Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/283531415

Optimization of Distribution Network with Backhaul

Article · July 2015

CITATIONS READS

0 131

1 author:

Chawis Boonmee
Muroran Institute of Technology
18 PUBLICATIONS 55 CITATIONS

SEE PROFILE

All content following this page was uploaded by Chawis Boonmee on 19 February 2018.

The user has requested enhancement of the downloaded file.


Optimization of Distribution Network with Backhaul
Chawis Boonmee, Chompoonoot Kasemset*

Department of Industrial Engineering, Chiang Mai University, Chiang Mai, Thailand


Email: chompoonoot.kasemset@cmu.ac.th*

Abstract part of transportation cost that makes the company


This article presents a mathematical model for waste its money when vehicles return to original
optimization of distribution network with backhaul. point with empty. To reduce backhauls, the selecting
The proposed mathematical model considered the of distribution centers and size of vehicle can be
selection of distribution centers, backhauls and the simultaneously considered to provide the effective
number of routes assigned for each vehicle size. The solution in order to reduce both fixed and variable
case study problem was solved to test the proposed costs in distribution network. However, to include
model when the distribution network composed of those issues in the distribution network problem, the
factories, distribution centers and suppliers. The problem is more complicated and difficult to find the
highlight of the case study was the reduction of optimal solution especially when the size of problem
backhauls introduced by the routes from distribution is increased as the real working situation. During the
centers to suppliers before vehicles going back to last decade, there are many research works using
factories. The proposed model was solved using optimization to solve many problems including the
Lingo software and the results of the case study problems in supply chain management.
showed the optimal total cost as 577,131 Baht per The objective of this study is to propose a
month. The sensitivity analysis was provided for mathematical model for distribution network
main related parameters. The interpretation of the optimization considering backhaul. The objective
results showed that the results from this model were function of this mathematical model is to minimize
valid and can be further implemented in the real the total cost of distribution network including the
working situation. selecting of distribution centers, reducing backhauls
and assigning the number of routes for vehicles with
different sizes. A case study is solved via Lingo
Keywords: Distribution Network, Optimization, software to verify and validate the proposed model.
Supply Chain Management, Backhaul The remainder of this paper is organized as follow.
Section 2 is literature reviews. The proposed
1. Introduction mathematical model is presented in section 3. Case
Supply chain management (SCM) emphasized on study is presented in section 4. Finally, the conclusion
planning system from upstream to downstream is provided in section 5.
consisting of producers, manufacturers, suppliers,
distribution centers, retailers and customers. A 2. Literature Reviews
distribution network problem is one of the most Supply chain management can be defined as a
comprehensive strategic decision problems that need design of efficient planning system from upstream to
to be optimized for long-term efficient operation of downstream to response to client’s need. The basic
the whole supply chain network. In this kind of components of supply chain are producers,
problem, the number, location, capacity and type of manufacturers, suppliers, distribution centers,
plants, warehouses, and distribution centers are retailers and customers. Many versions of definitions
determined to minimize the total cost of part or of supply chain management are provided in [2]-[6].
product distribution system. It also establishes The simple supply chain network [7] proposed in can
distribution channels, and amount of materials and be presented as shown in Figure 1.
items to consume, produce, and ship from suppliers to
customers [1].
A cost of supply chain system consists of fixed cost
and variable cost. For fixed costs, there are the costs
