like business relations
in nature and cannot be measured quantitatively thereby
and long sightedness in terms of future business prospects which
are qualitative in nature.
making it difficult for the decision maker to compare these
with the profit.
In this paper, we consider the problem of a supplier firm In this paper, we consider the problem of a supplier firm
which manufactures products that are used by its customer firms which manufactures products that are used by its customer
as spare parts to manufacture their specific products. We firms as spare parts to manufacture their particular products.
attempt to address the decision making on the extent of We address the decision making on the extent of satisfaction
satisfaction of contingent demand of each customer. While
of contingent demand of each customer. For this decision
deciding on the extent of demands of its customers, we try to
address the concern of the supplier firm to concentrate not just
making problem, on the extent of demands of its customers,
on the profit, but also on the future business and relationship we present to quantitatively measure the factors of future
with its client firms. For this, we try to quantize these business and present relations with customers. A multi-
linguistically expressible aspects, so as to use them in our objective optimization is proposed which is converted into a
optimization problem. single objective optimization problem using the classical
weighted sum approach and the weights (or priority) of
Keywords: Multiobjective optimization; contingent demand; various objective functions is obtained using experts
linguistic; membership function; AHP(Analytical Hierarchy knowledge and analytical hierarchy process. A survey is
Process) conducted with managers of a manufacturing firm with similar
profile and a similar situation generally faced by them. The
I. INTRODUCTION problem is solved for a sample data provided by the firm.
The planning of production and other operations in a firm
is done based on demand forecasts. But production decisions II. MOTIVATION AND RESEARCH GAP
get affected by actual realization of the demand. These Plethora of research has been done on the strategy building
decisions are taken keeping into consideration production and development of decision support systems (DSS) of firms
capacities. Sometimes sudden and unforeseen demand can for the supplier selection and order allocation [10]. Some of
arise from customers. This kind of situation is fairly common recent developments in the field are [3], [11]. Looking through
for the supplier firm producing those products which are the other direction of the supply chain, there is again ample
frequently purchased by other firms for further use in their development of DSS for production firms demand fulfillment
own production processes. The unforeseen demands at decisions. Some of well known developments are [1], [4], [7]
customer end can be either due to some new contracts and [8]. The literature suggests strategy building depending on
the known or estimated demand. But, so far no research has

978-1-4673-9939-5/16/$31.00 ©2016 IEEE

2271
addressed the issue of unforeseen demand from customers, proportional to the extent business relationship. So, we try to
especially for a supplier having its customers as business quantize the other two factors.
firms.
For quantizing the satisfaction or ease of doing business
The decision makers of any business organization, while with the customer using the above mentioned three types of
taking decisions on any deal or processes of supply chain, transactions, we propose to use a membership function defined
consider many strategy oriented factors directing towards long as follows.
term gains and sustainable business development. Experts
unsuccessful transactions
from industry opinion that major factors kept into = (1)
consideration during business to business (B2B) deals are total transactions
generally business relations with the organization, future
satisfactory transactions
business with the organization and future of the product = (2)
segment under negotiation with the organization. These total transactions
factors are qualitative in nature, in fact some aspects involved excellent transactions
in the assessment can be judged in a linguistic manner. So, no = (3)
work has tried to quantify these and therefore these are not total transactions
incorporated into quantitative DSS. Then using the weights with these functions (all three
weights being non-negative with their sum equals 1), which in
III. QUANTIFICATION OF QUALITATIVE FACTORS practice can be subjectively defined by the decision maker, the
Various components kept into consideration by strategic satisfaction factor may be calculated as following.
decision makers were identified through experts’ opinion from = +
industry for the above mentioned qualitative factors. We
develop the corresponding quantitative measures in context of + . (4)
our problem. Now for making the payment lead-time or grace period to
A. Developing Quantitative Measure for Business Relation reflect in the measure of business relations, one needs to
consider the subjectivity in the consideration of these time
For developing quantitative measure for the aspect of spans as small, medium or high. For this we propose to use the
business relation with a customer firm, following factors are three point membership function approach. As these lead-
identified to differentiate a customer firm from the other. times are defined in terms of number of days so demonstrate
(1) Fraction of the total demand of the customer which it these membership functions to be defined linearly as follows.
allocates to the decision maker supplier firm. This
1 0 ≤ < 10
gives a measurable value of the proportion of the
business for which the supplier is a partner of the = (30 − )/20 10 ≤ < 30, (5)
customer. 0 > 30
0 0 ≤ < 10
(2) The profit percentage (proportion) earned from the ⎧(
customer firm out of all the customers taken together. − 10)/20 10 ≤ < 30
= , (6)
(3) Satisfaction or ease of doing business with the ⎨(60 − )/30 30 ≤ < 60
customer. This can be measures in terms of a function ⎩ 0 > 60
of number transactions for payments which are 0 ≤ 30
categorized as unsuccessful, satisfactory and excellent. = ( − 30)/30 30 ≤ < 60 . (7)
Unsuccessful transactions are termed as those which 1 > 60
have failed due to some error or miscommunication.
Satisfactory are those which are successful but
delayed from promised due dates. The excellent one
are termed as those transactions which have been
performed and that too within the promised time
frame.
(4) Payment lead-time or grace period. This is measured
as the time span between the delivery of the placed
order and the receipt of its payment by the customer. Fig. 1. Membership functions of lead-time
Sometimes this is considered as the time span
We then propose to use the weights of these membership
promised by the buyer for the payment. The lead-time
functions to describe the preference or acceptance of each
for the payment is generally negotiated by the
category of lead-times as , and . Where,
customers in advance. The lesser the payment lead-
time, the better is the customer considered to do , , ∈ [0, 1] such that +
business with. + = 1.

