You are on page 1of 10



University of Maribor, Faculty of Organizational Sciences
Kidričeva cesta 55a, 4000 Kranj, Slovenia,,

This paper describes an approach to using simulation and visualization of discrete event oriented
simulation models for multi-criteria scheduling optimization with genetic algorithms. A simulation model
was used for fitness function computation of genetic algorithm results, as well as for visual
representation of process behavior of a chosen schedule following genetic algorithm optimization. In
this way, with the help of simulation and visualization, an additional verification of a schedule’s
suitability was conducted on optimization results. The described methodology provides the planner
with a quick and efficient scheduling method and enables him/her to experiment and decide which of
the suitable solutions will become the production plan. The scheduling system is composed of a
business information system - a database, a discrete event simulation model and a scheduling
algorithm. The purpose of the integrated system is to help operative management personnel with
production scheduling and planning. By comparing various scheduling methods, we established that
the system utilizing genetic algorithms and simulation yielded from 5% to 15% better scheduling within
a shorter time compared to manual scheduling.


Most real life optimization and scheduling problems are too complex to be solved completely. The
complexity of real life problems often exceeds the ability of classic methods (1). In such cases
decision-makers prepare and execute a set of scenarios on the simulation model and hope that at
least one scenario will be good enough to be used as a production plan. A long time goal for
scheduling optimization research has been to find an approach that will lead to qualitative solutions in
a relatively short time.

The development of decision-making methodologies is currently headed in the direction of simulation
and search algorithm integration. This leads to a new approach, which successfully joins simulation
and optimization. The proposed approach supports man-machine interaction in operational planning.
A group of widely known meta-heuristic search algorithms are genetic algorithms (GA). With computer
imitation of simplified and idealized evolution, an individual solution-chromosome represents a
possible solution to our problem. Chromosome fitness is calculated with a fitness function. After being
evaluated with a fitness function, each chromosome in population receives its fitness value. Optimal
fitness values depend on the problem. E.g. for a maximization problem the values should be maximal
and for a minimization problem they should be minimal. The final step of evolution, which transfers the
world of biology to idealized evolution in scheduling problem solving, is data transfer from the
database to the GA computer program. Data are presented to GA in the form of chromosomes.

A problem that many companies encounter is how to schedule production in a way that would
minimize production costs and evenly fill the production line schedule. Traditional methods prove
unsuitable when dealing with a large assortment of products and complex production procedures
since they are too time-consuming. Using GA with discrete event simulation (DES) solves the
complexity problem and enhances production transparency.

Companies need to be flexible to compete for the market share and adapt to market demands by
offering competitive prices and quality. This calls for a wide assortment of products or product types,
punctual deliveries, small production costs, etc. In order to solve assortment, quality, material costs
and similar problems, companies employ dedicated teams. However, we believe that an untapped
source of production cost reduction lies in scheduling and in product launch schedule planning (2). A
good schedule leads to smaller production costs and better personnel satisfaction.

To maintain a large product assortment, a company needs to employ a large number of production
operations and/or tools. Since the complexity of scheduling problems grows factorially with the number
of jobs in a schedule, traditional methods consume too much time to find a suitable solution. Davis (3)
and Goldberg (4) prove that GA is a search method with a good quality/speed ratio. The suitability and
feasibility of GA derived solutions need to be verified against real world production line limitations. In
(5) and (6) simulation is proposed as a verification tool, thus we have employed a DES model to verify
the results produced by GA in our system. Although there is considerable work devoted to the study of
combining GA and simulation methodology, there is a lack of its application in practice, especially in
small and medium-sized companies. The reason probably lies in the lack of suitable methodologies for
knowledge transfer to enterprises, and not in the methodology itself (7).