for operating factories and distribution centers that
are not related to the quantities of parts or products Figure 1 Supply chain network system [7]
passing through the distribution network. For variable
costs, there are transportation cost, storage cost and During the last decade, there were many
operation cost that are directly varied with the publications on the topic of supply chain management.
quantities of products or parts passing through the The distribution network design was one topic that
network. Thus, this kind of cost can be considered as many researchers worked on. Forecasting techniques
a critical issue in the distribution network problem. was proposed in [8] to find solution for supply chain
Transportation cost is one major cost of the network via the application of machine learning
distribution center for any company. Backhaul is one techniques in supply chain demand forecasting.
Learning technique was applied to manage the supply the fleet cost was dependent on the number of
chain system mentioned in [9]. Simulation model vehicles used and the total unused capacity. This kind
from [10] was used to evaluate several designs of a of problem was normally found in urban
supply network for a multi-site enterprise and transportation when several vehicles transiting on a
compared results from different control policies. demand point occurs were considered.
Transportation network optimization proposed in [11] The classification of distribution problems in the
presented the movement of indigenous goods in context of supply chain management according to
Amazonian Ecuador and optimization technique was two general specifications: problem definitions (or
adopted to obtain the minimum total cost of the model assumptions) and objective functions were
transportation network. presented as Table 2 based on coding system shown
Optimization was one of the important techniques in Table 1.
applied in many research works. It was the technique
used in finding an alternative with the most cost Table 1 Classification of distribution problems in
effective or highest achievable performance under the supply chain management
given constraints, by maximizing desired factors and Objective Cost C
minimizing undesired ones. This technique has been
Customer
applied to various problems such as production CR
responsiveness/backorders
planning, transport planning and many others ([12],
[13]). Activity time AT
Genetic algorithm approach was proposed in [1] Inventory I
for multi-objective optimization of supply chain Problem
networks in which three objectives were considered Customer Cus
Definition
as (1) minimization of total cost comprised of fixed Inventory levels/ordering
costs of plants and distribution centers, (2) IL
(Batch) size
maximization of customer service level, and (3)
Number of stages NS
maximization of capacity utilization balance for
distribution centers. Distribution center DC
Both qualitative and quantitative supply chain Plant P
performance measures were proposed in [3].
Supplier S
Qualitative measures include customer satisfaction,
flexibility, information and material flow integration, Production Pro
effective risk management, and supplier performance. Backhaul B
Quantitative performance measures include cost
Vehicle V
based measurements (i.e. cost minimization, sales
maximization, profit maximization, inventory
investment minimization, return on investment Table 2 concluded objective and problem definition
maximization) and customer responsiveness based found in literature reviews. Most objectives were
measurements including fill rate maximization, cost-based and problem definitions including plants,
product lateness minimization, customer response suppliers, distribution centers, vehicles and inventory
time minimization, lead time minimization, function levels.
duplication minimization.
Those objectives were considered under different Table 2 Literature Reviews Classification
constraints depended on the detail of each problem. Article’s Code
Reference
The constraints of supply chain management were No (Objective/Problem
Article(s)
material cost, internal transportation, external Definition)
transportation, fixed cost of factory, internal and Altiparmak et al.
1 C,CR/IL,NS,DC,P,S
external process cost, internal and external [1]