Once quantized, let us denote these factors as ( = = + + . (8)

1, 2, 3, 4), respectively. Among the above four factors, first To prefer shorter lead-times one should consider
two are quantitative in such a manner that they are directly wsmall > wmedium > whigh.

2272
 After quantizing all the four factors as described above, we For aggregating these three elements to represent the
now propose to aggregate these as a single factor as follows. reputation factor as a number on the scale of [0, 1], we first
quantize each rating on the scale of [0, 1] by dividing them by
= , (9) 3 and then take their arithmetic mean [2]. For example, if a
buyer is rated as 1 for financial stability, 1 for market status
where, is the weight assigned to each quantized factor and 0 for recent news/ results, then their scores are obtained as
0.33, 0.33 and 0. Then by obtaining the arithmetic mean of
for their preference or importance in the decision making
these scores we obtain the aggregated score for reputation of
process.
the buyer as 0.22.
B. Developing Quantitative Measure for Future Business
The third factor considered as customer satisfaction is
Prospects taken from the scorecards ratings given by each buyer to the
This factor is considered as an initiative with the vision of supplier. These ratings are given in terms of percentages so we
promoting the business with those customers and for those can convert these percentage scores on the scale of [0, 1].
products which are envisaged as key role players in future
business. So, here we should quantize this factor The fourth factor of friendly managerial policy of buyer is
corresponding to each product and each buyer. In our problem, again a linguistic term. A buyer can be judged for his policy
as the buyer is a firm who manufactures some products using by the market analyst and given a grade on a 5 point scale as
the products from the supplier as spare parts, so accordingly shown below. The friendlier the managerial policy of the
we identify the influencing aspects as following. buyer, the better the future business collaboration can be
(1) Number of years of business collaboration expected with the buyer. This grading can be quantized on the
(since when buyer purchasing that product from scale of [0, 1] using Harrera’s approach [5].
supplier) Feedback Grading Score
(2) Reputation of buyer as an enterprise in its market Extremely poor 1 0.2
(a) financial stability Poor 2 0.4
(b) market value Satisfactory 3 0.6
(c) recent news/ results Good 4 0.8
(3) Customer satisfaction Extremely Good 5 1
(4) Friendly managerial policy
(5) Market performance of product of the buyer which The fifth factor considered as the market performance of
uses ith product as spare part/ raw material the product of the buyer, which uses ith product as spare part/
raw material, is again such a factor that can only be judged
Once quantized, for a product k (k = 1, 2…, K), let us and graded on a 5 point scale. So, this also can be quantized
denote these factors as ( = 1, … , 5), respectively. Out of on similar line as described above.
the above listed factors only the first and last are product
specific in addition to being buyer specific but the other three After quantizing all these five factors, as described above,
are buyer specific only. we propose to aggregate these into a single factor as follows.
For quantizing and scaling the years of business = , (11)
collaboration, we propose to define three point membership
functions for partnership in its initial, intermediate and high
stages for purchase of each product under consideration. These where, is the weight assigned to each quantized factor
are illustrated in equations (20) – (22). Further, the weights of for their preference or importance in the decision making
these membership functions to describe preference of the process.
durations of collaboration be defined as , Once above described aspects of future business and
and . Where, , , ∈ [0, 1] such present business relations are quantized, these can be
that + + = 1. The corresponding incorporated as parameters in our decision making problem.
component can then be expressed as for the product k (k = 1,
2…, K). IV. PROBLEM DESCRIPTION
( ) = + The proposed model considers the problem of a supplier,
+ . (10) which is a firm manufacturing ‘K’ types of products. This
supplier firm is having a set of ‘I’ customers who purchase
The second factor of reputation of a business firm can only these products for using these in their own production. The
be described in linguistic terms and can be detailed by the supplier firm receives a sudden additional demand for its
market analyst of the supplier company as financial stability, products from its customers. The supplier has to decide on the
market value and recent news/ results. All of these three proportion of the respective demands to be satisfied. In case,
elements can only be graded linguistically by a market expert. the supplier is capable of satisfying these additional demands
We incorporate the grading done by expert for each of these through its present resources then the problem is trivial. So, let
three as ratings of the buyer on a 4 point scale {0, 1, 2, 3} as us assume the case when the supplier has limited resources
follows. (which cannot be expanded all of a sudden) to meet all such
0 (poor) 1 (satisfactory) 2 (good) 3 (excellent) demands. The supplier wants to decide on the demand