Simulation is a strong interactive tool that helps decision-makers improve the efficiency of enterprise
actions. The ability of simulation to show a real process on the computer with the consideration of
uncertainty is a big advantage when analysing system behavior in complex situations. In some
situations system behavior can be accepted as optimal, not just in the case of minimal make span, but
also when the risk of a decision does not exceed a defined point, etc. In practice most decision-
makers pay attention to specific limitations (law, social, performance) of the system, which in the
future can lead to unstable system behavior.

This paper presents a production scheduling system using a GA and DES model, resulting from the
further development of integrated management information systems (7). The scheduling system is
composed of a business information system (IS) – a database, a DES model and a scheduling
algorithm. The purpose of the integrated system is to aid operative management personnel in
production scheduling and planning. The scheduling is done on two levels: the first level consists of
approximate annual resource utilization planning using material stocks and delivery time for a known
customer’s data. The second level consists of the operative scheduling of a series of jobs without time
limits and without constraints related to the identity of the customer (e.g. for any customer). In a later
phase, this methodology is also used for raw material cutting in order to reduce raw material
assortment and related material order complexity (8), (9).

The main advantage of the presented system is to enhance man-machine interaction in production
planning, since the computer is able to produce several acceptable schedules using the given data
and a set of criteria. The planner then selects the most suitable schedule and modifies it, if necessary.


The system presented was developed to schedule the production of a mid-size Millboard and
Graphics Company where the type of production system is job-shop. There are four main universal
machines, which are the basis for production scheduling. Currently the scheduling of a product series
is done in two steps. In the first step a crude schedule is produced manually, and in the second step
the production line workgroup leaders alter and finalize the schedule. A consequence of such an
approach is that schedule quality varies significantly from day to day, depending on the planner and
workgroup leader's experience. The DES and GA were used for multi-criteria scheduling optimization
because of transparency and a holistic approach. Scheduling optimization with GA and visual
simulation are just parts of an entire system, which should exist in conjunction with complete
enterprise activity. Figure 1 shows the proposed integrated production scheduling system with the help
of GA and a simulation connected with business IS.

demands data
Products missing quantity
Raw material

Fig. 1: Diagram of data and information streams for scheduling optimization (10)
It consists of order database market demands, the warehouse, raw material and a system for
scheduling. A market with its demand for chosen products or services launches processes in the
enterprise, which represent market demands for a production process. Required numbers of different
products are stored as market demands in the database. The quantity of products to be produced is
calculated according to market demands and the quantity of different products in the warehouse. After
this step, the planner prepares schedules with GA according to the defined criteria. The most suitable
schedule becomes the production plan. Simulation of the real process is conducted in the GA
scheduling process and is used for the fitness function.

The system allows an automated scheduling process and offers superior control of the production
process flow. According to the order of information, GA is used to generate product schedules, while a
simulation model is used for schedule suitability and quality verification. The simulation model
embedded in the scheduling process is described in (11).

A good schedule means good resource utilization while respecting deadlines and production line
limitations. As previously mentioned, a manual schedule is produced weekly and then daily schedules
are modified on the fly. This leads to highly variable resource utilization and miscoordination due to a
time-consuming two-step scheduling process. Following this, it verifies plan feasibility and declares it

In the scheduling process the company uses an IS where the current state of orders, production
procedures and resource availability are stored. The scheduling system utilizes the following order
• setup and production times for different products on the machines in the production process,
• work operations and necessary procedures,
• ordered amount, and
• production deadline.

2.1. Genetic Algorithms

Of all the evolutionary techniques, the GA are best known and widely spread. They have been used
successfully to solve continuous functional optimization problems (12). GA use operators of selection,
crossover and mutation. They combine survival of the fittest in a series of structures with structured,
yet still random, information exchange in the form of search algorithms (13). GA represent solutions to
the problem in structures named chromosomes. In our case a chromosome is represented as a list of
ordered products, whose production should be scheduled according to prescribed criteria and
restrictions in the production process. As stated before, a simulation model was used for the fitness
function, so each chromosome gets a fitness value. The Model of the process for evaluation of
solution quality is often used in complex organizational processes where both randomness of the
process and restrictions should be considered. Thus, it is suitable to be used in production processes
(13), logistics (14) and elsewhere.

