Documente Academic
Documente Profesional
Documente Cultură
a r t i c l e
i n f o
Keywords:
Four-echelon supply chain architecture
Evolutionary approach
Non-dominated sorting algorithm
MOHPSO
Swarm intelligence
a b s t r a c t
This paper aims at multi-objective optimization of single-product for four-echelon supply chain architecture consisting of suppliers, production plants, distribution centers (DCs) and customer zones (CZs). The
key design decisions considered are: the number and location of plants in the system, the ow of raw
materials from suppliers to plants, the quantity of products to be shipped from plants to DCs, from
DCs to CZs so as to minimize the combined facility location and shipment costs subject to a requirement
that maximum customer demands be met. To optimize these two objectives simultaneously, four-echelon network model is mathematically represented considering the associated constraints, capacity, production and shipment costs and solved using swarm intelligence based Multi-objective Hybrid Particle
Swarm Optimization (MOHPSO) algorithm. This evolutionary based algorithm incorporates non-dominated sorting algorithm into particle swarm optimization so as to allow this heuristic to optimize two
objective functions simultaneously. This can be used as decision support system for location of facilities,
allocation of demand points and monitoring of material ow for four-echelon supply chain network.
2012 Elsevier Ltd. All rights reserved.
1. Introduction
Supply chain (SC) is an integrated system of facilities and activities that synchronizes inter-related business functions of material
procurement, material transformation to intermediates and nal
products and distribution of these products to customers. Supply
chain management is a set of approaches utilized to efciently
integrate suppliers, manufacturers, warehouses, and stores, so that
merchandise is produced and distributed at the right quantities, to
the right locations, and at the right time, in order to minimize system-wide costs while satisfying service level requirements
(Simchi-Levi, Kaminsky, & Simchi-Levi, 2000). Above denition reveals that there are many independent entities in a supply chain
each of which tries to maximize its own inherent objective functions in business transactions. This is a complicated problem as
too many factors are involved and needs more than one objective
to be satised simultaneously. Such a problem is called multiobjective optimization problem and has many Pareto solutions.
The nal decision is made taking the total balance over all criteria
into account. This balancing over criteria is called trade-off.
Since today, the success measures for the companies are
thought as lower costs, shorter production time, shorter lead time,
Corresponding author.
E-mail address: chen@jepson.gonzaga.edu (J.C.H. Chen).
0957-4174/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.eswa.2012.07.065
less stock, larger product range, more reliable delivery time, better
customer services, higher quality, and providing the efcient coordination between demand, supply and production, the trade-off
between cost investment and service levels may change over time.
Hence the supply chain performance needs to be evaluated continuously and supply chain managers should make timely and right
decisions (Shen, 2007).
The key issues in supply chain management can broadly be divided into three main categories: (i) supply chain design (ii) supply
chain planning and (iii) supply chain control. In the supply chain
design phase, strategic decisions, such as facility location decisions
and technology selection decisions play major roles. It is very
important to design an efcient supply chain to facilitate the
movements of goods. These strategic decisions lead to costly, time
consuming investment as the facilities located today, are expected
to remain in operation for long time. Environmental changes during the facilitys lifetime can drastically alter the appeal of a particular site, turning todays optimal location into tomorrows
investment blunder. Determining the best locations for new facilities is thus an important strategic challenge (Owen & Daskin,
1998). Once the supply chain conguration is determined, the focus shifts to decisions at the tactical and operational levels, such
as inventory management decisions on raw materials, intermediate products, and end products and distribution decisions within
the supply chain (Chopra & Meindl, 2005).
552
In traditional supply chain management, the focus of the designs of supply chain network is usually on single objective, minimum cost or maximum prot. But the design, planning and
scheduling projects are usually involving trade-offs among different incompatible goals such as fair prot distribution among all
members, customer service levels, ll rates, safe inventory levels,
volume exibility, etc. (Chen & Lee, 2004). Hence real supply
chains are to be optimized simultaneously considering more than
one objective. Many of the problems that occur in supply chain
optimization are combinatorial in nature and picking a set of optimal solutions in the case of multi-objective formulations requires a
algorithm that can efciently search the entire objective space
using small amounts of computation time. Literature shows that
evolutionary algorithms perform well in this respect and give good
optimal results when applied to many combinatorial problems.
This work proposes the utility of non-dominated sorting particle
swarm optimization algorithm for simultaneous optimization of
two objectives, minimizing total supply chain cost and maximizing
ll rate for a four-echelon supply chain architecture so as to arrive
at an efcient supply chain design and optimal transportation/
shipment plan which can be used as decision support system.
plants and customers. The risks arising from the use of heuristics in
distribution planning are discussed early on by Geoffrion and Vanroy (1979). They present three examples in the area of distribution
planning demonstrating the failure of common sense methods to
come up with the best solution.
In literature, another set of problems considered is called xed
charge facility location problems which consider xed charge associated with locating at each potential facility site. There are two
types of problems capacitated and uncapacitated plant location
problems. Uncapacitated and capacitated plant location models
are extensively dealt in Mirchandani and Francis (1990) and ReVelle, Eiselt, and Daskin (2008) and capacitated plant location models in Sridharan (1995). Also basic facility location problems are
given a new orientation with integrated approach. This is due to
the realization of fact that location decisions taken without considering inventory and shipment costs can lead to sub-optimality.