2273
fulfilment proportions based on criteria of maximum possible The muti-objective mathematical model representing our
profit to be obtained, present business relations with each problem is described as follows.
customer firm, and future business prospects. The scenario is
depicted in Figure 1. ( ) = ( − )∗ (13)

= ∗ (14)

= ∗ (15)

subject to constraints

∗ ≤ (16)

∗ ≤ (17)
Fig. 2. Depiction of problem structure

We attempt to measure the present business relations and ∗ ≤ (18)

future business aspects quantitatively from the linguistic
feedbacks on various relevant components of experts of the 0≤ ≤  , . (19)
firm. This quantitative measurement is explained in the case
studied in section V. For understanding the decision model The objective (13) is aimed to maximize the total profit to
we, for the time being, assume these to be measured and be earned by satisfying the demands of various products from
known parameters. customers fully or partially. The objective (14) aims to
maximize the business relationship of the vendor with all the
V. PROBLEM FORMULATION customers. The objective (15) aims to give more weight to
fulfilling the demand of those products whose demand may
A. Notations for the Proposed Model grow in future. Constraint (16) gives a bound on the total of
actually fulfilled demand because of production capacity
TABLE I. NOTATIONS constraints. Constraint (17) gives a bound on the same because
Indices of available storage space. Constraint (18) gives a bound on
I number of customers the same because of total available budget. Constraint (19)
K number of different types of products manufactured by supplier limits the number of units of any product to be supplied to a
i index for customers (i = 1, 2,…, I) customer to not more than those actually demanded and not
k index for products (k = 1, 2,…, K) less than zero.
Parameters C. Solution Algorithm
Dik Demand of kth product from ith customer/buyer
Step 1. The objectives are first normalized by solving three
C Total Budget/Money in hand
optimization problems each with a single objective
Cik Cost of producing and delivering kth product to ith buyer
to obtain the “Positive Ideal Solution” (PIS) and the
Pik Per unit price of kth product paid by ith buyer
“Negative Ideal Solution” (NIS) following the
eik Production resource consumption approach proposed by Lai and Hwang [6].
Sk Space occupied by kth product
Step 2. The weights for the objectives are calculated using
S Total storage available space
analytical hierarchical process (AHP), a popular
M Total production capacity multi-criteria decision making method introduced
Li Business Relationship factor with ith buyer by Saaty [9] in 80's through numerous research
Eik Factor corresponding to future demand of kth product by ith buyer articles and books.
Variables Step 3. Weighted sum approach is used to solve the multi-
objective decision making problem (P) by
Xik Number of units of kth product to be supplied to ith buyer converting it into single objective optimization
problem.
B. Model
Our assumption, that the vendor under its present capacity
cannot satisfy the contingent demands of all its customers for VI. IMPERICAL ILLUSTRATION
various products, can be formulated as follows. We considered a case of a supplier who faces a contingent
demand of two of its products from four buyers. Case oriented
< (12) assumed data is obtained from an industry expert of a
manufacturing firm. The data is tabulated below.