distribution network, inventory cost and Tavakkoli-Moghad
2 C/IL,B
transportation cost [14]. dam et al. [16]
Supply chain constraints were also addressed in Tavakkoli-Moghad
3 C/IL,S,V
[15] consisted of capacity, service compliance and the dam et al. [17]
extent of demand including supply chain decision 4 Bekrar et al.[18] I/C,IL,DC,P,S,P
variables, which are location, allocation, network 5 Day et at. [19] C/Cus,IL,P,V
structuring, number of facilities and equipment,
number of stage (echelons), service sequence, volume 6 Tang Jiafu [20] C/Cus,P,V
inventory level, size of workforce and the extent of Farahani and
7 C,I/Cus,NS,DC,P,S
outsourcing. Elahipanah [21]
Problems with other constraints were also C,CR/Cus,IL,NS,DC,P,
8 Chen et al. [22]
addressed in [16]. Capacitated vehicle routing Pro
problem with split service for minimizing fleet cost Alptekinoglu and
9 C/Cus,IL,DC,P,S,Pro
was proposed. The results of this work showed that Tang [23]
Backhaul occurred when a vehicle carried goods to i. Backhaul was reduced when introducing the routes
a destination, and then it returned empty to the from suppliers to plants. The mathematical model
starting point. In this situation, the transportation cost considered the selection of the number of appropriate
was almost twice the cost of the single shipment at distribution centers, the size of vehicles and the
destination. If a manufacturer or distributor of a route number of traveling time by each vehicle.
can bring a product back instead of an empty
backhaul, the backhaul costs can be reduced and the 3.1 Assumptions
efficiency of the distribution network can increase. To formulate the proposed mathematical model,
When backhaul was considered, decisions on the following assumptions were considered:
distribution of products are more complicated. 1. All used parameters are known, constant and
Memetic algorithm was proposed in [16] for solving deterministic.
vehicle routing problem with backhauls because the 2. The proposed mathematical model included
complexity and the large size of the problem make it plants, distribution centers and suppliers with plants
difficult to solve by the exact algorithm. as the last destinations (as shown and discussed as
Many research works on supply chain management Figure 2).
usually applied either mathematical model or
heuristic technique to find the optimal or 3.2 Mathematical Model
near-optimal solutions. During the last decade, the Indices:
mathematical model was a popular approach for
solving problems in supply chain management i = Number of factory (1, 2, 3, …, m)
especially for distribution network problems because j = Number of
the solution from this technique can guarantee as distribution center (1, 2, 3, …, q)
global optimal solution. However, the limitation of k = Number of supplier (1, 2, 3, …, k)
the mathematical model was addressed in [3] that this l = Vehicle Size (1, 2, 3, …, r)
technique was appropriate for solving small-size
problems with global optimal solution while the Decision Variable:
heuristics can be applied to solve the big-size
problems to obtain near-optimal solution within short 1 , if distribution center
computational time. Yj = j is open
For this study, the size of the case study problem is 0 , otherwise.
small so the mathematical model of the case study
was developed and solved using exact method in the R_IJijl = The number of routes from
following section. factory i to distribution center j
by vehicle size l
3. Proposed Mathematical Model R_JKjkl = The number of routes from
distribution center j to supplier
In this section, the proposed mathematical model k by vehicle size l
for distribution network optimization considering R_KIkil = The number of routes from
backhaul for this case study is presented as shown in supplier k to factory i by
Figure 2. vehicle size l
R_JIjil = The number of routes from
Plant (i) Distribution Centre (j) Supplier (k) distribution center j to factory i
by vehicle size l
R_IKikl = The number of routes from
factory i to supplier k by
vehicle size l
Tranijl = The number of product
transported from factory i to
distribution center j by vehicle
size l
Tran_matkil = The number of material
transported from supplier k to
factory i by vehicle size l
Figure 2 Distribution network of the case study T_routel = Total route of vehicle size l