Data for scheduling optimization are prepared from the database. A GA program based on extracted
data prepares a starting population of chromosomes. Each population has a limited, fixed size and is
called a generation. With the help of the fitness function, represented with a simulation model,
chromosomes in each population are evaluated. A selection of chromosomes is used to choose which
chromosome from the generation will survive to the next generation. The evolution of the surviving
chromosomes is executed according to genetic operators such as crossover and mutation and
therefore, a new generation of chromosomes evolves. The process of evolution is repeated as long as
the stopping criteria are not satisfied. The last generation of chromosomes represents a number of
suitable plans. Chromosomes with a better fitness value are then simulated on a visual simulation
model. The planner can therefore choose the most appropriate – optimal plan. By using visual
simulation and optimization integration in the system for production planning decision support, it is
possible to decide more easily and quickly, which plan is the most appropriate based on the current

The whole program for scheduling optimization with GA was developed with the C++ programming
language. For validation purposes as well as for fitness function the model was at first implemented in
a ProModel (15), a visual interactive modeling and animation simulation package that facilitates the
building and validation of a model. However, it was soon obvious that the simulation time was too long
for evolution with GA. Due to the excessive length of evolution time, it was decided that fitness
function implementation should be quicker. Therefore, the model was rewritten from ProModel to C++.
In this way, the model was implemented in the same language as GA, and the entire evolution of
solutions became very fast. The ProModel’s simulation model was still used for visual simulation. The
simulation of the model collected data and checked the sequence of operations, which was vital for
setting preparation and production start times for each operation. If the system was complex and if it
needed to be analyzed, we used Petri nets simulation tool ExSpect (16). ExSpect is a powerful
business-modeling tool giving organizations the ability to model and analyze their business processes
effectively and efficiently. With the support of ExSpect, current and future business processes can be
modeled and then subjected to computer analysis. This enables trying out "what-if" scenarios before
introducing them into practice. Visualization and animation of an ExSpect model is supported
automatically. The animation offers planner facilities to follow business processes in a step-by-step
mode as well as continuous simulation.

The process of GA scheduling and use of a simulation model are shown in Figure 2. In the scheduling
process, the missing product quantity represents production demands data as the starting (1
generation sorted according to the arrival time of orders. The chromosome presents one possible
schedule, while the gene presents a customer order. The quality of an individual schedule in each
generation is verified with the fitness function. The fitness function in the presented case is the
minimum of production time of the entire schedule, and is represented with the DES model. It has the
functions of each of the schedule’s fitness evaluations to verify the schedule’s feasibility. Fitness value
selection is done according to the individual schedule. Selection has the function of selecting the fittest
individuals and the elitist selection is used in the presented case.

Selection Last generation Data


New individual
Results of
fitness eval.
Set of production
demands data
Set of feasible

Fig. 2: Genetic algorithms and simulation evaluation in the process of scheduling (17)

Through GA evolution on the selected schedules, new schedules are produced with genetic operators.
We tested different crossover and mutation recombination operators (17), but in the described system
linear order crossover-LOX (Figure 3) and gene swapping mutation (Figure 4) were used, because
they perform the best for the studied problem.

Two random selected crossover positions (4 and 7) in selected chromosomes

Swapping genes with asterisks

Moving genes from crossover space

Selected genes from ch. 2 are moved into selected space in ch. 1 and vice versa
Chromosome 1: * 1 * 3 | 4 5 6 | 7 8 *
Chromosome 2: * 7 8 * | 2 0 9 | 1 * 3
Chromosome 1: 0 1 2 3 | 4 5 6 | 7 8 9
Chromosome 2: 5 7 8 4 | 2 0 9 | 1 6 3
Chromosome 1: 1 3 4 5 | * * * | 6 7 8
Chromosome 2: 7 8 2 0 | * * * | 9 1 3
Chromosome 1: 1 3 4 5 | 2 0 9 | 6 7 8
Chromosome 2: 7 8 2 0 | 4 5 6 | 9 1 3