2274
 TABLE II. Cost (per unit) of producing and supplying kth product The weighted avarage score thus obtained are as following.
to ith buyer (including logistic costs) (in Rs.): Cik
Lif Buyer 1 Buyer 2 Buyer 3 Buyer 4
Cik Buyer 1 Buyer 2 Buyer 3 Buyer 4
Product 1 2654 2521 2654 2654 Li (f=3) 0.545 0.565 0.56 0.62
Product 2 1600 1520 1600 1600
4. Payment lead-time or grace period
TABLE III. Pik: price paid by ith buyer for kth product (in Rs.)
Using the membership functions as defined in equations
Pik Buyer 1 Buyer 2 Buyer 3 Buyer 4 (5), (6) and (7) and weighted aggregate as defined in equation
Product 1 3800 3728 4000 4000 (8) for payment lead-times.
Product 2 2280 2232 2400 2400
wsmall wmedium whigh
TABLE IV. Contingent demand of kth product from ith buyer: Dik 0.5 0.35 0.15
Buyer Average μsmall μmedium μhigh
Dik Buyer 1 Buyer 2 Buyer 3 Buyer 4 Li (f=4)
lead-time
Product 1 50 110 35 20 Buyer 1 75 0 0 1 0.15
Product 2 65 110 35 20 Buyer 2 35 0 0.833 0.167 0.3166
Buyer 3 20 0.5 0.5 0.425 0.425
TABLE V. Production resource consumption eik Space occupied by Buyer 4 5 1 0 0 0.5
kth product (in cubic inch): Sk
Aggregating the above scores; the weights for the preferences
Product 1 Product 2 of these four factors given by the industry expert are listed
eik (machine-hours) 39.81 28.80 below in the following table showing the calculation of Li.
Sk (cubic inches) 735 506
Lif Li1 Li2 Li3 Li4
TABLE VI. Bounds of the model M, S and C
weights (wtlf) 0.2 0.5 0.2 0.1 Li
M (Available S (Total storage C (Total Budget 0.5 0.25 0.55 0.15
Buyer 1 0.35
machine hours) space available – in available)
cubic inches) Buyer 2 0.85 0.5 0.565 0.3166 0.56466
18070 259688 Rs. 9,03,500 Buyer 3 0.5 0.2 0.48 0.425 0.3385
Buyer 4 0.2 0.05 0.62 0.5 0.239
Now we demonstrate the quantification of the factors of
business relation Li and future business prospects Eik. B. Quantification of Future Business Prospects
A. Quantification of Business Relation 1. Years of Business Collaboration
1. The fraction of the regular demand of ith customer which The three point membership functions for categorizing the
he purchases from the supplier. years of business collaborations as initial, intermediate and
high have been defined as follows.
Lif Buyer 1 Buyer 2 Buyer 3 Buyer 4
Li (f=1) 0.5 0.85 0.5 0.2 1 0≤ <5
= (8 − )/3 5 ≤ <8 (20)
2. Profit percentage earned from ith customer. 0 >8
Lif Buyer 1 Buyer 2 Buyer 3 Buyer 4 =
%age 25% 50% 20% 5%
Li (f=2) 0.25 0.5 0.2 0.05 0 0≤ <5
⎧ (
− 5)/3 5≤ <8
3. Satisfaction or ease of doing business with the customer (21)
⎨(15 − )/7 8≤ < 15
Transactions ⎩ 0 > 15
type /Total Buyer 1 Buyer 2 Buyer 3 Buyer 4 0 ≤8
transactions = ( − 8)/7 8≤ < 15 (22)
Unsuccessful 5/100 2/80 5/50 0/10 1 > 15
Satisfactory 30/100 8/80 15/30 3/10
Excellent 65/100 70/80 30/50 7/10 The data of pertaining to years of collaboration of the four
Computing the three point membership values as following. buyers for each of the two products is as follows.
Years Buyer 1 Buyer 2 Buyer 3 Buyer 4
Buyer 1 Buyer 2 Buyer 3 Buyer 4
Product 1 20 10 3 1
μ i low 0.05 0.025 0.1 0 Product 2 10 4 1 1
μ i medium 0.3 0.1 0.5 0.3
μ i high 0.65 0.875 0.6 0.7
Using the prescribed weights as = 0.02, =
The weights for the three types of transacions prescribed by 0.02, = 0.02, the first factor is calculated as follows.
the expert are as follows.
Eik(f=1) Buyer 1 Buyer 2 Buyer 3 Buyer 4
weights Buyer 1 Buyer 2 Buyer 3 Buyer 4
wunsuccessful 0 0 0 0 Product 1 0.833 0.343 0.06 0.02
wsatisfactory 0.3 0.4 0.4 0.2 Product 2 0.343 0.08 0.02 0.02
wexcellent 0.7 0.6 0.6 0.8

