Journal of International Trade, Logistics and Law, Vol. 4, Num.

1, 2018, 101-110

DESIGN OF A GENETIC ALGORITHMIC APPROACH FOR REDUCING

ENERGY CONSUMPTION IN SINGLE MACHINE PRODUCTION SYSTEMS

Elif Tarakçı, (MSc)

Istanbul Commerce University, Turkey

Abdül Halim Zaim, (PhD)

Istanbul Commerce University, Turkey

Oğuzhan Öztaş, (PhD)

Istanbul University, Turkey

Received: April. 26, 2018 Accepted: May. 24, 2018 Published: June 1, 2018

Abstract:
In traditional production systems, production is generally based on performance criteria such as total completion time, price and quality.
Energy consumption or environmental impacts of production systems are not often considered. This article deals with a scheduling problem
aiming to increase energy efficiency, which has been noticed by companies in recent years. A mathematical model is set up to minimize the
total energy consumption in a single machine production system. Genetic algorithm method is used to solve this problem. To prove the
efficiency of our model and algorithm, we use the sample problems in the reference paper. According to the test results, our model and
algorithm have been shown to reduce energy consumption by 50% compared to the algorithm in the reference paper.

Keywords:
Energy Consumption, Single Machine, Genetic Algorithm

1. Introduction
Nowadays, competition for energy sources is becoming more intense than previous decade because of the increase in
energy prices and non-renewable resources. Including production, the industrial sector consumes about half of the
total energy produced in the World (Zhang et al., 2016).
Global climate change, limited energy sources, reduction of dependence on fossil fuels and CO2 emissions require
that energy management in production systems be integrated into decision-making processes (Mikhaylidi et al.,
2015). In addition, changes in customer behavior towards greener products and new environmental regulations
contribute to the key role of the energy management strategy in production organizations (Shrouf et al., 2017).
Briefly, various approaches and solutions for energy efficiency in production systems have been started to be
developed in recent years. Measures have been taken to increase efficiency factor components or to avoid inefficient
components. Generally, energy efficiency studies in the literature can be classified under two categories. The first,
studies are aimed at reducing energy consumption through technological advances in production processes
(Neugebauer et al., 2011). The second is the efforts to reduce energy consumption by adjusting the managerial
parameters of the production process called energy efficient production planning. In comparison to technological
infrastructure investments to increase energy efficiency in manufacturing systems, energy efficient production
planning studies are becoming increasingly popular in practice. Because production planning studies usually do not
require large investments. Especially, there has been an increase in the number of scientific publications in this field
in the last decade (Biel and Glock, 2016).
In this study, it is aimed to develop an efficient mathematical model considering energy efficiency for production
systems. A suitable genetic algorithm is proposed for the energy efficient scheduling model.

101
 Elif Tarakçı & Abdül Halim Zaim & Oğuzhan Öztaş

The outline of this study is as follows. In section 1, literature review is presented. In section 2, a single-objective
problem is proposed in a production system. In addition, an energy efficient model has been developed that
minimizes energy consumption in the single machine scheduling problem. In section 3 inproves genetic algorithm to
solve the model. In section 4, various problems were used to measure the computational performance of the model.
Finally in section 5, the results are discussed.