Hence facility location models are developed considering location
associated costs as well as production, inventory and distribution
costs. Integrated decision making models in particular focus at
coordination of any two of the three important supply chain decisions. Based on the factors considered they are catagorized into (1)
locationrouting (LR) models; (2) inventoryrouting (IR) models;
and (3) locationinventory (Li) models. These problems are extensively reviewed by Shen (2007).
Dynamic facility location models next evolved make an attempt
to capture many of the characteristics of real-world location problems. Ballou (1968) rst use dynamic programming to determine
optimal location and relocation strategy for the planning period.
Wesolowsky (1973) examines another, unconstrained, version of
the single facility location problem over a nite planning horizon
with explicit facility relocation costs. Scott (1971) develop multiple
dynamic facility location-allocation problem. Wesolowsky and
Truscott (1976) present an integer programming model to extend
the analysis of multi-period node location-allocation problems,
allowing facilities to be relocated in response to predicted changes
in demand. Erlenkotter (1981) compares the performance of several heuristic solution approaches on a single problem formulation.
He examines a dynamic, xed charge, capacitated, cost minimization problem with discrete time intervals.
In traditional supply chain management, minimizing costs or
maximizing prot is the primitive objective in most of the supply
chain network design models (Cohen & Lee, 1989; Tsiakis, Shah,
& Pantelides, 2001). But for a supply chain, producing products at
minimum cost is not the only objective, satisfying customers is
also equally important. Later some researchers start incorporating
more than one competing objectives such as improving customer
service and reducing cost in their models. Different methodologies
found in literature for treating multiobjective optimization problems are the weighted-sum method, the e-constraint method and
the goal-programming method, fuzzy method, etc. (Azapagic &
Clift, 1999; Chen & Lee, 2004; Cheng-Liang, Wang, & Wen-Cheng,
2003; Zhou, Cheng, & Hua, 2000).
Sabri and Beamon (2000) develop a model for supply chain
management by combining strategic and operational design and
planning decisions and solve it using an iterative procedure. They
adopt multi-objective decision analysis and optimize simultaneously cost, ll rate and exibility. Nozick and Turnquist (2001)
present an optimization model which minimizes cost and maximizes service. They use a linear function to approximate the safety
stock inventory cost function, which is then embedded in a xedcharge facility location model. For a review of other multi-objective location models, publication by Shen, Coullard, and Daskin
(2003) can be referred. Chen and Lee (2004) propose a model
which simultaneously optimize conicting objectives such as each
participants prot, the average customer service level, and the
average safe inventory level. Guillna, Melea, Bagajewiczb,
Espuaa, and Puigjanera (2004) formulate a multiobjective stochastic Mixed Integer Linear Programming model for supply chain
design, which is solved by using the standard e-constraint method,
and branch and bound techniques. This formulation takes into account not only Supply Chain prot and customer satisfaction level,
but also uncertainty by means of the concept of nancial risk,
which is dened as the probability of not meeting a certain prot
aspiration level. Shen (2006) addresses prot-maximizing supply
chain design model where in a company can choose whether to
satisfy a customers demand.
Particle swarm optimization (PSO) is one of the evolutionary
computation techniques and is a population-based search algorithm which is initialized with a population of random solutions
called particles. Each particle in PSO is also associated with a velocity. Particles y through the search space with velocities which are
dynamically adjusted according to their historical behaviors. Due
to this the particles have a tendency to y towards the better
and better search area over the course of search process. This technique is originally proposed by Kennedy and Russell Eberhart as a
simulation of social behavior, and it is initially introduced as an
optimization method in 1995 (Kennedy & Eberhart, 1995). PSO
can be easily implemented and it is computationally in expensive,
since its memory and CPU speed requirements are low. The main
difculty in Multi-objective Particle Swarm Optimization (MOPSO)
is to pick suitable global best (gbest) and personal best (pbest) to
move the particles through search space. In general, a good MOPSO
method must obtain solutions with (a) a good convergence, and (b)
a diversity spread along the Pareto-optimal front (Coello, 1999).
Parsopoulos and Vrahatis (2002) are one among to study rst
the performance of the PSO in multi-objective optimization problems and nd Pareto front in weighted aggregation cases. Hu and
Eberhart (2002) present a MOPSO that uses dynamic neighborhood
strategy to obtain the best local guide for each particle in multiobjective optimization problems. Coello and Lechuga (2002) propose a MOPSO where the objective space is divided to hyper cubes.