From Figure 2, the distribution network composed Parameters:


of plants, distribution centers and suppliers.
Vehicles left from factory i to distribute products to Demandi = The number of product needed
distribution center j and then vehicles went to pick up to be supplied from factory i
raw material at supplier k before went back to factory Mati = The number of material needed
by factory i
Warej = Capacity of distribution center j  R _ IJ (i, j, l )   R _ IK (i, k , l )   R _ KI (k , i, l ) 
Cap_prol = Capacity of vehicle size l for j k k

transporting products  R _ JI ( j, i, l )
Cap_matl = Capacity of vehicle size l for j

transporting materials i, l (8)


Cost_capl = Maintenance cost of vehicle size
l per time
 R _ KI (k , i, l )   R _ JK ( j, k , l )   R _ IK (i, k , l )
i j i
Fixj = Fixed operating cost of k , l (9)

C_IJijl =
distribution center j
Transportation cost from factory i  R _ IJ (i, j, l ) T _ route(l )
i j
l (10)
to distribution center j by vehicle
size l Y ( j)  Ware _ m
j
(11)
C_JKjkl = Transportation cost from
distribution center j to supplier k Tran _ mat (k , i, l )  Sup(k )
i l
k (12)
by vehicle size l Y(j) = Binary Variable (13)
C_KIkil = Transportation cost from supplier Tran(i,j,l) = General Integers (14)
k to factory i by vehicle size l R_IJ(i,j,l) = General Integers (15)
C_JIjil = Transportation cost from R_JK(j,k,l) = General Integers (16)
distribution center j to factory i R_KI(k,i,l) = General Integers (17)
by vehicle size l
C_IKikl = Transportation cost from factory i The proposed model was a mixed integer linear
to supplier k by vehicle size l programming model. The objective function in
Ware_m = The number of distribution center equation (1) was to minimize the total cost of
in the network distribution network including transportation costs
Supk = Capacity of supplier k and costs of operating distribution centers. Equation
(2) was to limit the number of products transported
Objective: not to exceed the capacity of each distribution center.
Equation (3) was to ensure that for the total products
Minimize Z (Total cost of distribution network) transported should be equal to the number of product
= supplied for each factory i. Equation (4) was to
 Fix( j)*Y ( j)   Cost _ cap(l )*T _ route(l ) 
j l
confirm that the total amount of material received
should be equal to the amount of material needed for
 C _ IJ (i, j, l )* R _ IJ (i, j, l ) 
i j l
each factory i. Equations (5) and (6) was to ensure
that the total amount of transported products and
 C _ JK ( j, k , l )* R _ JK ( j, k , l )  material of each time should not exceed the capacity
j k l (1) of each vehicle size. Equations (7) to (9) was the
 C _ KI (k , i, l )* R _ KI (k , i, l ) 
k i l
constraint to ensure that the number of routes going
into each node should be equal to the number of
 C _ JI ( j, i, l )* R _ JI ( j, i, l ) 
j i l
routes departure from that node. Equation (10) was to
confirm that the summation of routes for each vehicle
 C _ IK (i, k , l )* R _ IK (i, k , l ) 
i k l
size should be equal to the total number of traveling
routes of each vehicle size. Equation (11) ensures that
the limit on the number of opened distribution centers
Subject to should be less than or equal to the total number of
available distribution center in the network. Equation
Tran(i, j, l )  Ware( j)*Y ( j) j (2) (12) was to confirm that the summation of the amount
i l of material transported should not be exceed the
Tran(i, j, l )  Demand (i) i (3) amount of material supplied by supplier k. Equation
j l (13) was a binary constraint and equations (14) to
Tran _ Mat (k , i, l )  Mat (i)
k l
i (4) (17) were the integer variable constraints.

R _ IJ (i, j, l )* Cap _ pro(l )  Tran(i, j, l ) 4. Case Study


(5)
i, j, l The case study problem consisted of two factories,
R _ KI (k , i, l )* Cap _ Mat (l )  Tran _ Mat (k , i, l ) four distribution centers and three suppliers. There
were three different vehicle sizes. The parameters
k , i, l (6) used in this case study were shown in Tables 3 to 9.
i
R _ IJ (i , j , l )  
k
R _ JK ( j , k , l )   R
i
_ JI ( j , i , l ) The product demand and raw material needed were
monthly data.
j, l (7)
Table 3 Product demand and raw material needed Table 8 Transportation cost from supplier to
for each factory (unit: pallet) factory (unit: Baht)
Factory 1 Factory 2 Supplier
Factory Vehicle
Product Demand 887 782 1 2 3
Raw Material Needed 950 850 1 4,384 4,372 4,303
1 2 4,575 4,562 4,490
Table 4 Capacity and fixed cost of each distribution 3 5,261 5,246 5,164
center 1 4,298 4,286 4,219
DC 1 2 3 4 2 2 4,485 4,472 4,402
Capacity 3 5,158 5,143 5,063
785 785 500 2,000
(unit: pallet)
Fixed Cost Table 9 Limit of materials of each supplier
20,000 20,000 10,000 3,5000
(unit: Baht1) Supplier 1 2 3
Limit of
Table 5 Capacity and maintenance cost of each 1,500 1,000 500
materials
vehicle
Size of Vehicle 1 2 3 4.1 Lingo Model
Product 10 12 20 The proposed mathematical model was solved by
Capacity pallets pallets pallets Lingo software. Lingo model for the case study was
Material 25 28 50 presented as Figure 3.
Capacity pallets pallets pallets From Figure 3, the model was divided into three
Maintenance 1,000 1,200 1,700 parts. The first part was how to define the used set
Cost Baht Baht Baht and set members (variables and parameters). The
second part was the proposed mathematical model in
Table 6 Transportation cost from factory to the form of Lingo code written following equation (1)
distribution center (unit: Baht) to equation (17). The third part was input data for the
Factory case study.
DC Vehicle
1 2
1 61 35
Part 1
1 2 64 41
3 75 52
1 83 74
2 2 88 79
3 95 86
1 74 83 Part 2
3 2 79 88
3 86 95
1 4,209 4,119
4 2 4,433 4,302
3 4,995 4,801

