DISCRETE SIMULATION

Discrete models are mainly concerned with the study of queues with the
objective of determining measures such as the average waiting time and the length of the queue. These
measurements change only when a customer enters or leaves the system. The moments when they occur
Changes at specific discrete points in time (arrival and departure events) origin the name
discrete event simulation. (TAHA, 2012)

ELEMENTS OF DISCRETE EVENT SIMULATION

The final objective of the simulation is to estimate some desirable performance measures that
describe the behavior of the simulated system. For example, in a service installation, the
associated performance measures can include the average wait time until a customer
is attended, the average queue length and the average utilization of the facilities
service. (TAHA, 2012)

MODEL COMPONENTS
Simulation clock
The time marked by the simulation clock makes the passage of time explicit throughout the
model. This makes it possible to indicate the beginning and end of the simulation, and the occurrence of events.
clinical moments precisely, without the need for fixed-duration cycles. This allows for a
more efficient and realistic simulation based on the occurrence of events.
Entities
They are the dynamic elements that are simulated through the model, change status, affect and are
affected by other entities and are the protagonists of the clinical events studied in the model.
Attributes
They are the characteristics of the entities, with which they are described and differentiated. For example, they are
attributes of the patients, such as age, sex, the duration of their illness, the severity of it
and the fulfillment of the treatment. An important attribute in economic evaluations,
fundamental in cost-utility studies, is quality of life. This can be studied from
detailed form including it through an attribute.
Events
It is everything that can happen during the simulation based on the process we are in.
studying. For example, an event can be an adverse effect, a hospital admission, discharge
of the hospital, a change in dosage or a sick leave. The concept of event goes beyond the
transitions of Markov models, since the occurrence of an event does not necessarily imply
the change in the patient's health status (e.g., a doctor's visit).
Despite the fact that there is a given timeline, events can happen almost
simultaneously and they can do it in any logical sequence related to the plausible history of
study process (e.g., failure in a target organ or multiple organ failure in a patient with sepsis).
Unlike in Markov models, these events do not have any type of restriction.
by heart.
The risks of different events occurring can take the form of dependent functions.
from the data and depend on some attributes and variables. These functions can change during the
simulation.
(José Manuel Rodríguez Barriosa, 2008).

MONTE CARLO SIMULATION

The Monte Carlo method provides approximate solutions to a wide variety of problems.
mathematicians enabling the conducting of experiments with number samplings
pseudo-random in a computer. The method is applicable to any type of problem, whether it is
stochastic or deterministic. Unlike numerical methods that are based on evaluations in
N points in an M-dimensional space to produce an approximate solution, the Monte Carlo method
Carlo has an absolute error of the estimate that decreases as per the limit theorem.
central (Huizar Perez, 2009).
The type of simulation used is sequential, since the times of exit and restoration
generated are accumulated to obtain the total operating time of the primary circuit under
study. Simulation is an iterative process in which observations are made over a period of time.
interest, the operational states that appear in the primary circuit due to random events
exit and restoration of the components. In each operating state, the points are determined
load affected by the failure of a given component. Once the simulation is complete, it
count for each charging point the number of outputs that affected it and the time of
unavailability. With these two basic indices, the other reliability indices are calculated. For
to apply this method, it is necessary to know the probability distributions that model the times
for the output and restoration of each of the components.
The simulation is implemented as software that uses the system's database, where it has been
registered for each primary circuit: components with their probability distributions, points
of charging, number of users per charging point and total demand per charging point. A matter
the key in building the database of the system is the encoding method of the
components of each primary circuit, in such a way that it allows to know which load points
are affected by the removal of any of the components.
This is the most versatile analysis method given that:
• Allows the use of any distribution to model the times for departure and restoration of the
components.
• It allows solving systems for which there is no analytical solution. For example, systems
where some of the components have modeled time for output or restoration through
the Gaussian distribution.
It allows obtaining the probability distributions of the reliability indices of the points of
load, which is very useful for assessing the risk of different values of the indices occurring.
Changes in the system are made in the database without the need to make changes in
the software (Carlos Zapata, 2004).
Monte Carlo simulation is a quantitative technique that uses statistics and the
computers to imitate, using mathematical models, the random behavior of systems
non-dynamic reals (generally, when it comes to systems whose state is changing with the
As time goes by, one resorts to either discrete event simulation or simulation of
continuous systems). The Monte Carlo method is a non-deterministic or statistical method.
numerical used to approximate complex and costly mathematical expressions to evaluate with
accuracy.
The Monte Carlo method provides approximate solutions to a wide variety of problems.
mathematicians enabling the conduct of experiments with number sampling
pseudo-random numbers on a computer. The method is applicable to any type of problem, whether
stochastic or deterministic. Unlike numerical methods that are based on evaluations in
N points in an M-dimensional space to produce an approximate solution, the Monte method
Carlo has an absolute error of the estimation that decreases as per the limit theorem.
central.
Key of the Monte Carlo method.

