Available online at www.sciencedirect.

com

ScienceDirect
Procedia CIRP 57 (2016) 522 – 526

49th CIRP Confference on Manufacturi

M ing Systemss (CIRP-CM
MS 2016)

A New
w Model for Susttainable Changeeability and
a Prodduction Planning
P g
S. Ghan
nei, T. AlG
Geddawy േ
Departm
ment of Mechanicaal and Industrial Engineering, University of Minneesota Duluth, MN
N 55812, USA

* Corresponding autthor. Tel.: +1-2188-726-6810; fax: +1-218-726-8596

+ 6. E-mail address:
s geddawy@d.um
mn.edu

Abstract

Chaangeable Manuufacturing Systeems have the capability

c to adaapt to varying production plan
ns and differen
nt product desiggns by changing their
connfiguration andd layout. This paper
p presents a new linear mixed integer mathematical model to maxximize sustainaability of Channgeable
Maanufacturing Syystems based on n the daily varyying energy priccing. The daily production demmand of severaal product variannts has to be saatisfied
by corresponding configuration of the manuffacturing system m. System connfiguration plaanning consistss of machine arrangement
a annd job
seqquencing for eacch planning dayy. The proposeed linear mixedd integer mathem matical model is
i solved by CP PLEX solver inn GAMS software for
ninne different prooblem sizes. Th he new LMI modelm finds the optimum connfiguration plaan and job seqquence in a reaasonable time, which
illu
ustrates the efficciency and praccticality of the proposed
p modell.

Keyywords: Changeaable Manufacturinng System; Sustaainability; Linear Mathematical Model; Gams;

1. Introduction
I
Midnight Midnight Midnig
ght
8.7 Offf-peak
ۖ/kW
Changing tthe configu uration andd layout enable e
maanufacturing ssystems to addapt to differrent product designd 13.2
ۖ/kW Mid-peak
andd production pplans. Howevver, the cost of o reconfigurinng the
Chhangeable Mannufacturing Systems is nott only dependdent on 18 On
n-peak
ۖ/kW
thee difficulty of the transition from a configguration to thee next, Noon Noon Noon
n h
butt also on the ttime of the daay through whhich it is perfoormed, Summer Weekends Winteer
sin
nce energy priccing changes throughout
t the day.
The consumpption of energgy in industrial sector hass been
inteensively increeased over lastt decades. 50%% of the total global Figgure 1. A sample of TOU electriciity price(Ontario energy Board) [1
13]
eneergy consumpption belongs to industrial sector which is the
There are several studies through the last decade which
moost important energy
e consummer in the woorld [1]. Theree is an
shoow the barriiers and the policies of energy efficciency
inccreasing pressuure on manuffacturers to deecrease their carbon
c
proogram implem mentation [5,,6]. Demand--Side Managgement
foootprint level ddue to cost off energy and climate changge [1].
(DS SM) considerss peak decreassing and load shifting to offf-peak
Imp proving the ennergy efficienncy of manufaacturing proceess and
from on-peak ennergy tariff rattes in order too reduce energgy cost
maanagement tactics help redu uce energy connsumptions [2].
[7]. Great endeaavor in DSM have h resulted in many beneefits in
The literaturee on energy efficient
e manuufacturing proocesses
ecoonomic and environmenta
e l points of vview. A varieety of
andd machinery is rich [3]. However,
H in a mass prod duction
straategies has beeen implementted in some coountries to decrease
envvironment, thhe majority of o the energyy is consum med by
eneergy consumpption cost such h as adopting time-of-use (TOU)
(
ind
direct productiion operationss [4]. The focus of energy saving
s
priccing with greeat differencess between onn-peak and off ff-peak
is to
t improve m machines or prrocesses consuumption sepaarately,
tariiff rate [8]. Th
he changes in energy pricess play a cruciaal role
whhich results in missing the opportunity
o off energy savinng by a
in energy
e consummption costs.
sysstem-level appproach [1].