2275
2. Reputation of buyer as an enterprise in its market C. Obtaining Weights of Multi-objectives problem using AHP
Eif Buyer 1 Buyer 2 Buyer 3 Buyer 4 Using 25 responses from the management personals of the
Financial Stability 3 2 1 1
firm, the weights corresponding to the three objectives of our
Market Value 3 2 1 1
model have been obtained using AHP as
Recent Results 3 2 1 0 w1 = 0.46, w2 = 0.31 and w3 = 0.23.
Then the single objective integer programming problem
The scores for the reputation factor are thus obtained as
was solved using the MATLAB 2014a.
follows.
Eif Buyer 1 Buyer 2 Buyer 3 Buyer 4
Ei (f=2) 1 0.67 0.33 0.2
VII. RESULTS AND ANALYSIS
3. Customer satisfaction factor from scorecard ratings are
taken as follows. The solution of the single objective problem obtained after
normalizing the multi-objective problem is obtained as
Eif Buyer 1 Buyer 2 Buyer 3 Buyer 4 follows.
%age 90% 95% 80% 50% TABLE VII. Demand Fulfilment Decision
Ei (f=3) 0.9 0.95 0.8 0.5
Xik Buyer 1 Buyer 2 Buyer 3 Buyer 4
4. Friendly Managerial Policy Product 1 50 110 35 7*
Linguistically characterized managerial policy of the Product 2 65 110 35 20
each of the buyers is quantized as following.
The optimal decision on fulfillment of demand of the four
Eif Buyer 1 Buyer 2 Buyer 3 Buyer 4 buyers for the two products is found to satisfy all the demands
Points 2 5 4 3 completely except that of Buyer 4 for the Product 1.
Ei (f=4) 0.4 1.0 0.8 0.6
Let us try to analyse and verify these results with the actual
5. Market Performance of product of kth buyer which uses ith situation and managerial decision making practices.
product as spare part/ raw material.
In our case, let the Buyer 1 is a buyer having good
Similarly, the market performances are quantized through business relationship with the supplier and is having
qualitative grading as follows. significantly long collaboration for 25 years. Buyer 2 is
Market Buyer 1 Buyer 2 Buyer 3 Buyer 4 signified with its very high demand and over that this buyer
Performance purchases an 85% of his demand from this supplier. So, this is
Product 1 5 4 3 2 also one of the buyers having good business relations and
Product 2 4 3 3 4 future business prospects with the supplier. Among the rest of
two buyers, Buyer 3 is having comparatively higher demand,
EiK (f=5) Buyer 1 Buyer 2 Buyer 3 Buyer 4 purchasing more demand proportions and contributes to more
Product 1 1 0.8 0.6 0.4 profit than the Buyer 4.
Product 2 0.8 0.6 0.6 0.8
Comparing these features, the demands of first three
Aggregating the scores; the weights for the preferences of buyers have been fulfilled which is within the suppliers
these four factors given by the industry expert are listed capacity of its available resources. Lastly, falling short of
below, followed by the tabulation of calculation of scores for resources, the supplier has decided on the demand of Product
the factor Eik. 2 fully but that of Product 1 partially. This is due to the fact
that the market performance of the product of this buyer using
Business Reputation Customer Managerial Market
Weights Partnership in Market Satisfaction Policy Performance
Product 1 is good but that of the Product 2 is poor. So, looking
(wtef) 0.35 0.2 0.15 0.15 0.15 through his own future business aspect this is justified.
VIII. CONCLUSIONS
Eik1 Eik2 Eik3 Eik4 Eik5 In this paper we propose a multi-objective optimization
wtef : 0.35 0.2 0.15 0.15 0.15 Eik model to decide on satisfying contingent demands of various
Product 1 0.5 1 0.9 0.4 1 0.72 buyers of various products for a supplier with a vision on
Buyer 1 present business relations and future business aspect along
Product 2 0.5 1 0.9 0.4 0.8 0.69
with the profit making objectives. For this we have
Product 1 0.85 0.67 0.95 1 0.8 0.844
Buyer 2 demonstrated the quantification of linguistic terms featuring
Product 2 0.85 0.67 0.95 1 0.6 0.814 many factors which are intended to be considered in the
Buyer 3
Product 1 0.5 0.33 0.8 0.8 0.6 0.571 decision making process. Using weighted sum approach, the
Product 2 0.5 0.33 0.8 0.8 0.6 0.571 problem is converted into a single objective optimization
Product 1 0.2 0.2 0.5 0.6 0.4 0.335 problem which can then be solved using the standard
Buyer 4 optimization techniques.
Product 2 0.2 0.2 0.5 0.6 0.8 0.395