Identify the A is taking place After repeating in several

Crear un modelo
mathematician of the variables whose experiment sometimes this experiment,
behavior consists of generating we will have
system that is
random determines the random samples observations on the
wants to analyze
behavior stop the variables behavior of the
global ofm
[Link] system

Our analysis will be more precise the larger the number n

of experiments that we will carry out.
APPLICATIONS OF DISCRETE SIMULATION IN THE
ORGANIZATIONS
Discrete simulation application for a military logistics system based on cases
historical accounts of World War II.

In this research work, discrete simulation is used for the proposal of

computational models that provide support for various military logistics systems
under a series of scenarios. The discrete simulation will reduce the high uncertainty
existing in various complex environments that may arise in military contexts,
be it disaster, catastrophe, humanitarian aid, or peace missions. Already since the time of
Napoleon considered military logistics as the third fundamental pillar of the art of
the war, complemented by strategy and tactics.
According to experts, the final outcome of the battle could be explained by seventy percent.
for the performance of logistics systems. (Soto, 2009).

An application of discrete simulation to the improvement of load capacities

in a warehouse.

In a Central Warehouse, a non-compliance has been evidenced regarding the order cycle.
delivery to the clients of Havana has prompted complaints of 44.4%
They.
Associated with this issue, underutilization of capacity has been detected as a deficiency.
of load, since it is used at 66%, impacting the level of service
perceived by customers regarding the order delivery cycle that is
20.415%, considered very low.

This paper presents the results of the diagnosis carried out and a proposal.
of increasing the utilization of load capacity, which is evaluated based on a
simulation model using the ARENA simulation package, which allowed
evaluate the proposed plan and determine that this indicator will be increased to 97%,
existing a relationship between the availability of resources associated with distribution
with
Relation to the demand of 134.02%. The expected economic impact is shown that
it rationalizes resources such as fuel and maintenance accessories. (Ileana Pérez)
Vergara, 2010).

design of the operation of a tourist information call center, through

discrete simulation. case of application in Guatemala.

As society and economies have evolved, a development has taken place.

growing preference for the service culture, both in the demand for these, as
in the labor field that has become the most important source of jobs. Evidence of
This is the rise that Call Centers have had, which emerge as an effective response,
to timely address the different ways in which they may arise
customer needs.
 This study aims to show how Simulation can be useful for sizing the
operational capacity -initial configuration- of a Call Center that seeks to provide
tourist information in Guatemala, especially when this service does not exist and does not
it has system data (Quijada, 2008).

Discrete event simulation of the logistics chain for commodity export.

This work uses discrete event simulation technique as methodology

support for the decision in a complex problem with multiple variables; the goal is
model the logistical operations involved in the maritime export of commodities
in bulk and bulk load presentations from twelve producing companies until the
seaport. The simulation allows analyzing the sensitivity of the main
parameters and variables of the system, in order to define different planning models
of dispatches. The scenarios show that it is possible to achieve a reduction in the
inventories and maintenance costs, as well as improve the indicators of
service and as a consequence increase the rate of resource utilization (Guerrero,
2012).

Cost Prediction Model in Health Services Supported by Simulation

Discrete.

A decision support model has been developed for cost analysis.

provision of health services. Discrete simulation is used, evaluating different
scenarios of service provision by a Health Service Provider Entity
EPS: general medical consultation, emergencies, hospitalization, and catastrophic illnesses,
among others, The model was created for a typical EPS to simulate the cost of service provision
services for one year, based on indicators taken from health statistics of the order
national whose values were adjusted to the specific behavior of the EPS. The
results obtained with the model allowed us to determine that the calculation approach
cost is effective, and that the national averages of certain parameters must be
modified for each case. Otherwise, significant errors can occur in the
budget not including the variation in these parameters (Vargas, 2013).