Fieldsend and Singh (2002) propose a MOPSO which uses a dominated tree for the search of the best local solution. Mostaghim and
Teich (2003) propose a sigma method and turbulence operator to
improve the convergence and diversity for multi-objective optimization. Hu, Eberhart, and Shi (2003) adopt a secondary population
called extended memory to improve their previous dynamic neighborhood PSO approach presented in (2002). Mostaghim and Teich
(2004) propose a new method which divides population of the covering MOPSO into sub-swarms. The sub-swarms try to cover the
gaps between the non-dominated solutions found in the initial
run. In later research in the eld of MOPSO, different approaches
have been introduced to identify the local best solution based on
memorizing all the non-dominated solutions. These approaches
when implemented on standard test problems show that keeping
the particle archive improves signicantly the effectiveness of
the technique (Jrgen & Sanaz, 2006). Mahnam, Yadollahpour, Famil-Dardashti, and Hejazi (2009) develop an inventory model for
an assembly supply chain network where in the performance of
supply chain is assessed by two criteria total cost and ll rate using
a fuzzy expert system. To solve this bi-criteria model, hybridization
of multi-objective particle swarm optimization and simulation
optimization are considered. Results indicate the efciency of proposed approach in performance measurement. Guo, Li, Mileham,
and Owen (2009) apply effectively a variant of PSO to integrate
process planning and scheduling. Recently Guo and Hou (2010)
develop a simulation model to describe the operations of a
third-party logistics provider. They develop an MOPSO algorithm
combining with the simulation model to identify non-dominated
solutions that constitute the trade-off curve between ll rate and
total inventory. Marinakis and Marinaki (2010) propose a new
hybrid algorithmic nature inspired approach based on particle
553
554
conicting objectives; minimizing total supply chain cost consisting of production, transportation and distribution cost and maximizing ll rate using MOHPSO algorithm. The proposed strategy
should lead to a nal SC design which would represent the desired
compromise among the different objectives from the decision-makers perspective.
Objective functions
To minimize total cost of supply chain which includes raw
material making and transportation, plant location, production
and inventory costs in plant, distribution cost from plant to
DC, throughput cost in DC, distribution cost from DC to CZ, etc.
To maximize the ll rate.
Product availability reects a rms ability to ll a customer
order out of available inventory. A stock-out results if a customer
order arrives when product is not available. Product ll rate (fr)
and cycle service level (CSL) are the ways to measure product
availability. Product ll rate is the fraction of product demand that
is satised from product in inventory. CSL is the fraction of replenishment cycles that end with all the customer demand being met
(Chopra & Meindl, 2005). This work considers a constrained
bi-objective problem formulation for a pull based supply chain
and proposes the utility of using hybrid MOPSO algorithm for the
simultaneous optimization of both cost and product ll rate.
4. Problem description
This work considers four-echelon capacitated plant locationallocation network model as shown in Fig. 1. The notations used
in this network model are listed in Table 1 with their meaning.
The model denes decision variables as shown in Table 2.
The problem is then formulated as the following mixed integer
model:
Objective 1 : Min
Product produced needs three raw materials.
Capacities of suppliers for different raw materials, cost of making each raw material at each supplier, cost of transportation
per raw material per supplierplant pair are known in advance.
Potential location sites for plants and DCs are known.
Production, inventory costs per product and distribution costs
per plantDC pair are known and xed.
Inventory costs per product and cost of transportation per DC
CZ pair are known and xed.
Minimum ll rate is to be maintained.
The overall problem can thus be stated as follows:
Given
Number of suppliers and their capacities for each raw material.
Number of potential plants, DCs and their capacities.
Cost of making of each raw material at each supplier and transportation cost per supplierplant pair.
Production and inventory costs in each plant.
Distribution costs per product per plantDC pair.
Product throughput cost in each DC and transportation cost per
DCCZ pair.
Number of CZs and their demands.
Minimum ll rate (fraction of demand satised) to be
maintained.
To determine
Quantity of each raw material to be shipped from each supplier
to a plant.
Number and location of plants.
Flow of products from located plants to DCs.
Number and location of DCs.
Allocation of CZs to DCs.
p
n
t
n X
l X
X
X
X
fi yi
fe ye
cchi xhci
e1
i1
i1 h1 c1
n X
t
t X
m
X
X
cie xie
cej xej
i1 e1
Pt
e1
Objective 2 : Max
e1 j1
Pm
Pm
j1 xej
j1 Dj
Subject to
3
n
X
xhci 6 Sch for h 1; . . . :l; c 1; ::p
i1
t
X
xej 6 Dj for j 1; . . . ; m
e1
t
X
xie 6 K i yi for i 1; . . . :n
e1
m
X
xej 6 K e ye for e 1; . . . ; t
6
7
j1
l
t
X
X
xhci
xie P 0 for i 1; . . . ; n;
h1
e1
c 1; . . . ; p
n
m
X
X
xie
xej P 0 for e 1; . . . ; t
i1
0:80 6
j1
Pt
e1
8
9
Pm
Pm
j1 xej
j1 Dj
61
yi ; ye 2 f0; 1g for i 1 . . . . . . n; e 1; . . . ; t
10
11
The objective function (1) minimizes the total cost (xed + variable)
of setting up and operating the network and objective function (2)
maximizes ll rate.
The constraint in Eq. (4) species that the total quantity shipped
from a supplier cannot exceed the suppliers capacity. The
555
Fig. 1. Four-Echelon supply chain network of 3 suppliers, 5 plants, 6 dcs and 7 CZs.
Table 1
Notations and their meaning used in the mathematical formulation.