2. Literature Review
The first systematic approach to scheduling problems was introduced in the mid-1950s. From past to present,
thousands of articles have been published in the literature on different scheduling problems. (Allahverdi et al., 2008).
In the past, operational and planning decisions in production systems were primarily based on traditional measures
such as cost, quality, flexibility. (Vijayaraghavana and Dornfeld, 2010). The research on energy scheduling is limited.
(Zhang et al., 2016).
Mouzon et al. (2007) is based on the principle that energy consumption is performed by idle-running non-bottleneck
machines. For this purpose, they have developed operational methods that minimize the energy consumption of
production equipment. In other words, they aimed to reduce total energy consumption together with other
production planning targets. In addition, they have proposed a multi-objective mathematical programming model to
reduce the energy consumption and total completion time in the single CNC machine. As a result, when there is
non-bottleneck in the machine, turning on/off operation of the machine can be decreased the idle energy
consumption of the machine.
Mouzon and Yildirim (2008) proposed an NP-hard problem to minimize total energy consumption and total
tardiness. They developed a new greedy randomized meta-heuristic to solve the single machine problem. They used
analytic Hierarchical Process to find the feasible solution from non dominated solutions.
Fang et al. (2011) proposed a multi objective mixed-integer mathematical model with flow-shop scheduling problem
that minimizes makespan, carbon emissions and peak total power consumption.
Dai et al. (2013) have developed an energy-efficient scheduling model for flexible flow shop. A genetic-simulated
annealing algorithm is used to obtain the appropriate solution in the model. Experimental results have shown that
there is a conflicting between makespan and energy consumption.
Liu et al. (2014) have developed a model that aims to reduce total electricity consumption and total weighted
tardiness. The study focuses on the classical job shops in the manufacturing industry. Non-dominated sorting genetic
algorithm method is used to solve 10 * 10 job shop scheduling case study.
Lu et al. (2017) have developed a multi-objective scheduling problem of permutation flow shop that minimizes
makespan and total energy consumption. They used a hybrid multi objective backtracking search algorithm to solve
the problem. For this problem, a new energy saving scenario is presented extending the life of the machines.
Backtracking search algorithm is compared is the two well-known NSGA-II and MOEA/D methods. The results
show that the performance of the backtracking search algorithm method in the study was better.

3. Problem Definition and Mathematical Model

3.1. Problem Definition
Generally, when the first product arrives, the machines are opened and not closed until completion of the last
product in the manufacturing industry . If there is a long time between two consecutive jobs' arrival times, the
machine runs idle (Liu et al., 2014). If the time between two consecutive jobs is longer than the break even duration,
closing the machine between the two jobs can provide significant energy savings. Thus, energy consumption can be
reduced by controlling the machine (Mouzon et al., 2007).

Assume that the energy consumption per unit time is PI when the machine stays idle. The energy consumption for
turning on and turning off machine is Eon-off. The time for turning on and turning off machine is Ton-off. TBED
is defined as the minimum time of required for turning on/off machine.
Mathematically,

TBED = max ( Eon-off /PI , Ton-off ). (1)

102
 Design of a Genetic Algorithmic Approach for Reducing Energy Consumption in Single Machine
Production Systems

There are two important key factors in this problem. First, the job scheduling and start times of jobs (or completion
times) should be determined. Second, it is essential to decide whether to turn on and off the machine between two
consecutive jobs.

2.2. Assumptions in The Model

The following assumptions are accepted in the model.
i. Arrival time and processing time of each job are known.
ii. The machine can only process one job at a time.
iii. Preemption is not allowed. When an operation is started, it must be completed without interruption.
iv. Jobs are independent each other.
v. The machine is always available during the job scheduling.
vi. The total turning on/off energy consumption of the machine per unit time is fixed.
vii. The idle energy consumption per unit time is fixed when the machine stays idle.
viii. The objective function is the turning on/off and idle energy consumption of the machine. A job can only
be processed once on the machine.
ix. Setup time of each job includes in processing time.

Parameters

n = The number of jobs

i = The index of job (i= 1,2,…,n)
j = The index of job (j= 1,2,…,n)
Pj = The processing time of job j
Cj = The completion time of job j
rj = The arrival time for job j
PI = The idle energy consumption per unit time when machine stays idle.
Eon-off = The total energy consumption for turning on/off the machine at a time.
Ton-off = The time required for turning on/off the machine
ET = The total energy consumption for producing all jobs
TBED = Minimum time of used for turning on/off machine. (It means that when both the time and the energy
consumption of turning on/off machine is less than the idle time and the energy consumption of the idle machine,
we should turn on/off the machine to consume less time and energy)

Decision variables

Sj = The starting time of job j

Yij = The state of the machine in the time between job i and job j
Xij = Processing state of j after job I

3.3. Mathematical Model

Mathematical model of a single machine scheduling problem that minimizes total energy consumption is presented
the following.