Fig. 3: Linear order crossover (LOX)

Selected genes in parent chromosome are swapped in child chromosome
Parent: 1 2 3 4 5 6 7 8 9 10
Index: * *
Child: 1 2 6 4 5 3 7 8 9 10

Fig. 4: Gene swapping mutation

The evolution cycle of generating a new population is:
1. A starting population is formed from 60 randomly generated individuals
2. A simulation model is used as a fitness function and validates an individuals' quality – gives a
fitness value to individuals
3. The selected individuals are used for generating new individuals with genetic operators
4. Non-selected individuals are replaced with new individuals made with genetic operators
5. Point 2 repeats if the stopping criteria is not satisfied

During the selection stage:
1. We selected a fixed number of chromosomes from the current population
2. Using genetic operators, evolution occurs in the selected population in order to obtain a new

Newly produced schedules represent the new (n
) generation and are evaluated with the fitness
function. Evolution through GA runs until the stopping criteria is fulfilled. The rate of the result's
progress towards a perceived optimum is used as a stopping criteria. If twenty successive generations
bring no improvement, the evolution is halted. The last generation is a set of feasible schedules. Using
GA, everyone should be aware that there is nearly always a better (closer to optimal) schedule
available, but to find it may take too much time (11). In the evolution, the choice of individual selection
and genetic operators has a major influence on both the run time and resulting solution’s quality. More
details on GA basics as well as using them for scheduling and their operation are described in (3), (4),
and (18) respectively.

For each schedule produced by GA, the simulation model produces resource utilization statistics and
material expenditure. Therefore the presented system, in addition to quality schedule production, also
helps optimize material stocks and workforce placement, helps plan production utilization and
determines due dates.

After completion of the scheduling process, the most suitable production plans are simulated on the
visual model of the system. With process animation, using chosen parameters and according to
defined criteria, the decision-maker is motivated to search for results which will have the most
advantageous influence on the whole production process. The schedule, which is selected after
simulation on the visual simulation model, becomes the production plan and is written into the
database. This production plan will be carried out completely if no urgent demands arise from the
market during the process of production plan execution. In the case of urgent demands from the
market, the scheduling process with visual simulation is repeated. The newly selected production plan
should fulfill urgent demands, as less important tasks are scheduled at a later point in time. Programs
for scheduling and a visual simulation model are integrated into the decision support system, so that
the planner can quickly and efficiently search through possible schedules. Figure 5 shows the process
of scheduling and decision-making for production plan selection.

Scheduling process Decision process
Scheduling data
Result selection -
production plan

Fig. 5: Scheduling and decision process for production plan selection

Simulation and visualization of possible schedules on the simulation model also helps to better
understand the process and in that way contributes to a quicker decision-maker response in the case
of unpredictable events that influence the system.


The experiment was conducted in the case company with real world data for the duration of four
workdays, where each day a new schedule was produced. Using order data, we used the GA and
DES model in ProModel, which hides model complexity behind visual and animated representation to
obtain the final GA generation of possible schedules. The planner therefore does not focus on process
methodology but on searching for the most appropriate plan to fulfill the system requirement, leading
to a better system operation. Visually representing the production process on the simulation model
and graphically representating simulation results further simplify decision-making in the search for an
“optimal” plan. In this way, better cognition of the studied system should lead to the optimal selection
of a plan and the best operation of a production process.

We produced schedules for the four selected successive days and updated them according to the
orders that arrived during the process. One hundred and twenty (120) orders were confirmed. The
number of orders per day varied from 10 to 45. If there were important changes in the production
process, the orders were rescheduled with new data.

Changes that warrant rescheduling are:
• order cancellations,
• new orders,
• machine breakdown, and
• unexpected obstruction in the production process.