Notations
used
Meaning
t
n
l
p
Dj
Ki
Ke
Sch
fi
fe
cchi
cie
cej
CSch
STCchi
ci
ICi
Ici
PTCie
Table 2
Decision variables used in the mathematical formulation.
Notations used
Meaning
yi
ye
xhci
xie
xej
5. Methodology
constraint in Eq. (5) requires that the demand at each regional
market be satised to the maximum extent. The constraint in Eq.
(6) states that no plant can supply more than its capacity. Capacity
is 0 if the plant is closed and Ki if it is open. Similarly Eq. (7) states
556
Kennedy and Eberhart (1995). This algorithm models a set of potential problem solutions as a swarm of particles moving about
in a virtual search space. PSO is based on the behavior of communities that have both social and individual communication, similar
to birds searching for food. The bird would nd food through social
cooperation with other birds around it (within its neighborhood).
In PSO algorithm, each individual (particle) represents a solution in a n-dimensional space. Each particle also has knowledge
of its previous best experience and knows the global best experience (solution) found by the entire swarm. Each particle updates
its direction using the following equations:
to represent a positive random number drawn from a uniform distribution with a predened upper limit so that the two u limits sum
to 4.0. Thus the formula for binary decision is
where
P(xid(t) = 1) is the probability that individual i will choose 1.
(xid(t)) is the current state of individual i.
t is the current time step, and t 1 is the previous step.
vid(t 1) is a measure of the individuals current probability of
deciding 1.
pid is the best state found so far, for example, it is 1 if the individuals best success occurred when xid was 1 and 0 if it was 0. It
is referred as pbest.
pgd is the neighborhood best, again 1 if the best success attained
by any member of the neighborhood was when it was in the 1
state and 0 otherwise. It is referred as gbest.
The parameter vid(t), will determine the probability threshold. If
vid(t) is higher, the individual is more likely to choose 1, and lower
values favor the 0 choice. Such a threshold stays in the range
[0.0, 1.0]. The following sigmoid function is used to determine
the probability threshold.
sv id
1
1 expv id
557
Vendors
Components
V1
V2
V3
C1
C2
C3
36
40
42
62
65
70
50
55
60
Plants
XCapacity
Fixed Cost
P1
P2
P3
P4
P5
18
24
37
22
41
7650
3500
500
4100
2200
Warehouses
Capacity
WH1
WH2
WH3
WH4
WH5
WH6
15
12
14
13
12
16
Table 4
Cost of making and shipping cost matrix per unit for rst stage of SCN.
Vendors
Components
V1
C1
C2
C3
300
115
90
10
6
3
13
7
4
8
5
5
11
8
4
15
4
5
V2
C1
C2
C3
320
120
85
17
6
6
14
5
6
12
7
5
12
5
6
15
7
4
V3
C1
C2
C3
290
125
75
13
6
3
12
5
6
14
3
3
11
4
2
9
5
3
Table 5
Manufacturing cost, inventory cost, transportation cost matrix per unit for second stage of SCN.
Plants
P1
P2
P3
P4
P5
Warehouses
WH1
WH 2
WH 3
WH 4
WH 5
WH 6
7
12
8
10
8
12
10
10
12
10
15
11
14
13
11
17
13
15
14
15
18
15
18
13
11
20
17
21
18
12
3900
2010
1945
1855
1975
50
45
55
48
52
Table 6
Handling cost and transportation cost matrix per unit for third stage of SCN.
Warehouses
WH1
WH2
WH3
WH4
WH5
WH6
Monthly demand Dj
Customer zones
CZ1
CZ2
CZ3
CZ4
CZ5
CZ6
CZ7
8
5
9
3
7
5
3
3
8
3
9
6
6
5
9
7
8
2
3
7
4
6
6
6
2
9
8
6
7
3
7
5
4
3
4
3
2
5
4
9
2
5
4
8
4
8
4
9
3
55
50
60
54
55
45
558
Total xed
cost
Total
variable cost
Total SC
cost
Demand
satised
Actual
demand
1
2
3
4
5
6
7
30,050
27,150
29,750
15,600
26,950
20,900
17,950
112,256
114,235
110,879
104,304
90,868
82,390
82,530
142,306
141,385
140,629
119,904
117,818
103,290
100,480
30
29
28
27
26
25
24
30
30
30
30
30
30
30
Table 8
The optimal material ows from suppliers to plants corresponding to Pareto front
points.
Supplier plants
Fig. 2. Feasible solutions where in series 2 represents Pareto-front with six nondominated points showing trade-off between SC cost and ll rate and series 1
represents other inferior points.
method, goal programming method etc. But all these methods convert a multi-objective optimization problem into a single-objective
optimization problem. These methods are subjective and only one
Pareto-optimal solution can be expected in one simulation run.
Since then there has been increasing interest to use evolutionary
algorithms in solving multi-objective optimization because of their
ability to nd multiple optimal solutions in single run and complexity of location-allocation problems.
5.6. Selecting the global best and personal best
In MOHPSO, each particle has to change its position as guided
by two leaders pbest and gbest which must be selected from the
updated set of non-dominated solutions stored in the archive.