ET = min ( PI ∑𝑛i=1 ∑𝑛j=1≠i (𝑆j − 𝐶i)) (1 − 𝑌ij) 𝑋ij + ∑ni=1 ∑nj=1≠i (Eon − off) Yij Xij ) (2)

0, (Sj - Ci) ≤ TBED (if the machine is idle) (3)

Yij =
 Elif Tarakçı & Abdül Halim Zaim & Oğuzhan Öztaş

1, (Sj - Ci) > TBED (if the machine is turned off)

1, job i and job j are consecutive jobs (4)

Xij =
0, otherwise

Sj ≥ rj ∀𝑗 = 1,2, … , 𝑛 (5)

Sj ≥ Ci ∀𝑗 = 1,2, … , 𝑛 ; ∀𝑖 = 0,1,2, … , 𝑛 ≠ 𝑗 (6)

∑𝑛𝑖=0 𝑋ij = 1 ∀𝑗 = 1,2, … , 𝑛 ≠ 𝑖 (7)

Cj = Sj + Pj ∀𝑗 = 1,2, … , 𝑛 (8)

Sj ≥ 0 ∀𝑗 = 1,2, … , 𝑛 (9)

In this model;
Equation (2) aims to minimize the total energy consumption of the machine. Equation (3) indicates that the
machine runs idle if the time between two consecutive jobs is smaller than TBED, otherwise the machine shuts
down. Equation (4) identifies that the sequence of consecutively processed jobs. Equality (5) implies that a job
can not start to be processed in the machine without coming to the production system. Equation (6) imposes that
the starting time of a job must be equal and greater to completion time of a preceding job. Equation (7) ensures
that each job is processed once on the machine. Equation (8) shows that the completion time of a job is equal to
the sum of the processing time and starting time. Equation (9) indicates that the starting time of each job is
greater than zero.

4. Solution Method: Genetic Algorithm

Genetic algorithms have been extensively researched in the literature (Shrouf et al., 2014). The genetic algorithm
originally proposed by John Holland (1975) in the 1960s and later developed by Goldberg (1989) is the heuristic
search algorithm that simulates the natural selection and evolutionary process. In order to reach successful solution
values, it is important to determine gen structure , chromosomes structure, population size and genetic operators
used in algorithm construction well (Elmas, 2016).

The steps of the genetic algorithm are as follows:

Step 1: Generate initial population of solutions.
Step 2: Calculate the fitness value of each solution in the population.
Step 3: Apply the selection operation (solutions with better fitness values are represented more in the new
population).
Step 4: Perform the crossover operation (two new childs are produced from the two available solutions).
Step 5: Perform the mutation operation
Step 6: Calculate the fitness values of the individuals in the new population.
Step 7: Perform the elitism operation
Step 8: Run the algorithm until the stopping criterion
Step 9: Output the best solution

104
 Design of a Genetic Algorithmic Approach for Reducing Energy Consumption in Single Machine
Production Systems

Chromosome structure, generation of the initial population, selection strategy, genetic operators and stopping
criterion affecting the performance of the genetic algorithm are explained below.

4.1. Coding
Alphanumeric or real digital coding rather than binary coding is prefered in scheduling problems (Elmas, 2016). In
this problem, each job is represented by an integer. Consecutive integers from 1 to N are generated for N jobs. Each
gene identifies a job. Each chromosome contains genes as the number of jobs. Each chromosome in the search
space gives a solution point.

4.2. Generation of The Initial Population

Population is a community of chromosomes. An each chromosome that consists job scheduling represents a
solution point. Under some constraints, some random solutions (chromosomes) is created for the initial population.

4.3. Fitness Function

The fitness function is used to determine the quality of solutions in the current population. The value of the
objective function for each individual (chromosome) in the population is calculated. The calculation value is as
follows:

Function value= Pı ∑ni=1 ∑nj=1≠i (Sj − Ci)) (1 − Yij) Xij + ∑ni=1 ∑nj=1≠i (Eon − off) Yij Xij (10)

Equality (10) represents the total energy consumption of the machine for processing of all jobs. This value consists
of two parts. The first section shows the energy consumption when the machine is idle. The second part calculates
the energy consumption for turning on/off of the machine.