2212-8271 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the 49th CIRP Conference on Manufacturing Systems
doi:10.1016/j.procir.2016.11.090
 S. Ghanei and T. AlGeddawy / Procedia CIRP 57 (2016) 522 – 526
In this paper, TOU pattern of energy pricing is considered
(figure 1). TOU is identified as a successful pricing pattern to
find electricity demand response [8]. TOU structure divides
the full day into the on-peak and off-peak periods in most
cases. Sometimes a mid-peak period is also considered as a
time period [8]. The on- and mid-peak periods are defined as
daylight hours while the off-peak period is defined as night
time hours. All these period times are the Local Standard
Time [8]. For instance, each state of U.S.A has unique on-,
mid-, off- peak periods with different rates of energy price. In
the United States, industrial sector accounts for approximately
34% of the country’ total energy consumption [1].
This paper introduces a novel linear mixed integer
mathematical model to maximize sustainability of the
Changeable Manufacturing Systems. The daily production
demand of several product variants has to be satisfied by
corresponding configurations of the manufacturing system
according to unstable energy price. System configuration
planning consists of machine arrangement and job sequencing.
The proposed model considers three main factors that affect
system sustainability in the environmental and economic
domains, which are, 1) the change pattern in energy prices
throughout the day, 2) the transportation cost of jobs between
machines, which depends on machines locations in the system,
3) the setup cost of each machine, which is dependent on the
job sequence. The model output is a system configuration
plan, indicating arrangement of machines in the system, and
the sequence of jobs, which need to be produced on that day.
In the following section, a brief literature review is
provided, and the mathematical model is described in detail in
section 3. Section 4 presents numerical experiments. Finally,
the study’s conclusion and future research are presented in
section 5.
2. Literature review
Flexible and changeable manufacturing systems have
played an effective role in energy saving. Many studies about
the energy and resources efficiency of manufacturing systems
has been conducted based on increasing attention to global
environmental and energy saving concerns in manufacturing.
Choi and Xirouchakis proposed a linear holistic production
planning model to minimize energy consumption and
maximize the output in a changeable manufacturing system.
The proposed model considers numerous part handling
systems in detail to find an optimum energy consumption
model [9]. Rajabinasab and Mansour consider different set-up
planning and pallet arrangements as alternative process plans
in the dynamic flexible job shop scheduling problem to
minimize the energy consumption [10]. Fang et al. presented a
new multi-objective mixed integer mathematical model to
optimize energy consumption and peak power load of shop
scheduling problem [1].
Kuster et al. proposed a new approach of changeable
manufacturing system utilizing evolutionary algorithm to
rearrange and optimize the production schedule with the aim
of energy consumption reduction. Shifting energy consuming
processes and their interconnected sub-process required
reconfiguring of the individual processes to minimize the
523
energy cost [11]. Yusta et al. presented a mathematical model
to optimize production schedule and maximize the system
throughput by simulating hourly electricity price differences
and demand of a machining process [12].
Luo et al. proposed a multi-objective model for hybrid flow
shop scheduling problem based on ant colony optimization
which aims to increase the efficiency of electricity
consumption based on time of use prices (TOU) and minimize
the makespan [13]. Tan et al. considered the scheduling
problem in the steel manufacturing process in two phases
based on the variable electricity price [14].
Moon et al. proposed a hybrid genetic algorithm to
maximize the energy and production efficacy of the dissimilar
parallel machine scheduling problem. Decision maker can use
this approach to find a solution by the use of weighted sum
objective to minimize the makespan of production and time-
dependent electricity costs [15].
Wang et al. developed a systematic optimization approach
for scheduling and process planning of shop floor problem in
order to optimize multi objectives including enhancement of
adaptability, productivity and energy efficiency of system.
The proposed approach includes a process stage and a system
stage. In this problem, milling a part in the process stage and
scheduling of whole shop floor in the system stage are
optimized in order to increase energy efficiency and
productivity [8].
Shrouf et al. provided a mathematical model for single
machine scheduling problem to optimize the energy
consumption costs. In this model, job processing time on the
machine, idle time, “turning off” and “turning on” of machine
are determined by applying genetic algorithm to implement
the sustainable production scheduling and minimize energy
consumption throughout peak periods [16].
Several optimization-based approaches have been studied
for production planning in changeable manufacturing systems,
including heuristics, metaheuristics, systematic approaches,
and mathematical programming. None of these studies,
considered energy related objective in linear mathematical
modeling of the system configuration plan, including
arrangement of machine, and the sequence of jobs in
changeable manufacturing system.
3. Problem description
3.1. Assumptions
The system consists of a number of machines which have
to be arranged in a closed loop layout. A number of different
jobs are required to satisfy the daily production demand. Each
job should be processed by a unique set of machines in a
specified order. The sequence of required machines for
processing each job is given. A job represents the demand of a
specific product or product variant. A job requires at least one
machine and at most all machines in the line in order to be
finished. There is a unidirectional material handling system to
transport work-in-process from one machine to another. A job
may need the same machine more than one time to be done.
Each job starts processing immediately when the last
preceding job processes are completely finished. A maximum
524 S. Ghanei and T. AlGeddawy / Procedia CIRP 57 (2016) 522 – 526