The main difculty in MOHPSO is to pick suitable global best
(gbest) and personal best (pbest) to move the particles through
search space to attain a good convergence and diversity along
the Pareto-optimal front.
In this study, method of selection of pbest and gbest used is inspired by NSGA-II. For selecting pbest a method called Prandom is
employed, according to which a single pbest is maintained. Pbest is
replaced if new value < pbest, otherwise, if new value is found to
be mutually non-dominating with Pbest, one of the two is
randomly selected to be the new pbest. ( Eveson, Fieldsend, &
Singh, 2002). To select gbest from archive, this work makes use
Component 1
Component 2
Component 3
V1
V2
V3
V1
V2
V3
V1
V2
V3
Scenario 1
P1
P2
P3
P4
P5
0
5
15
11
0
0
0
0
0
0
8
11
0
0
0
10
10
18
0
0
0
12
0
12
0
0
0
0
0
0
0
0
0
15
0
12
22
12
0
0
0
0
0
0
0
Scenario 2
P1
P2
P3
P4
P5
0
0
10
0
0
1
18
9
0
11
4
0
0
0
0
8
16
0
0
0
0
0
17
0
10
0
0
0
0
0
8
9
0
0
0
0
12
0
0
9
0
0
20
0
0
Scenario 3
P1
P2
P3
P4
P5
7
0
0
8
0
0
0
0
0
0
10
6
0
0
0
13
0
0
10
0
5
8
0
0
0
0
0
0
0
0
5
0
0
7
0
0
0
0
0
0
14
9
0
0
0
Scenario 4
P1
P2
P3
P4
P5
0
0
23
0
0
0
0
0
0
0
0
0
0
15
17
0
0
25
0
20
0
0
0
16
0
0
0
0
0
0
0
0
21
0
0
0
0
0
0
0
0
0
0
18
22
Scenario 5
P1
P2
P3
P4
P5
0
0
0
10
0
11
6
0
0
0
0
0
0
0
0
0
0
0
9
0
0
6
0
0
0
12
0
0
2
0
8
0
0
12
0
3
0
0
0
0
0
7
0
0
0
Scenario 6
P1
P2
P3
P4
P5
0
0
6
0
0
0
7
5
0
0
0
11
0
0
9
0
0
10
0
0
0
0
0
0
12
0
14
0
0
0
0
19
12
0
0
0
0
0
0
7
0
0
0
0
0
Scenario 7
P1
P2
P3
P4
P5
0
0
17
0
0
0
0
0
0
0
9
0
0
0
0
0
0
18
0
0
0
0
0
0
0
10
0
0
0
0
0
0
0
0
0
0
0
20
0
0
11
0
0
0
0
D1
D2
D3
D4
D5
D6
Optimal location
of plants
Scenario 1
P1
P2
P3
P4
P5
0
1
8
0
0
0
3
0
2
0
0
1
0
2
0
2
7
0
2
0
0
0
0
0
0
4
0
2
0
0
1
1
1
1
0
Scenario 2
P1
P2
P3
P4
P5
1
2
2
0
0
0
0
0
0
0
0
0
9
0
0
1
6
0
0
3
0
4
1
0
0
1
2
2
0
4
1
1
1
0
1
Scenario 3
P1
P2
P3
P4
P5
0
0
0
0
0
4
0
0
2
0
1
2
0
2
0
8
0
0
3
0
4
0
0
0
0
0
4
0
0
0
1
1
0
1
0
Scenario 4
P1
P2
P3
P4
P5
0
0
2
2
2
0
0
0
0
0
0
0
0
0
0
0
0
5
4
0
0
0
5
2
1
0
0
2
2
9
0
0
1
1
1
Scenario 5
P1
P2
P3
P4
P5
1
0
0
0
0
3
0
0
0
0
4
0
0
3
0
0
0
0
0
0
3
0
0
4
0
0
6
0
3
0
1
1
0
1
0
Scenario 6
P1
P2
P3
P4
P5
0
0
5
0
0
0
3
4
0
0
0
7
0
0
0
0
0
0
0
6
0
4
0
0
1
0
0
0
0
0
0
1
1
0
1
Scenario 7
P1
P2
P3
P4
P5
3
0
0
0
0
4
0
5
0
0
0
0
0
0
0
1
0
5
0
0
0
0
0
0
0
1
0
6
0
0
1
0
1
0
0
Niche count nci for ith solution is calculated using sharing function sh(dij) as follows:
8
< 1 d a ; if d 6 r
N
X
share
r
share
nci
shdij and shd
:
j1
0;
otherwise
559
6. Numerical example
In order to illustrate the capabilities of the proposed model and
algorithm, a hypothetical case study of four echelon supply chain
network connecting 3 suppliers, 5 plants, 6 DCs and 7 CZs as shown
in Fig. 1 is considered. As per the problem, the company has manufacturing plants located in ve different geographical locations
with total capacity of 142 which serve markets in seven different
places with total demand of 30 through six warehouses with total
capacity of 82. In the rst stage of SCN, each vendor supplies three
components which are required for producing one product. Supplier capacities, plant capacities, warehouse capacities and xed
costs per month at each plant are given in Table 3. Supplier cost
of making a component and supplier transportation cost per unit
shipped are given in Table 4. Variable inventory, production and
transportation costs per unit shipped and demand at different
CZs are given in Table 5. Handling cost and transportation cost matrix per unit for third stage of SCN is given in Table 6. This problem
is analyzed with respect to two conicting objectives, minimizing
total cost and maximizing ll rate, with the condition that minimum ll rate to be attained is 80% of actual demand (24). There
are totally 128 decision variables to be optimized and 48 constraints. This problem results in seven Pareto solutions. The nal
decision is made among them taking the total balance over all criteria into account. This is a problem of value judgment of decision
maker. Also these decisions will be revisited every year as demand
and costs change.