4.4. Selection Operator

The selection operation is a method used for the selection of feasible individuals and the elimination of infeasible
individuals in the population. The selection mechanism ensures the best survival. Chromosomes, which generally
have feasible fitness values, are more likely to be found in future generations. There are many selection operators,
such as tournament method, fitness proportional selection, and local selection (Liu et al., 2014). Stochastic universal
sampling operation is used in this study. Stochastic universal sampling operation is similar to the roulette wheel
selection operation. The most important difference is that the outer part of the circle is divided into equal parts. The
number of these parts equals the number of individuals in the population. When the selection is made, the circle is
rotated only once. Thus, all of new individuals are selected. Not only the selecting of the best individuals but also the
disappearance of the bad ones is prevented with this method. it prevents the algorithm from sticking to a local point.

4.5. Crossover Operator

Crossover is a genetic operator that the displacement of the reciprocal genes of two individuals. There are many
methods such as one-point crossover, two-point crossover, position-based or order-based crossover. In this study,
an order-based crossover operation is used (Elmas, 2016). The steps of the method are shown below.

Step 1: Two individuals are randomly selected from the population.

Step 2: Generate 0 and 1 numbers as the number of genes. The genes are determined for the crossover process
according to more numbers of 0 or 1. These genes are replaced with the genes on the other chromosome
respectively. The same process is applied on the other chromosome.
Step 3. This crossover procedure is applied for all chromosomes.

4.6. Mutation Operator

The mutation operator is the modification of several gene values of individuals in the population. In addition, this
operator makes to increase diversity of the target population and speed up the comprehensive search. Various
mutation operators such as adjacent two-job change, arbitrary two-job change, arbitrary three-job change, Shift
change are used in the genetic algorithm (Murata et al., 1996).
 Elif Tarakçı & Abdül Halim Zaim & Oğuzhan Öztaş

Arbitrary two-job change operators are used in this paper. Two genes are randomly selected according to
determining mutation rate and these genes displace. In this way, new chromosomes are obtained.

4.7. Elitism
Elitism is a strategy for preserving the elite solutions in the evolutionary process. This operator can usually help to
speed up the convergence of genetic algorithms (Zhang and Chiong, 2015). The used elitism strategy is explained
below.

Step 1: Sk is determined as the best solution in the Pk population.

Step 2: Run the algorithm for the Pk population. Sk+1 is determined as the best solution in the solutions of new
Pk+1 population.
Step 3: Sk and Sk+1 solutions are compared. The highest fitness-valued solution is updated in the Pk+1 population.
In other words, the worst fitness-valued chromosome is eliminated from the current population. The best fitness-
valued chromosome is added to the new population. (Rajkumar and Shahabudeen, 2009).

4.8. Stopping Criteria

The stopping criterion is the number of generations specified. When the algorithm reaches this stage, it stops. The
best fitness-valued chromosome in the population is the optimal solution of problem (Yildirim and Mouzon, 2012).

5. Computational Performance
In the literature, Liu et al. (2014) tried to minimize the total carbon dioxide emissions and the total completion time
with First Come First Served rule in a single machine manufacturing system. They achieved some feasible solutions
with NSGA-II method. We used data sets of three-jobs and five-jobs problems in the paper of Liu et al (2014) to
verify our proposed algorithm. We have created job sequences. We evaluated the performance of our model. We
compared solution results of our genetic algorithm and their NSGA-II outcomes. In addition, we considered the
delivery time of all jobs in this paper. Liu et al (2014) used the Ctcf = λ.Etec (λ=0.785 kgCO2/kwh ) formula to
convert energy values to total carbon dioxide emission values. We used same formula to convert total carbon
dioxide emission values to energy values.
Below is a comparative analysis based on the obtained solutions for these problems.

5.1. A Problem With Three Jobs

Jobs have processing times of 10 h, 10 h and 20 h, respectively. The arrival times of the jobs are 0 h, 40 h and 60 h,
respectively. Delivery time of all jobs is 100 h. Kilowatt (kW) is used as a unit of measure for energy consumption.
Processing energy consumption and idle energy consumption are 3 kW and 0.4 kW respectively. The total energy
consumption for turning on/off the machine at a time. is 4 kW hours and the turning on/off on time is 12 hours.
According to the above datas, we find the TBED value in the equation below (Liu et al., 2014).