working hour in a day is 12 hours. The energy consumption σெ

ெሺெାଵሻ (2)
௠ୀଵ ‫ݍ‬௠ ൌ ଶ

cost is considered based on the TOU pattern of energy pricing
throughout a day. ‫ݍ‬௠ െ ‫ݍ‬௡ ൒ ߝǤ ݄௡௠ െ ‫ܯܤ‬Ǥ ݄௠௡  ‫݉׊‬ǡ ݊Ǣ ݉ ് ݊ (3)
‫ݍ‬௠ ൒ ͳ  ‫݉׊‬ (4)
3.2. Mathematical model
݄௠௡ ൅ ݄௡௠ ൌ ͳ ‫݉׊‬ǡ ݊ Ǣ ݉ ് (5)
In this study, a new mixed integer linear mathematical ݊
model is presented to minimize the total energy consumption
cost, the total transportation cost of jobs between machines ‫ݔ‬௠௡ ൑ ‫ݍ‬௠ െ ‫ݍ‬௡ ൅ ‫ܯܤ‬ሺͳ െ ݄௡௠ ሻ‫݉׊‬ǡ ݊ (6)
and the total set up cost of machines simultaneously. The ‫ݔ‬௠௡ ൑ ‫ ܯ‬െ ‫ݔ‬௡௠ ൅ ‫ܯܤ‬ሺͳ െ ݄௠௡ ሻ‫݉׊‬ǡ ݊ (7)
following notations are used in the mathematical model.
‫ݔ‬௠௡ ൒ ‫ݍ‬௠ െ ‫ݍ‬௡ െ ‫ܯܤ‬ሺͳ െ ݄௡௠ ሻ ‫݉׊‬ǡ ݊ (8)
3.2.1. Indexes ‫ݔ‬௠௡ ൒ ‫ ܯ‬െ ‫ݔ‬௡௠ െ ‫ܯܤ‬ሺͳ െ ݄௠௡ ሻ‫݉׊‬ǡ ݊ (9)
σ௃௝ୀଵ σெ ெ
௠ୀଵ σ௡ୀଵ ܶܵ௝௠௡ Ǥ ‫ݔ‬௠௡ ൑ ܶ‫ܥ‬ (10)
݉ǡ ݊: Machine, m= 1,2, …, M, where M is the number of
machines
݆ǡ ݅ǡ ݈: Job, j= 1,2, …, J , where J is the number of jobs
‫ݐ‬: Time, t= 1,2, …, T, where T is working time of a day σ௃௜ୀଵ ‫ݕ‬௜௝ ൑ ͳ ‫݆׊‬ (11)

3.2.2. Parameters σ௃௝ୀଵ ‫ݕ‬௜௝ ൑ ͳ‫݅׊‬ (12)

‫ݕ‬௜௝ ൅ ‫ݕ‬௝௜ ൌ ͳ ‫݅׊‬ǡ ݆ Ǣ ݅ ് (13)

ܶ‫݌‬௝ ǣ Processing time of job j
‫ܧ‬௝ ǣ Energy consumption of job j ݆
‫݌ܧ‬௧ ǣ Energy price in time t ‫ݕ‬௜௝ ൅ ‫ݕ‬௝௟ ൅ ‫ݕ‬௜௟ ൑ ʹ ‫݅׊‬ǡ ݆ǡ ݈ Ǣ ݅ ് (14)
ܶܵ௝௠௞ ǣ Transportation cost of job j from machine m to
machine k ݆ ് ݈
ܵܶ௝௜௠ ǣ Setup cost of machine m for processing job j when σ௃௝ୀଵ σ௃௜ୀଵ ‫ݕ‬௜௝ ൌ ‫ ܬ‬െ ͳ (15)
job i precedes job j ௝ஷ௜
‫ܯܤ‬ A large number
ߝ A small number σெ ௃ ௃
௠ୀଵ σ௝ୀଵ σ௜ୀଵ ܵܶ௝௜௠ Ǥ ‫ݕ‬௜௝ ൑ ܵ‫ܥ‬
(16)

‫ܥ‬௜ ൅ ܶ‫݌‬௝ െ ‫ܯܤ‬൫ͳ െ ‫ݕ‬௜௝ ൯ ൑ ‫ܥ‬௝ ‫݅׊‬ǡ ݆ Ǣ ݅ ് (17)