6.1. The proposed MOPSO implementation
The proposed MOHPSO based approach was implemented using
C language and the developed software program was executed on a
1.8-GHz Pentium 4 personal computer. The random seeds were
generated by a random number generator incorporated in the program. Initially, several simulation runs have been done with different values of the PSO key parameters such as the maximum
allowable velocity. Other parameters selected are: number of particles n = 20, decrement constant = 0.8, c1 = 1.05, c2 = 0.05. The time
560
Table 10
Optimum distribution of products from DCs to CZs corresponding to Pareto front points.
Supply DC
CZ1
CZ2
CZ3
CZ4
CZ5
CZ6
CZ7
Scenario 1
DC1
DC2
DC3
DC4
DC5
DC6
Total quantity supplied
Demand
0
0
1
1
0
1
3
3
0
0
2
2
0
1
5
5
2
1
0
1
0
0
4
4
3
1
0
2
0
0
6
6
0
1
0
1
0
2
4
4
1
0
0
2
0
2
5
5
2
1
0
0
0
0
3
3
1
1
1
1
0
1
30
30
Scenario 2
DC1
DC2
DC3
DC4
DC5
DC6
Total quantity supplied
Demand
1
0
1
1
0
0
3
3
1
0
2
0
1
1
5
5
1
0
1
0
2
0
4
4
1
0
2
2
1
0
6
6
0
0
0
1
0
3
4
4
1
0
0
0
0
4
5
5
0
0
0
2
0
0
2
3
1
0
1
1
1
1
29
30
Scenario 3
DC1
DC2
DC3
DC4
DC5
DC6
Total quantity supplied
Demand
0
1
0
1
1
0
3
3
0
0
0
2
1
1
4
5
0
0
0
3
0
0
3
4
0
3
3
0
0
0
6
6
0
1
0
2
0
1
4
4
0
1
0
3
1
0
5
5
0
0
2
0
0
1
3
3
0
1
1
1
1
1
28
30
Scenario 4
DC1
DC2
DC3
DC4
DC5
DC6
Total quantity supplied
Demand
0
0
0
0
2
0
2
3
0
0
0
4
0
1
5
5
0
0
0
0
0
3
3
4
2
0
0
0
3
1
6
6
2
0
0
0
0
2
4
4
2
0
0
0
2
0
4
5
0
0
0
0
0
3
3
3
1
0
0
1
1
1
27
30
Scenario 5
DC1
DC2
DC3
DC4
DC5
DC6
Total Quantity Supplied
Demand
0
0
1
0
1
0
2
3
0
0
1
0
2
1
4
5
0
0
1
0
3
0
4
4
0
3
0
0
0
3
6
6
0
0
2
0
0
1
3
4
0
0
1
0
0
4
5
5
0
0
1
0
1
0
2
3
0
1
1
0
1
1
26
30
Scenario 6
DC1
DC2
DC3
DC4
DC5
DC6
Total Quantity Supplied
Demand
0
0
1
0
1
0
2
3
0
2
0
0
2
0
4
5
2
1
0
0
0
0
3
4
1
2
2
1
0
0
6
6
0
0
0
3
0
0
3
4
0
1
3
0
0
0
4
5
2
0
1
0
0
0
3
3
1
1
1
1
1
0
25
30
Scenario 7
DC1
DC2
DC3
DC4
DC5
DC6
Total quantity supplied
Demand
0
1
0
0
0
1
2
3
0
3
0
1
0
0
4
5
0
0
0
1
0
2
3
4
0
3
0
2
0
0
5
6
2
0
0
0
0
2
4
4
0
2
0
0
0
2
4
5
0
0
0
2
0
0
2
3
1
1
0
1
0
1
24
30
Fig. 3 for total demand of 27 for both best and worst cases. The
solution space has seven non-dominated optimum decision points,
whose corresponding objective values are given in Table 7 and corresponding decision variables representing optimum material
ows from suppliers to plants for seven different scenarios are
given in Table 8, optimum location and distribution of goods from
5 plants to 6 DCs are tabulated in Table 9 and optimum location
and ow of goods from 6 DCs to 7 CZs are tabulated in Table 10.
These points clearly demonstrate tradeoffs in objective functions
total SC cost to be minimized and ll rate to be maximized from
which an appropriate solution can be compromisingly chosen as
per the requirements of the corresponding industrial environment.
8. Conclusion
In this paper, an analytical model is formulated for the location
and allocation of facilities of four-echelon supply chain network for
the optimal facility location and capacity allocation decisions.