TBED = max ( Eon-off/ PI , Ton-off )= max (10,12)= 12 (11)

The machine must be switched off when both the on-off energy consumption and the on-off time of the machine
are less than the idle energy consumption and time between the consecutive two jobs. Otherwise, the machine
should not be turned off. According to the data, some feasible solutions were obtained as shown in Table 1.

Table 1. Some feasible solutions of the three-job by different algorithms

GA (Proposed in this paper)

S1 S2 S3 Y12 Y23 ET

106
 Design of a Genetic Algorithmic Approach for Reducing Energy Consumption in Single Machine
Production Systems

Solutions (1) 0 40 60 1 0 8
S2 S1 S3 Y21 Y13 ET
Solutions (2) 40 50 60 0 0 0
S3 S1 S2 Y31 Y12 ET
Solutions (3) 60 80 90 0 0 0
S2 S3 S1 Y23 Y31 ET
Solutions (4) 40 60 80 0 0 4

NSGA-II (Liu et al., 2014)

S1 S2 S3 Y12 Y23 ET
Solutions (1) 10 40 60 0 0 16
S1 S2 S3 Y12 Y23 ET
Solutions (2) 10 40 60 1 0 12
S1 S2 S3 Y12 Y23 ET
Solutions (3) 10 50 60 1 0 8
S1 S2 S3 Y12 Y23 ET
Solutions (4) 40 50 60 0 0 0

Table 1 shows that; Solution (1) of the genetic algorithm and Solution (3) of NSGA-II (Liu et al., 2014) gives same
result value for same job scheduling. This result shows the success of our model and algorithm. In addition, We
achieved zero energy consumption values with different job sequences except First Come First Served rule in
Solution (2) and Solution (3).

5.2. A problem With Five Jobs

The proccessing time, arrival time and the delivery time of each job are given in Table 2. Table 3 shows the genetic
parameters used in this problem. These parameter values are taken from Liu et al. (2014). The other parameter values
are the same with three jobs problem above. This problem is solved with the MATLAB program. Some feasible
solutions are presented in Table 4.

Table 2. Basic input data of each product

Jobs 1 2 3 4 5

Arrival time 0 30 50 80 150

Processing time 10 20 20 30 10
Time for delivery 200 200 200 200 200
Source: Liu et al., 2014
Table 3. The parameter values
Parameter Value
 Elif Tarakçı & Abdül Halim Zaim & Oğuzhan Öztaş

Population size 300

Maximal number of generations 100
Crossover probability 0,98
Mutation probability 0,5
Source: Liu et al., 2014

Table 4. Some feasible solutions of the five-job by different algorithms

GA ( Proposed in this paper)

S3 S2 S1 S4 S5 Y32 Y21 Y14 Y45 ET
Solutions (1) 50 70 90 100 150 0 0 0 1 4
S4 S3 S2 S1 S5 Y43 Y32 Y21 Y15 ET
Solutions (2) 80 110 130 150 160 0 0 0 0 0
S3 S4 S2 S1 S5 Y34 Y42 Y21 Y15 ET
Solutions (3) 50 80 110 130 150 0 0 0 0 8
S1 S2 S3 S4 S5 Y12 Y23 Y34 Y45 ET
Solutions (4) 0 80 100 120 150 1 0 0 0 4

NSGA-II (Liu et al., 2014)

S1 S2 S3 S4 S5 Y12 Y23 Y34 Y45 ET
Solutions (1) 10 30 55 119 165 0 0 1 1 14
S1 S2 S3 S4 S5 Y12 Y23 Y34 Y45 ET
Solutions (2) 10 45 65 100 150 0 0 1 0 26,93
S1 S2 S3 S4 S5 Y12 Y23 Y34 Y45 ET
Solutions (3) 10 66 105 130 162 1 1 0 0 10,8
S1 S2 S3 S4 S5 Y12 Y23 Y34 Y45 ET
Solutions (4) 10 84 129 152 183 1 1 0 0 9,61
S1 S2 S3 S4 S5 Y12 Y23 Y34 Y45 ET
Solutions (5) 10 106 128 154 186 1 0 0 0 8
S1 S2 S3 S4 S5 Y12 Y23 Y34 Y45 ET
Solutions (6) 10 76 137 157 188 1 1 0 0 8,4