3.2.3. Decision Variable
݆
‫ݔ‬௠௞ Distance between machine m and machine k
‫ݕ‬௜௝ Equal to 1 if job i precedes job j, and 0 otherwise ܶ‫݌‬௝ ൫ͳ െ ‫ݕ‬௜௝ ൯ ൑ ‫ܥ‬௝  (18)
݄௡௠ Equal to 1 if machine m precedes machine n in ‫݅ ׊‬ǡ ݆ Ǣ ݅ ് ݆
layout
‫ݓ‬௝௧ Equal to 1 from beginning of a day to completion ‫ܥ‬௝ െ ‫ ݐ‬൅ ͳ ൑ ‫ܯܤ‬൫‫ݓ‬௝௧ ൯ ‫݆׊‬ǡ ‫ݐ‬ (19)
time of job j ‫ܥ‬௝ െ ‫ ݐ‬൅ ͳ െ ܶ‫݌‬௝ ൑ ‫ܯܤ‬൫‫ݓ‬Ԣ௝௧ ൯‫݆׊‬ǡ ‫ݐ‬ (20)
‫ݓ‬Ԣ௝௧ Equal to 1 from beginning of a day to start time of
job j ‫ܥ‬௝ െ ‫ ݐ‬൒ െ‫ܯܤ‬൫ͳ െ ‫ݓ‬௝௧ ൯ ‫݆׊‬ǡ ‫ݐ‬ (21)
‫ݍ‬௠ Position of machine m (1, 2,…,M)
‫ܥ‬௝ െ ‫ ݐ‬െ ܶ‫݌‬௝ ൒ െ‫ܯܤ‬൫ͳ െ ‫ݓ‬Ԣ௝௧ ൯‫݆׊‬ǡ ‫ݐ‬ (22)
‫ܥ‬௝ Completion time of job j
‫ܥܧ‬௝௧ Energy consumption cost of job j in time t ൫‫ݓ‬௝௧ െ ‫ݓ‬Ԣ௝௧ ൯ Ǥ ‫݌ܧ‬௧ Ǥ ‫ܧ‬௝ ൑ ‫ܥܧ‬௝௧ ‫݆ ׊‬ǡ ‫ݐ‬ (23)
ܶ‫ܥܧ‬ Total energy consumption cost
ܵ‫ܥ‬ Total setup cost σ௃௝ୀଵ σ்௧ୀଵ ‫ܥܧ‬௝௧ ൑ ܶ‫ܥܧ‬ (24)
ܶ‫ܥ‬ Total transportation cost
The problem is formulated as follow:

3.2.4. Problem Formulation

‫ ܿܶ݊݅ܯ‬൅ ܵ‫ ܥ‬൅ ܶ‫ܥܧ‬ (1)