Table 7 Transportation cost from distribution center


to supplier (unit: Baht)
Supplier
DC Vehicle
1 2 3
1 4,415 4,403 4,334
1 2 4,607 4,594 4,523
3 5,298 5,283 5,201 Part 3
1 4,434 4,422 4,372
2 2 4,627 4,614 4,562
3 5,321 5,306 5,246
1 4,384 4,374 4,303
3 2 4,575 4,564 4,490
3 5,261 5,249 5,164
1 286 361 305
4 2 298 377 318
3 343 433 366

1
Approximately 32 Baht = 1$ Figure 3 Lingo model for the case study
4.2 Results Table 14 Number of routes from each distribution
The solution from Lingo showed the optimal total center to each factory
cost of distribution network was 577,131 Baht per DC Factory 1 Factory 2
month (shown as Figure 4). The distribution center Vehicle 1 2 3 1 2 3
number 1, 2 and 3 were selected. The number of 1 - - 6 1 1 21
transportation and the number of used vehicles were 2 1 - 19 - - -
concluded as in Tables 10 to 14 and Figure 5. 3 - - - - - -
4 Closed

Plant (i) Distribution Centre (j) Supplier (k)


[3,3]
[6,3]

3 ]
[6,

]
[1

1,3
1,
3]

[2
Distribution (1)

2],
[1,
1],
Supplier (1)

[1,
[1,1], [19,3]
[1,1], [19,3]

Plant (1) [25


,3] Distribution (2) ]

]
8,3
,3
[6

[3
[9,3] Supplier (2)

2],
[1,
[10,3

1],
]