In Table 4, we presented some feasible solutions of our algorithm. Solution (4) in our algorithm and Solution (5) in
NSGA-II (Liu et al., 2014) have with same job sequence. The value of solution (4) in our algorithm is better than
solution value Solution (5) in NSGA-II (Liu et al., 2014). We were able to reduce energy consumption by 50%. As a
result, it proves that our model can find much better solutions. Besides, we have created with various job sequences
except First Come First Served rule and found zero energy consumption in Solution (2). The our algorithm
succeeded to achieve a zero energy consumption value by creating an appropriate job sequence with different job
sequences except First Come First Served rule.
108
 Design of a Genetic Algorithmic Approach for Reducing Energy Consumption in Single Machine
Production Systems

6. Conclusion and Future Work

The machines consume waste energy while idling or switching on and off. In single-machine systems, keeping the
machine idle running instead of turning it on and off or vise versa would reduce the waste of energy. Therefore it is
important from the point of energy consumption to decide on when to turn the machine on/off and keep it idle. In
this paper, we have developed a model and algorithm that minimize energy consumption in a single machine
production system. We aimed to reduce energy consumption with decision whether the machine should be idle or
switched on or off between consecutive jobs.
Liu et al. (2014) set up a model that reduces the total carbon dioxide emissions and the total completion time. They
used NSGA-II method to solve three-jobs and five-jobs problems with First Come First Served rule in a single
machine menufacturing system. They obtained some feasible solutions. We used data sets of three-jobs and five-
jobs problems in the paper of Liu et al (2014) to verify our proposed algorithm. We generated a simulation
environment on MATLAB and obtained some feasible results. We have achived job sequences. We evaluated the
performance of our model. We compared solution results of our genetic algorithm and their NSGA-II outcomes. we
considered the delivery time of all jobs in this paper. We used Ctcf = λ.Etec (λ=0.785 kgCO2/kwh ) formula to
convert total carbon dioxide emission values to energy values (Liu et al., 2014). Solutions of the problem with three
jobs was presented in Table 1. Solution (1) of the genetic algorithm has same value and same job scheduling as
Solution (3) of NSGA-II (Liu et al., 2014). It seems that our solutions in three jobs problem were not worse than the
solutions provided by Liu et al. (2014). In addition, We reached zero energy consumption values with different
job sequences except First Come First Served rule in Solution (2) and Solution (3) thanks to our genetic algorithm.
We presented some feaseble solutions for the problem with five jobs in Table 4. Solution (4) in our algorithm and
Solution (5) in NSGA-II (Liu et al., 2014) have same job sequence. While Solution (5) energy value is 8 in NSGA-II
(Liu et al., 2014), solution (4) energy value is 4 in our algorithm. We succeeded to decrease energy consumption rate
50% with our genetic algorithm. Besides, we generated various job scheduling except First Come First Served rule in
this problem. Energy consumption value of Solution (2) was calculated zero. As a result, that proves our genetic
algorithm generating much better solutions than NSGA-II of Liu et al. (2014).
In this article, a scheduling problem with single objective was studied in a single machine system. There are in the
real world that a lot of target need to be optimized together for companies. Therefore, multi-objective mathematical
models can be established in future resarch. In addition, an enegy efficient model can be developed with sequence-
dependent setup times.