S.T
 S. Ghanei and T. AlGeddawy / Procedia CIRP 57 (2016) 522 – 526 525

Figgure 2. Best arranngement of machines

Eqquation (1) reepresents the objective funnction addresssing

minim mization of associated coosts to energgy consumpttion,
sequeence-dependennt setup of machines
m and transportationn of
jobs between macchines in an attempt to find the optiimal
solution. Equationns (2) to (4) guuarantee that each
e machine has Figuure 3. Electricity consumption cosst based on TOU pattern
p
a uniique position nnumber betweeen set of 1 too M. Equationn (5)
ensurres that sequuence of eacch two machhines is uniqque. Reegarding the survey
s of TOUU electricity pricing
p in the U.S
U
Equaations (6) to (9) calculate distances beetween each two conduucted by Wanng and Li, tim me and price of o on-, mid-, off-
mach hines based on o their sequeence while prroduction line is peak periods of electricity are varies
v in eachh state [8]. In this
closeed loop and thhere is one-waay to transporrt jobs in the line.
l paperr, it is assumeed that manuffacturing systeem is in AZ states
It shoould be notedd that equatioons (7) and (99) guarantee that as onne of most exp pensive electrricity pricing zones
z of U.S [8].
sum ofo distances frfrom machine m to machinee n and vice veersa Basedd on the surveey, the TOU pattern
p of elecctricity pricingg in
is eqqual to M. Inn other word ds, these equuations guaranntee AZ State
S has fourr periods for 12 hours (frrom 8 A.M to t 8
mach hines to be arrranged in the closed
c loop linne. Equation (10)
( P.M): 1) the period of 8 to 11 A.MA (from 1 to 18 TU) is off-
comp putes total traansportation coost of jobs beetween machinnes. peak period. 2) The period of 111 A.M to 2 A.M M (from 19 too 36
Equaations (11) to (14) determine the sequeence of jobs and TU) isi Mid-peak period.
p 3) The period of 2 too 7 P.M (from m 37
ensurre that just a uunique job precedes the othher job. Equaation to 65 TU) is on-peeak and 4) the period of 7 too 8 P.M (from m 65
(15) shows that exxactly (J-1) off sequences shhould be equaal to to 722 TU) is again n Mid-peak period.
p Besidees, the electricity
1 whhen there are J numbers of jobs. Equatioon (16) calculates pricinng of AZ staate in on-, mid-
m and off-ppeak periods are
total setup cost which depennds on sequeence of jobss to 0.2033, 0.11, 0.05 $ per hour [8].
proceeed. Equations (17) to (18) calculate com mpletion timees of
jobs. Equations (19) to (22) dettermine time between
b start and Table 1. Basic Characteristics
comp pletion of eacch job processing. Equatioons (19) and (23) (
calcuulate the enerrgy consumpttion cost of each e job in each
e Te
est
Numberof Number Sizeof Computational
hour of processingg with help of o equations (19)
( to (23). The Prob
blem
Machine
es ofJobs Problem Time(Second)
Noo.
last equation calcculates total energy consuumption costt of
produuction. 1 5 5 25 5.51
2 5 10 50 11.97
3 10 5 50 29.33
4 15 5 75 41.97
4. Nu
umerical expeeriments 5 5 15 100 106.18
6 10 10 100 467.16
To o demonstratee the efficienncy of the prooposed modeel, 9 7 15 10 150 1005.94
8 10 15 200 877.07
numeerical examples are consid dered. These examples
e incllude 9 15 15 300 1014.01
comb bination of thrree different number
n of maachines and joobs.
Basicc characteristiics of these problems
p are indicated
i in taable
1. It should be nooticed that en nergy consumpption of jobs are Teest problems are
a solved by GAMS softw ware on a desk
ktop
distriibuted uniforrmly in the range of 5 to 12 kWh per Core i7, 3.40 GHzz with 16 GB RAM. Compuutational timee for
mach hines. Transpportation costt of jobs annd setup costt of each problem is shown in table t 1. As it is expeccted,
mach hines for each job are unifoormly generateed respectively in compputational tim
me gradually increased wiith problem size
the range of 35 too 45 and 25 to t 40 U.S. doollar per hourr. In increasing. As it mentioned before, the model output is i a
addittion, working time is supp posed to be 122 hours in a day m configuraation plan, indicating arrangement of
system
and time
t unit (TU)) is assumed to
t be 10 minuutes. For instannce, machhines in the syystem, and thhe sequence of
o jobs, whichh is
the period of 8 A.MM to 11 A.M is i defined as 118 time units.
526 S. Ghanei and T. AlGeddawy / Procedia CIRP 57 (2016) 522 – 526