Fixed location and variable material cost, production, inventory
and transportation costs are considered while making strategic
decisions. Two objective functions of minimizing total SC cost
and maximizing ll rate are considered. A heuristic based hybrid
MOPSO is used as optimizer. This algorithm is mainly aimed at
characterizing the Pareto Optimal front by computing well-distributed non-dominated solutions. These solutions represent trade-off
solutions which facilitate decision makers to develop management
policies under a changing environment. Further this can be used as
decision support system for strategic supply chain design and
monitoring of material ow. Above explained model and algorithm
further can consider safety stock costs, risk related and reliability
costs. Also the model can optimize more than two conicting
objectives simultaneously.
References
Azapagic, A., & Clift, R. (1999). The application of life cycle assessment to process
optimisation. Computer and Chemical Engineering, 10, 15091526.
Ballou, R. H. (1968). Dynamic warehouse location analysis. Journal of Marketing
Research, 5, 271276.
Che, Z. H. (2012). A particle swarm optimization algorithm for solving unbalanced
supply chain planning problems. Applied Soft Computing, 12, 12791287.
Cheng-Liang, Chen, Wang, B. W., & Wen-Cheng, Lee. (2003). Multi-objective
optimization for a multi-enterprise supply chain network. Industrial and
Engineering Chemistry Research, 42, 18791889.
Chengming, Qi. (2011). Application of improved discrete particle swarm
optimisation logistics distribution routing problem original research. Procedia
Engineering, 15, 3673-3677.
Chen, Cheng-Liang, & Lee, Wen-Cheng (2004). Multi-objective optimization of
multi-echelon supply chain networks, with uncertain product demands and
prices. Computers and Chemical Engineering, 28, 11311144.
Chopra, Sunil, & Meindl, Peter (2005). Supply chain management, strategy, planning
and operation. India: Pearson Prentice Hall.
Church, R. L., & ReVelle, C. S. (1976). Theoretical and computational links between
the p-median location set-covering and the maximal covering location problem.
Geographical Analysis, 8, 406415.
Coello, C. A. C. (1999). A comprehensive survey of evolutionary-based
multiobjective optimization techniques. Knowledge and Information Systems,
1(3), 269308.
Coello, C. A. C., & Lechuga, M. S. (2002). MOPSO: A proposal for multiple objective
particle swarm optimization. Congress on Evolutionary Computation 2002 (Vol. 2,
pp. 10511056). Piscataway, New Jersey: IEEE Service Center.
Cohen, M. A., & Lee, H. L. (1989). Resource deployment analysis of global
manufacturing and distribution networks. Journal of Manufacturing and
Operations Management, 81104.
Daskin, M. S. (1995). Network and Discrete Location: Models Algorithms and
Applications. New York: Wiley.
Deb, K. (2001). Multi-objective Optimization Using Evolutionary Algorithms.
Chichester: John Wiley and Sons.
Erlenkotter, D. (1981). A comparative study of approaches to dynamic location
problems. European Journal of Operational Research, 6, 133143.
Eveson, R.M., Fieldsend, J.E., Singh, S., 2002. Full elite sets for multi-objective
optimisation. In Parmee, I.C. (Ed.), Adaptive computing in design and manufacture
(pp. 343354).
561
Fieldsend, J.E., & Singh, S. (2002). A multi-objective algorithm based upon particle
swarm optimization, an efcient data structure and turbulence. In Proceedings
of the 2002 UK workshop on computational intelligence, Birmingham, UK (pp. 37
44).
Geoffrion, A. M., & Graves, G. W. (1974). Multi commodity distribution system
design by benders decomposition. Management Science, 20, 822844.
Geoffrion, A. M., & Vanroy, T. J. (1979). CautionCommon-sense planning methods
can be hazardous to your corporate health. Sloan Management Review, 20,
3142.
Guillna, G., Melea, F. D., Bagajewiczb, M. J., Espuaa, A., & Puigjanera, L. (2004).
Multiobjective supply chain design under uncertainty. Chemical Engineering
Science, 15351553.
Guo, Shin-Ming, & Hou, Chih Wei (2010). Multi- objective optimization for VMI
operations in the LCD supply chain. Computers and Industrial Engineering, 16,
16.
Guo, Y. W., Li, W. D., Mileham, A. R., & Owen, G. W. (2009). Applications of particle
swarm optimisation in integrated process planning and scheduling. Robotics
and Computer-Integrated Manufacturing, 25, 280288.
Hakimi, S. L. (1964). Optimum locations of switching centers and the absolute
centers and medians of a graph. Operations Research, 12, 450459.
Hu, X., & Eberhart, R. C. (2002). Multiobjective Optimization Using Dynamic
Neighborhood Particle Swarm Optimization. Congress on Evolutionary
Computation, 2, 16771681.
Hu, X., Eberhart, R.C., & Shi, Y. (2003). Particle swarm with extended memory for
multiobjective optimization. In Proceedings of the 2003 IEEE swarm intelligence
symposium, Indianapolis, IN, USA (pp. 193197).