[1,
Distribution (3)
Plant (2)
Supplier (3)
Figure 4 Solution from Lingo
Distribution (4)
Table 10 Number of routes for each vehicle size [10,3]
[17.3]
Vehicle Size 1 2 3
Figure 5 The optimal distribution network for the
Total Routes 2 1 82 case study [The number of route, the size of vehicle]
Table 11 Number of routes from each factory to The minimum total cost 577,131 Baht can be
each distribution center separated as the total fixed cost of opened distribution
Factory centers 50,000 Baht, the total maintenance cost
DC Vehicle
1 2 142,600 Baht, the total transportation cost from
1 - 1 factories to distribution centers 6,090 Baht, the total
1 2 - 1 transportation cost from distribution centers to
3 - 38 suppliers 188,560 Baht, the total transportation cost
1 1 - from suppliers to factories 186,375 Baht and the total
2 2 - - transportation cost from distribution centers to
3 19 - factories 3,506 Baht.
1 - -
3 2 - - 4.3 Sensitivity Analysis
3 25 - Sensitivity Analysis was the study of how
1 uncertainty in the output of a model can be
4 2 Closed apportioned to different sources of uncertainty in the
3 model input [24]. The mathematical model of this
study was considered as a mixed integer
Table 12 Number of routes from each distribution programming (MIP) problem that included both
center to each supplier non-negativity and integer variables. Solving by
DC Supplier 1 Supplier 2 Supplier 3 Lingo software, there was no report for sensitivity
Vehicle 1 2 3 1 2 3 1 2 3 analysis when MIP was study. Thus, for this case
1 - - - - - 11 - - - study, the sensitivity analysis can be provided by
2 - - - - - - - - - manually changing values of each parameter. The
3 - - 6 - - 9 - - 10 range of parameter without causing any of the
optimal values of the decision variables to change can
4 Closed
be explained for 4 parameters as follows.
For (1) product demand and (2) raw material
Table 13 Number of routes from each supplier to
needed, they were right-side values of constraints.
each factory
These values should be within ranges provided in
Supplier Factory 1 Factory 2 Table 15.
Vehicle 1 2 3 1 2 3
1 - - 6 - - -
2 - - 3 - - 17
3 - - 10 - - -
Table 15 Range of product demand and raw Reference
material needed with unchanged basic variables [1] F. Altiparmak , M. Gen , L. Lin and T. Paksoy,
(unit: pallet). “A genetic algorithm approach for multi-objective
Factory 1 Factory 2 optimization of supply chain networks”,
Current 887 782 Computers & Industrial Engineering, Vol.51(1),
Product
Min 885 781 pp.196-215, 2006.
Demand
Max 890 782 [2] D. M. Lambert, M. C. Cooper and J. D. Pagh,
Raw Current 950 850 “Supply Chain Management: Implementation
Material Min 929 829 Issues and Research Opportunities”, The
Needed Max 950 850 International Journal of Logistics Management,
Vol.9(2), pp.1–19, 1998.
From Table 15, changeable range of both product [3] B. M. Beamon, “Supply chain design and
demand and raw material needed were quite narrow, analysis: Models and methods”, International
so these can be interpreted that the optimal solution Journal of Production Economics, Vol.55(3),
was sensible to both of these parameters. pp.281-294, 1998.
For (3) maintenance cost of each vehicle and (4) [4] C. Vercellis, “Business Intelligence”, United
fixed cost of opening each distribution center, they Kingdom: TJ International, 2009.
were values of objective function’s coefficient that [5] K. Apipatchayasakun, “Manufacturing Logistics
directly gave effect on the objective value. and Operation Management”, C.Y.C. Printing,
For maintenance costs, these values were affected Nontapure: focus media and publicing. Inc, 2006.
on the number of vehicles used for each vehicle type. [6] G. Arzu Akyuz and T. Erman Erkan, “Supply
Currently, the minimum maintenance cost per unit chain performance measurement: a literature
carried was at vehicle 3. From sensitivity analysis, if review”, International Journal of Production
this cost of vehicle size 1 and 2, current values as Research, Vol.48(17), pp.5137-5155, 2010.
1,000 and 1,200 Baht, were reduced to less than 833 [7] S. Phonsuwan and V. Kachitvichyanukul,
and 1,072 Baht, respectively, the number of vehicles “Simulation for OTOP supply chain with hybrid
used for all vehicle sizes will be changed. push-pull flow control”, International Journal of
For fixed cost of opening each distribution center, Services and Operations Mannagement, Vol.15(2),
the current solution was to open distribution center 1, pp.196-214, 2013.
2 and 3. From sensitivity analysis, if the fixed cost of [8] R. Carbonneau, K. Laframboise and R. Vahidov,
distribution center 4 is decreased to lower than “Application of machine learning techniques for
25,599 Baht (current cost at 35,000 Baht), the supply chain demand forecasting”, European
solution will change to select distribution 1, 3 and 4. Journal of Operational Research, Vol.184(3),
This can be said that the reduced cost of decision pp.1140-1154, 2008.