If an order is cancelled or a new order arrives, the scheduling can be performed from scratch or from
the existing schedule (with new orders appended and cancelled orders erased). The time needed for
re-scheduling is much shorter than for scheduling from scratch, as has been confirmed through
experiments. A similar problem was the subject of research by Fang et al. (19), with similar findings.

Figure 6 shows the difference between the two rescheduling methods. At first, the GA took about 25
minutes to prepare schedules for the four days. Before the production start, another three orders
arrived and rescheduling was necessary. Starting from scratch with three new orders, approximately
another 25 minutes were needed to produce the new schedule. However, when the GA started using
the existing schedule, the new orders were simply appended and the new schedule was ready in only
seven minutes. Both methods produced the same final schedule.

Fig. 6: Comparison of the times needed and result quality of the two rescheduling methods

A simulation model was used to verify and evaluate the historical schedules, and then new schedules
were produced with GA for the same order data. Two schedules were produced for each series of
orders: one starting from scratch and one from the historical schedule. Table 1 shows results for the
four-day period. The results display that manual scheduling was the most time-consuming (~120 min).
Complete re-scheduling - Re-scheduling
0 5 10 15 20 25 30 40 45 50 Time (min)
Production time (min)
Arrival of new orders
Complete re-scheduling
The shortest method was the rescheduling method with the GA and simulation model (~7 min).
Manual scheduling consistently produced the poorest schedules according to criteria. The presented
approach with the GA and simulation model and rescheduling using the GA and a simulation model
system, produced similar final scheduling results except that the second was less time consuming.

Table 1: Scheduling production methods comparison

Day 1 Day 2 Day 3 Day 4
Scheduling method
t (min) result t (min) result t (min) result t (min) result
Manual (real data) ~120 491,2 ~120 475,6 ~120 512,6 ~120 487,4
GA and sim. model 25 439,7 22 449,4 18 465,2 20 455,3
Rescheduling;GA and
simulation model
7 436,8 5 446,3 4 463,3 4 455,3

In the studied example the type of production process is job-shop, where we have jobs that must be
processed by a set of machines in a specific, predetermined order. It is not necessary that each job be
processed on all machines. Figure 7 shows an optimized schedule of five jobs for a selected day
where some free working time is available. When a new order arrives it is placed at the end of the
schedule. In Figure 8 a non-optimized schedule containing five jobs (1-5) and a job order that has just
arrived (new) are shown. The schedule utilizes two (if necessary) workstations (ID1 and ID2). The
available workstation time of each (8 hours) is represented by triangles. After the new job arrives it is
placed on the non-optimized schedule, and the available time on workstation ID2 is exceeded. Figure
9 shows an optimized schedule for the five scheduled jobs and one new job. The new order can be
accepted, since none of the workstations’ available time is exceeded.

Fig. 7: Optimized schedule for five jobs

Fig. 8: Non-optimized schedule with the new job added on previous schedule – available time is

Fig. 9: Optimized schedule with six jobs (five + new)

After scheduling with GA, the scheduling program shows the most appropriate plans which have the
smallest (minimum) simulated times, so parts of the final solutions (last GA generations) are shown in
Table 2. The first three schedules (plans GA1-GA3) are very much alike in terms of production time.
The fourth schedule (plan GA4) also does not exceed the maximum production time of 480 minutes in
our case. All other plans exceed the maximum allowed schedule time. For comparison with GA
scheduling, a manual schedule for the same data is also shown. As can be seen, the manual schedule
slightly exceeds the maximum available time and the business day must be prolonged. Which
schedule will be used as the final schedule, depends on the planner. She/he is aware of vital
information outside of the computer scheduling system, such as idle times, production and workstation
blocking and wait states in the production process. The actual schedule is produced a day ahead, and
is based on confirmed customer orders.