needed to be produced. For instance, in the first problem, optimize the energy consumption reduction in uncertain
figure 2 shows the best arrangement of machines and, figure 3 manufacturing systems considering the volatile demand and
depicts the trend of energy consumption cost and sequence of energy price.
jobs based on the TOU pattern of energy pricing in AZ State
during a day. Figure 3 shows the proposed model schedules References
jobs with lower energy consumption throughout on-peak
period in order to minimize the energy consumption cost. [1] Fang K, Uhan N, Zhao F, Sutherland JW. Glocalized Solutions for
Sustainability in Manufacturing. 2011;
[2] Weinert N, Chiotellis S, Seliger G. Methodology for planning and
Table 2. Objective function operating energy-efficient production systems. CIRP Ann. - Manuf.
Technol. 2011;60:41–4.
SC TEC TC Z [3] Dahmus JB, Gutowski TG. An Environmental Analysis of Machining.
Test Problem No. Int. Mech. Eng. Congr. RD&D Expo. 2004;1–10.
1 312.35 345.77 301.4 959.52 [4] Gutowski T, Murphy C, Allen D, Bauer D, Bras B, Piwonka T, et al.
Environmentally benign manufacturing: Observations from Japan,
2 229.35 124.5 366.27 720.13 Europe and the United States. J. Clean. Prod. 2005;13:1–17.
3 90.07 348.4 644.68 1083.16 [5] Bunse K, Vodicka M, Schönsleben P, Brülhart M, Ernst FO.
Integrating energy efficiency performance in production management
4 204.28 368.3 553.58 1126.17 – gap analysis between industrial needs and scientific literature. J.
Clean. Prod. 2011;19:667–79.
5 338.43 111.91 440.01 877.86 [6] Thollander P, Rohdin P, Moshfegh B, Karlsson M, Söderström M,
6 160.78 167.09 687.46 1015.34 Trygg L. Energy in Swedish industry 2020 – current status, policy
instruments, and policy implications. J. Clean. Prod. 2013 [;51:109–
7 186.68 141.86 693.07 1021.62 17.
[7] Michaloski JL, Shao G, Arinez J, Lyons K, Leong S, Riddick F.
8 119.8 163.22 1030.12 1324.361
Analysis of Sustainable Manufacturing Using Simulation for
9 146.65 111.3 1205.1 1463.06 Integration of Production and Building Service. SimAUD ’11 Proc.
2011 Symp. Simul. Archit. Urban Des. 2011;93–101.
[8] Wang Y, Li L. Time-of-use electricity pricing for industrial customers:
Table 2 reports objective function and three associated A survey of U.S. utilities. Appl. Energy [Elsevier Ltd; 2015;149:89–
costs of energy consumption, sequence-dependent setup of 103.
[9] Choi Y-C, Xirouchakis P. A holistic production planning approach in a
machines and transportation of jobs in nine different test reconfigurable manufacturing system with energy consumption and
problems. It shows the efficiency and practicality of the environmental effects. Int. J. Comput. Integr. Manuf. 2014;28:379–94.
proposed model, since it finds the optimum configuration plan [10] Rajabinasab A, Mansour S. Dynamic flexible job shop scheduling with
alternative process plans: an agent-based approach. Int. J. Adv. Manuf.
and job sequence with minimum objective function in a Technol. 2011;54:1091–107.
reasonable time. Since test problems are in different sizes, all [11] Küster T, Lützenberger M, Freund D, Albayrak S. Distributed
parameters are different and objective functions are not evolutionary optimization for electricity price Responsive
manufacturing using multi-agent system technoloy. Int. J. Adv. Intell.
comparable. Syst. 2013;6:27–40.
[12] Yusta JM, Torres F, Khodr HM. Optimal methodology for a machining
5. Conclusion process scheduling in spot electricity markets. Energy Convers.
Manag. Elsevier Ltd; 2010;51:2647–54.
[13] Luo H, Du B, Huang GQ, Chen H, Li X. Hybrid flow shop scheduling
A new linear mixed integer mathematical model proposed considering machine electricity consumption cost. Int. J. Prod. Econ.
in order to configure the manufacturing system to make it Elsevier; 2013;146:423–39.
more efficient and maximize its sustainability. Machine [14] Tan Y, Guan J, Karimi HR. The Impact of the Subsidy Policy on Total
arrangement and job sequencing are key factors in system Factor Productivity: An Empirical Analysis of China’s Cotton
configuration planning. In this study three main elements are Production. Math. Probl. Eng. 2013;2013:1–8.
[15] Moon J-Y, Shin K, Park J. Optimization of production scheduling with
considered which affect system sustainability in the both time-dependent and machine-dependent electricity cost for industrial
environmental and economic point of view. These three energy efficiency. Int. J. Adv. Manuf. Technol. 2013;68:523–35.
factors are the change pattern in energy prices during a day, [16] Shrouf F, Ordieres-Meré J, García-Sánchez A, Ortega-Mier M.
transportation cost of jobs moving from one machine to Optimizing the production scheduling of a single machine to minimize
another and the setup cost of each machines. Result of total energy consumption costs. J. Clean. Prod. Elsevier Ltd;
2014;67:197–207.
proposed model is a system configuration plan which shows
machines arrangement and job sequences. To get the result
out of the proposed model GAMS CPLEX solver has been
used for nine different problems in size. The new linear mixed
integer model finds the optimum system configuration plan in
a reasonable time. This output depicts the efficiency and
practicality of the proposed model.
Future extension of this work will consider different factors
such as uncertainty in changeable manufacturing systems. In
addition, different parallel machines in changeable
manufacturing system can affect the system configuration
planning to reduce energy consumption. Based on different
energy consumption of machine, different shifting costs and
continues changes in energy price, sustainability of
changeable manufacturing system will be optimized. Another
extension considers the robust optimization algorithm to