variable Y4 are 9,401. [9] P. J. Martínez-Jurado and J. Moyano-Fuentes,
“Lean Management, Supply Chain Management
5. Conclusions and Sustainability: A Literature Review”, Journal
This research paper presented the mathematical of Cleaner Production, pp.1-17, 2013.
model for distribution network optimization with [10] F. Pirard, S. Iassinovski and F. Riane, “A
backhaul. The objective of this model was to simulation based approach for supply network
minimize total cost when considering the selection of control”, International Journal of Production
distribution centers, backhauls and the number of Research, Vol.49(24), pp.7205-7226, 2011.
routes for different sizes of vehicles. The distribution [11] J. Kelley, M. Kuby and R. Sierra,
network of this study composed of factories, “Transportation network optimization for the
distribution centers and suppliers that physical flows movement of indigenous goods in Amazonian
were different comparing with other cases. Since, the Ecuador”, Journal of Transport Geography,
reduction of backhauls was proposed by introducing Vol.28, pp.89-100, 2013.
vehicles going to pick up raw material when they left [12] W. L. Winston and M. Venkataramanan,
from distribution centers before going back to “Introduction to mathematical programming”,
factories. The sensitivity analysis showed that the United States of America: Curt Hinrichs, 2003
optimal solution was sensible to parameters of [13] Lindo Systems inc., “Optimization Modeling
product demand and raw material needed due to the with LINGO”, United States of America: Lindo
narrow allowable to decrease and increase ranges. Systems inc, 2007.
The results of the proposed model were valid and [14] J. F. Shapiro, “Modeling the Supply Chain”,
this can proof that the logic of the proposed model United States of America: Transcontinental
was correct. Thus, the results from this study can be Printing/Louiseville, 2007.
implemented in the real situation of work to help the [15] M. Hokey, J. Vaidyanathan and S. Rajesh,
company to reduce the cost in product distribution. “Combined location-routing problems: A
synthesis and future research directions”,
European Journal of Operation research Vol.108,
pp.1-15, 1998.
[16] R. Tavakkoli-Moghaddam, N. Safaei, M. M. O. Chawis Boonmee is a
Kah and M. Rabbani, “A New Capacitated Vehicle graduate student in the
Routing Problem with Split Service for Department of Industrial
Minimizing Fleet Cost by Simulated Annealing”, Engineering, Faculty of
Journal of the Franklin Institute, Vol.344(5), Engineering, Chiang Mai
pp.406-425, 2007. University, Thailand. He
[17] R. Tavakkoli-Moghaddam, A. R. Saremi and M. received the B.Eng. degree in
S. Ziaee, “A memetic algorithm for a vehicle Industrial Engineering from
routing problem with backhauls”, Applied Chiang Mai University in
Mathematics and Computation, Vol.181(2), 2012. He interested in the field
pp.1049-1060, 2006. of industrial engineering
[18] A. Bekrar, D. Trentesaux and B. Beldjilali, including optimization, modern production &
“Multi-stage Optimization in Supply Chain: An industrial system, operation & supply chain
Industrial Case study”, Proceedings of management, simulation and Material Flow Cost
MOSIM'12, 2012. Accounting (MFCA).
[19] J. M. Day, P. Daniel Wright, T. Schoenherr, M.
Venkataramanan and K. Gaudette, “Improving Chompoonoot Kasemset
routing and scheduling decisions at a distributor of is an assistant professor in the
industrial gasses”, Omega, Vol.37(1), pp.227-237, Department of Industrial
2009. Engineering, Faculty of
[20] J. Tang, J. Zhang and Z. Pan, “A scatter search Engineering, Chiang Mai
algorithm for solving vehicle routing problem University, Thailand. She
with loading cost”, Expert Systems with received a D.Eng. in Industrial
Applications, Vol.37(6), pp.4073-4083, 2010. Engineering and Management,
[21] R. Z. Farahani and M. Elahipanah, “A genetic School of Engineering and
algorithm to optimize the total cost and service Technology at Asian Institute
level for just-in-time distribution in a supply of Technology in 2009. Her
chain”, International Journal of Production research interests include operations management,
Economics, Vol.111(2), pp.229-243, 2008. applied operations research, simulation application in
[22] C. Chen and W. Lee “Multi-objective production management and Material Flow Cost
optimization of multichannel supply chain Accounting (MFCA). Her special field is Theory of
net-works with uncertain product demands and Constraints (TOC).
prices”, Computers & Chemical Engineering,
Vol.28, pp.1131–1144, 2004.
[23] A. Alptekinoglu and C. S. Tang, “A model for
analyzing multichannel distribution systems”,
European Journal of Operational Research,
Vol.163 (3), pp.802–824, 2005.
[24] A. Saltelli, M. Ratto, T. Andres, F. Campolongo,
J. Cariboni, D. Gatelli, M. Saisana and S.
Tarantola, “Global Sensitivity Analysis”, England:
John Wiley & Sons Ltd, 2008.

View publication stats

You might also like