Jrgen, B., & Sanaz, M. (2006). About selecting the personal best in multi-objective
particle swarm optimization. In Proceedings of the 9th international conference on
parallel problem solving from nature PPSN IX, Reykjavik, Iceland (pp. 523532).
Kadadevaramath, R. S., Chen, J. C. H., Latha Shankar, B., & Rameshkumar, K. (2012).
Application of particle swarm intelligence algorithms in supply chain network
architecture optimization. Expert Systems with Applications, 39, 1016010176.
Kennedy, J., & Eberhart, R.C., 1995. Particle swarm optimization. In Proceedings of the
IEEE international conference on neural networks, Perth, Australia (pp. 1942
1948).
Kennedy, J., & Eberhart, R. C. (2001). Swarm intelligence. Morgan Kaufmann
Publishers.
Lien, Li-Chuan, & Cheng, Min-Yuan (2012). A hybrid swarm intelligence based
particle-bee algorithm for construction site layout optimization. Expert Systems
with Applications, 39(10), 96429650.
Mahnam, Mehdi, Yadollahpour, Mohammad Reza, Famil-Dardashti, Vahid, & Hejazi,
Seyed Reza (2009). Supply chain modeling in uncertain environment with biobjective approach. Computers & Industrial Engineering, 56, 15351544.
Marinakis, Yannis, & Marinaki, Magdalene (2010). A hybrid multi-swarm particle
swarm optimization algorithm for the probabilistic traveling salesman
problem. Computers & Operations Research, 37(3), 432442.
Mirchandani, P. B., & Francis, R. L. (Eds.). (1990). Discrete location theory. Wiley.
Mostaghim, S., & Teich, J., 2003. Strategies for nding good local guides in
multiobjective particle swarm optimization (MOPSO). In Proceedings of the 2003
IEEE swarm intelligence symposium, Indianapolis, IN, USA (pp. 2633).
Mostaghim, S., & Teich, J. (2004). Covering pareto-optimal fronts by subswarms in
multiobjective particle swarm optimization. In Proceedings of the IEEE congress
on evolutionary computation 2004, Portland, Oregon, USA (pp. 14041411).
Mousa, A. A., El-Shorbagy, M. A., & Abd-El-Wahed, W. F. (2012). Local search based
hybrid particle swarm optimization algorithm for multi-objective optimization.
Swarm and Evolutionary Computation, 3, 114.
Mukhopadhyay, Sumona, & Banerjee, Santo (2012). Global optimization of an
optical chaotic system by Chaotic Multi Swarm Particle Swarm Optimization.
Expert Systems with Applications, 39(1), 917924.
Navalertporn, T., & Afzulpurkar, N. V. (2011). Optimization of tile manufacturing
process using particle swarm optimization. Swarm and Evolutionary
Computation, 1(2), 97109.
Nozick, L., & Turnquist, M. (2001). Inventory, transportation, service quality and the
location of distribution centers. European Journal of Operations Research, 129,
362371.
Owen, S. H., & Daskin, M. S. (1998). Strategic facility location: A review. European
Journal of Operational Research, 111, 423447.
Parsopoulos, K.E., & Vrahatis, M.N., 2002. Particle swarm optimization method in
multiobjective problems. In Proceedings of the ACM 2002 symposium on applied
computing (pp. 603607).
Prasanna Venkatesan, S., & Kumanan, S. (2012). A multi-objective discrete particle
swarm optimisation algorithm for supply chain network design. International
Journal of Logistics Systems and Management, 11(3), 375406.
ReVelle, C. (1986). The maximum capture or sphere of inuence location problem:
Hotelling revisited on a network. Journal of Regional Science, 26(2), 343358.
ReVelle, C. S., Eiselt, H. A., & Daskin, M. S. (2008). A bibliography for some
fundamental problem categories in discrete location science. European Journal
of Operational Research, 184, 817848.
Sabri, E. H., & Beamon, B. M. (2000). A multi-objective approach to simultaneous
strategic and operational planning in supply chain design. Omega, 28(5),
581598.
Scott, A. J. (1971). Dynamic location-allocation systems: Some basic planning
strategies. Environment and Planning, 7382.
Shen, Z. J. (2006). A prot maximizing supply chain network design model with
demand choice exibility. Operations Research Letters, 34, 673682.
562
Shen, Z.-J. M. (2007). Integrated supply chain design models: A survey and future
research directions. Journal of Industrial and Management Optimization, 3(1),
127.
Shen, Z. J., Coullard, C. R., & Daskin, M. S. (2003). A Joint location-inventory model.
Transportation Science, 37, 4055.
Simchi-Levi, D., Kaminsky, P., & Simchi-Levi, E. (2000). Designing and managing the
supply chain. New York: Irwin McGraw-Hill.
Sridharan, R. (1995). The capacitated plant location problem. European Journal of
Operational Research, 87, 203213.
Tsiakis, P., Shah, N., & Pantelides, C. C. (2001). Design of multi-echelon supply chain
networks under demand uncertainty. Industrial and Engineering Chemistry
Research, 40, 35853604.
Tsou, Ching-Shih, Yang, Dong-Yuh, Chen, Jyun-Hao, & Lee, Ying-Hao (2011).
Estimating exchange curve for inventory management through evolutionary