Table 2: Results of genetic algorithm and manual scheduling

Plan: Sim. Time (min): Products order in each plan (Art no.)
GA1 436,8 5 8 1 3 10 4 7 9 2 6
GA2 438,4 3 5 8 1 7 4 10 9 2 6
GA3 440,5 3 5 8 1 7 4 10 2 9 6
GA4 465,6 4 5 8 1 3 9 6 2 10 7
Manual 491,2 8 3 1 5 7 4 10 2 6 9

The results show a schedule of product order in an individual plan according to the quantity in the
database. Simulated times are the times from the simulated model, built into the optimization program.
Simulation of plan GA1 from Table 2 in the ProModel is shown in Figure 10. A simulation model is
used for animated simulation, which helps planners to determine the properties of each appropriate
plan and helps them to find the best plan (20). After conducting visual simulation on all suitable plans,
planners can also compare the statistical simulation results of each plan.

Fig. 10: Visual simulation of studied process after scheduling optimization

According to the observation of the visualized process and graphical representation of simulation
results, it is obvious that there are not major differences in the process when comparing the different
plans. Regardless of the chosen schedule, the system functions well. In the case of bottlenecks or
workstation over-utilization we chose a schedule that enables the system to operate more smoothly.
Decision-makers select a suitable system for implementation based on schedules and animated visual
simulation results. The schedule selected begins execution immediately and is followed through. In
case of major changes in the system due to workstation failures, blocks or urgent orders, a part of the
schedule is followed through, while a new schedule is developed with the new data. Depending on
needs and demands, other criteria may be used to evaluate quality – e.g. station utilization, total
production time, production costs, etc. Figure 11 shows statistical simulation results (main station
utilization) for plan GA1. Figure 12 shows the comparison between the manually made plan and the
plan made with the help of GA, location utilization for the manual plan according to Table 2.. Locations
are less utilized in the manually made plan which means less probability of bottlenecks and some
reserve in case of increased market demands, but the time needed to finish the plan is longer.

0% 20% 40% 60% 80% 100%
Packing Operation

Fig. 11: Location utilization for plan 1 made with GA

0% 20% 40% 60% 80% 100%

Fig. 12: Location utilization for manually made plan

Using GA and simulation, the planner can produce schedules in less time, while being able to
immediately verify schedules, compare them to existing schedules and possibly improve existing
schedules using their experience and soft information.

The benefits of the presented system are:
• real-time production resource utilization overview,
• aid in quality schedule development,
• greater flexibility (from weekly changes to daily changes),
• reduced impact of human factor on scheduling,
• fast rescheduling in case of order changes, seasonal deviations and machine downtime, and
• easier production tracking.

We can summarize that using GA and simulation yields 5-15% production timesaving, which
translates to less resource utilization and larger production throughput of the existing production


Multi-criteria scheduling methodology with genetic algorithms (GA) and its evaluation with visual
simulation are described. Different schedules – possible plans, generated by GA are tested on the
visual simulation model. A system is connected to the corporate database, so the transfer of the
presented system to the real process is simple. After scheduling optimization, a visual simulation
model makes it possible to visually represent production process behavior according to different

The presented system can be quickly and easily implemented within a modern networked e-business
information system. The basic quality of the described method is quick schedule development and
consistent schedule quality. The main advantage is better schedule quality and resource utilization.
This helps avoid occasional system overloads, while giving a better overview of the current production
process and control over new order insertion to the planner. The model of scheduling algorithm and
simulation coupling, providing visual system representation, shows an actual system operation for the
given schedule. In this way the decision-maker knows what to expect and can avoid possible
problems or at least make precautions. After the implementation of the presented system, the planner
better understands the workings of the production system and can therefore find new ways to improve
the scheduling process. Using visual simulation the planner verifies the influence of stochastic
variables e.g. delivery deadlines, expenses and customer satisfaction on the system or system part
operation. The presented methodology supports the human thinking process, improving the limited
rationality in the decision-making process.