References
Allahverdi, A., Ng, C. T., Cheng, T. C. E., Kovalyov, M. Y. (2008), ‘’ A survey of scheduling problems with setup
times or costs’’, European Journal of Operational Research, 187, 985–1032.
Biel, K. and Glock, C. H. (2016), ‘’ Systematic literature review of decision support models for energy efficient
production planning’’, Computers & Industrial Engineering, 101, 243–259.
Dai, M., Tang, D., Giret, A., Salido, M. A. and Li, W. D. (2013), ‘’Energy-Efficient Scheduling for a Flexible Flow
Shop Using an Improved Genetic-Simulated Annealing Algorithm’’, Elsevier, Robotics and Computer-
Integrated Manufacturing, 29(5), 418–429.
Elmas, Ç. (2016), ‘’Yapay Zeka Uygulamaları’’, Seçkin Yayıncılık, Ankara.
Fadi Shrouf, S., Bing Gong, B., Joaquin Ordieres-Mer!e., J. (2017), ‘’ Multi-level awareness of energy used in
production processes‘’, Journal of Cleaner Production, 142, 2570-2585.
Fang, K., Uhan, N., Zhao, F., Sutherland, J. W. (2011), ‘’ A new approach to scheduling in manufacturing for power
consumption and carbon footprint reduction’’, Journal of Manufacturing Systems, 30, 234– 240.

Goldberg, D. E. (1989), ‘’Gnetic Algorithms İn Search, Optimization And Machine Learning’’, Addison Wesley
Publishing Company, USA.
Holland, J., H. (1975), ‘’Adaptation In Natural And Artıfıcıal Systems’’, MIT Press, USA.
 Elif Tarakçı & Abdül Halim Zaim & Oğuzhan Öztaş

Liu, C. G., Yang, J., Lian, J., Li, W. J., Evans, S., Yin, Y. (2014), ‘’ Sustainable performance oriented operational
decision-making of single machine systems with deterministic product arrival time’’, Journal of Cleaner
Production, 85, 318-330.
Liu, Y., Dong, H., Lohse, N., Petroviç, S. and Gindy, N. (2014), ‘’An Investigation into Minimising Total Energy
Consumption and Total Weighted Tardiness in Job Shops’’, Journal of Cleaner Production, 65, 87–96.
Lu, C., Gao, ., Li, X., Pan, Q., Wang, Q. (2017), ‘’ Energy-efficient permutation flow shop scheduling problem using
a hybrid multi-objective backtracking search algorithm‘’, Journal of Cleaner Production, 144, 228-238.
Mouzon, G., Yildirim, M. B., Twomey, J. (2007), ‘’Operational methods for minimization of energy consumption of
manufacturing equipment’’, International Journal of Production Research, 45(18-19), 4247-4271.
Mouzon, G., and Yildirim, M. B. (2008), ‘‘A Framework to Minimize Total Energy Consumption and Total
Tardiness on a Single Machine A Framework to Minimize Total Energy Consumption and Total Tardiness
on a Single Machine’’, Wichita State University, International Journal of Sustainable Engineering.
Murata, T., Ishıbuchı, H., Tanaka, H. (1996), ’’ Genetic Algorithms For Flowshop Scheduling Problems’’,
Computers ind. Engng, 30(4), 1061-1071.
Neugebauer, R., Wabner, M., Rentzsch, H. and Ihlenfeldt, S. (2011), ’’Structure Principles of Energy Efficient
Machine Tools’’, CIRP Journal of Manufacturing Science and Technology, 4(2), 36–147.
Shrouf, F., Mere, J. O., Sanchez, A. G. and Mier, M. O. (2014) ‘‘Optimizing the Production Scheduling of a Single
Machine to Minimize Total Energy Consumption Costs’’, Journal of Cleaner Production, 67, 197–207.
Vijayaraghavan, A., and Dornfeld, D. (2010), ‘’Automated Energy Monitoring of Machine Tools’’, CIRP Annals -
Manufacturing Technology, 59(1), 21–24.
Yildirim, M. B. and Mouzon, G. (2012), ‘‘Single-Machine Sustainable Production Planning to Minimize Total Energy
Consumption and Total Completion Time Using a Multiple Objective Genetic Algorithm’’, IEEE
Transactions on Engineering Management, 59(4), 585–597.
Zhang, R. and Chiong, R. (2015), ‘‘Solving the energy-efficient job shop scheduling problem: a multi-objective
genetic algorithm with enhanced local search for minimizing the total weighted tardiness and total energy
consumption’’, Journal of Cleaner Production, 1-15.
Zhang, Z., Tang R., Peng, T., Tao, L., Jia, S. (2016), ‘’A method for minimizing the energy consumption of
machining system: integration of process planning and scheduling’’, Journal of Cleaner Production 137,
1647-1662.

110