2276
 [5] F. Herrera, and E. Herrera-Viedma, “Linguistic decision analysis: steps
Acknowledgment for solving decision problems under linguistic information”, Fuzzy Sets
The first author is thankful to the University Grant and Systems, vol. 115, 2000, p. 67 – 82.
Commission, India, for research grant. [6] Y. J. Lai, C. L. Hwang, Fuzzy Multiple Objective Decision Making:
Methods and Applications, Lecture Notes in Economics and
Mathematical Systems, Springer, 1994.
References [7] J. Li, “Customer demand satisfaction in production systems: a due-time
performance approach”, Robotics and Automation, IEEE Transactions,
vol 17 (4), 2002, p. 472 – 482.
[1] Y. Amer, L. Luong, S-H. Lee, “Case study: Optimizing order fulfillment
in a global retail supply chain, Int. J. Production Economics, vol. 127, [8] A. Nagurney, “Optimal supply chain network design and redesign at
2010, p. 278–291. minimal total cost and with demand satisfaction”, International Journal
of Production Economics, vol 128 (1), November 2010, p. 200–208.
[2] C.T. Chen, W.S. Tai, “Measuring the intellectual capital performance
based on 2-tuple fuzzy linguistic information”, in: Proceedings of the [9] T. L. Saaty, The Analytic Hierarchy Process, New York: McGraw Hill.
10th Annual Meeting of Asia Pacific Region of Decision Sciences International, Revised editions, Paperback (1996, 2000), Pittsburgh:
Institute, APDSI, Taiwan, 2005. RWS Publications, 1980.
[3] L. X. Cui, K.L. Mak, and S.T. Newman, “Optimal supplier selection and [10] M. Setak, S. Sharifi and A. Alimohammadian, “Supplier Selection and
order allocation for multi-product manufacturing featuring customer Order Allocation Models in Supply Chain Management: A Review”,
flexibility” International Journal of Computer Integrated Manufacturing, World Applied Sciences Journal, vol. 18 (1), 2012, p. 55-72.
vol. 28 (7), 2015 p. 729 – 744. [11] S.A. Torabi, M. Baghersad, S.A. Mansouri, “Resilient supplier selection
[4] A. Gupta, C. D. Maranasa, and C. M. McDonald , “Mid-term supply and order allocation under operational and disruption risks,
chain planning under demand uncertainty: customer demand satisfaction Transportation Research Part E: Logistics and Transportation Review,
and inventory management”, Computers & Chemical Engineering, vol. vol. 79, July 2015, p. 22–48.
24 (12), December 2000, p. 2613-2621.

2277