After comparing various scheduling methods we established that combinations of GA and simulation
yield to better production results than manual scheduling, where the scheduling time is included. The
presented rescheduling method using existing schedules proved to be very fast and effective. We
believe that with minor modification to the model and optimization of GA, the time for GA evolution can
be further reduced to about 2 to 3 minutes, which is an acceptable time for interactive business use.

The next phase of integrated system development is reducing material stock, enabling the
development of a material carving optimization module. The module would be connected to order
database and material stock database. Annual simulation of the production would produce valuable
information for major material suppliers and help rationalize the production process and make
scheduling easier, faster and better.


The Ministry of Education, Science and Sport of the Republic of Slovenia supported this research;
program no. PP-0586-501.


(1) Glover F., Kelly J.P. and Laguna M.: New Advances for Wedding Optimization and Simulation,
Proceedings of the 1999 Winter Simulation Conference, 1999.
(2) Ovacik M. I., Uzsoy R.: Decomposition methods for complex factory scheduling problems, Kluwer
Academic Publishers, 1997.
(3) Davis L.: Genetic algorithms and simulated annealing, Pitman, 1987.
(4) Goldberg D. E.: Genetic algorithms in search, optimization & machine learning, Addison Wesley,
(5) Kljajić M., Bernik I., Leskovar R., Škraba A.: Simulation approach to decision assessment in
enterprises, IASTED, Modelling and simulation, USA, 1998, 308-311.
(6) Dijk J. N. et. al.: Visual interactive modelling with SimView for organizational improvement,
Simulation, 67(2), 1996, 106-120.
(7) Kljajić M., Bernik I. and Škraba A.: Simulation approach to decision assessment in enterprises,
Simulation, 75 (4), 2000, 199-210.
(8) Sobotka A.: Inventory control: A simulation and decision support system, Simulation, 141(3),
1998, 141-153.
(9) Gradišar M., Resinovič G., Kljajić M.: A hybrid approach for optimization of one-dimensional
cutting, European Journal of Operational Research, 119(316), 1999, 719-728.
(10) Kljajić M., Breskvar U., Bernik I.: Production Planning Using a Simulation Model and Genetic
Algorithms, IASTED, Modelling and simulation, USA, 2002, 54-58.
(11) Breskvar U., Kljajić M., Bernik I.: Uporaba simulacije za planiranje v embalažni industriji,
Electronical and computer science conference, Portorož, 2000, 243-246.
(12) Azadivar F.: Simulation Optimization Methodologies, Proceedings of the 1999 Winter Simulation
Conference, 1999.
(13) Ronald S.: Routing and Scheduling Problems, Practical Handbook of Genetic Algorithms vol. I,
ed. Lance Chambers, CRC, 1995.
(14) Bruzzone A., Signorile R.: Simulation and Genetic Algorithms for Ship Planning and Shipyard
Layout, Simulation, 71:2, 1998, 74-83.
(15) ProModel: ProModel 4 user's guide, ProModel Corporation, 1997.
(16) Deloitte & Touche Bakkenist, ExSpect User Guide, 1999.
(17) Bernik I., Kljajić M.: Uporaba simulacijskih modelov za večkriterijsko razporejanje z evolutivnim
računanjem, Management and European Union, Portorož, 2002, 839-846.
(18) Syswerda G.: Schedule optimization using genetic algorithms, edited: Davis L., Handbook for
Genetic Algorithms, Van Nostrand Reinhold, 1991.
(19) Fang H.L., Ross P. and Corne D.: A promising genetic algorithm approach to job-shop
scheduling, rescheduling, and open-shop scheduling programs, Proc. of the Fifth International
Conference on Genetic Algorithms, Urbana-Campaign, USA, 1993, 375-382.
(20) Kljajić M., Bernik I., Škraba A. and Leskovar R.: Integral simulation approach to decision
assessment in enterprises, Shaping future with simulation: proceedings of the 4th International
Eurosim 2001 Congress, Delft University of Technology, 2001.