Improvement strategies in the waiting line system at Bancamía S.A. using

discrete simulation.
Financial entities as service providers identify that the way
offering and processing their service portfolio has become increasingly important in recent
times, due to the increasing competition that offers a comprehensive and varied portfolio that
In addition to the quality of the service, they integrate factors of efficiency and effectiveness in their system.
corporate. In addition to the above, these entities have opted to structure three
key components of your service offering: Products, infrastructure, and customer service
client, in this sense the level of satisfaction with the service by the clients and the
The achievement of organizational goals is an indicator of responsiveness.
of the entity in economic, organizational, and sectoral terms. The present
research shows the application of queueing theory in a financial entity, to
which is supplemented by a computational tool with the aim of modeling the
system behavior and propose improvement strategies that meet the
system requirements and specifications (Chingaté Ávila, 2012).
 SIMULATION SOFTWARE AND SIMULATORS MORE
USED

GASP IV
It is a collection of FORTRAN subroutines, designed to facilitate the simulation of sequence of
events. About 30 subroutines and functions that provide numerous facilities, including:
Time advancement routines.
Management of future event list.
Addition and removal of entities.
Collection of statistics.
Random variable generator.
Standard report.
The programmer only provides a main program, an update routine, routines for
events, custom report generators and a subroutine called EVNTS. The program
main must include the statement CALL GASP; being GASP a subroutine that determines the eminence
event, invoking to EVNTS written by the user and obtaining the NEXT index.
GASP IV is a simulation language developed by Alan B. Pristker and N. Hurst in 1973. It is a
hybrid language because it can be used for discrete, continuous, and simulation programmers
combined; being the first to fully integrate these two functional environments of the
time. GASP IV is a derivative of GASP II, and it differs in the definition of the space-event.
state (state space event).
SIMSCRIPT II.5
Developed at the RAND Corporation by H. Markowtz in the early sixties. SIMSCRIPT
11.5. It is a simulation language with a focus on events and processes; it is hybrid because it has
facilities for discrete and continuous systems simulation. A SIMSCRIPT 11.5 programmer
consists of the following parts:
Preamble.
Main program.
Event routines.
Ordinary routines.
SIMSCRIPT 11.5, produced by CACI Products Company (La Jolla, California), was used in the
based on large and complex simulations, as is the case with queue-unaware models;
for example military combat models. It is available in PC version highlighting its
The [Link] 11.5 environment is based on entities, attributes, and sets.
Visualize the world to be simulated as a set of entities that can be described through
of its attributes and the events that appear over time.
SIMAN/Cinema
The original version of SIMAN (Simulation and Analysis) was developed by Dennis Pegden in
the University of Alabama, when it was the leader of the development group for the original version of
SLAM (based on the GASP and Q~GER-r software from Pristker and Associates). Later,
Pegden begins his work at Penn State University where he designs it as a language of
modeling for general purposes, including very useful manufacturing facilities in
modeling of complex manufacturing systems.
Since its initial implementation in 1984, it has been continuously refined by System Modeling
Corporation, y en 1998 y 1989 el lenguaje fue completamente rediseñado dando origen a
SIMAN/Cinema.
The modeling environment in SIMAN is developed between Modeling and Experiment; in the
first, the components of the system and their interactions are described, and in the second, the definitions are made
experiment conditions (length of the run, initial conditions).
SIMAN models a discrete system using process orientation; that is, in a model of
particular system, the entities that move through the system are studied. An entity for
SIMAN is a client, an object that moves in the simulation and has unique characteristics.
known as attributes. The processes denote the sequence of operations or activities through
of which the entities move, being modeled by the block diagram.
You build a block diagram in a graphical flowchart, selecting and combining
blocks. Then, interactively, using a special editor, the automatic generator is activated.
the statements of the model from the graphical environment. The SIMAN blocks are classified into 10
basic types.
SLAM II
SLAM II (Simulation Language for Alternative Modeling) is a simulation language through which
models can be built with a process or event orientation. SLAM was developed in
1979 by Dennis Pedge and Alan Pritsker and is distributed by Pritsker Corporation (Indianapolis,
Indiana). The part of SLAM that focuses on processes employs a network structure composed of
through symbols of nodes and branches such as queues, servers, and decision points. Modeling
it means incorporating those symbols into a network model that represents the system and where the
entities (items) pass through the network. SLAM contains a processor that converts the
visual representation of the system to a set of sentences.
ARENA
Arena software is a modeling and simulation tool for improvement.
systems and processes. It is used to analyze business operations, services, processes of
manufacturing and flow systems in general.
(Leirca, 1999)
REFERENCES

Ileana Pérez Vergara, R. G. (November 29, 2010). AN APPLICATION OF

THE SIMULATION. Obtained from AN APPLICATION OF THE SIMULATION:
[Link]
José Manuel Rodríguez Barriosa, D. S. (April 1, 2008). The models of
discrete event simulation in economic evaluation of
technologies and health products. Barcelona, Spain.
Soto, D. A. (April 20, 2009). "APPLICATION OF DISCRETE SIMULATION FOR
A SYSTEM. Conception, Chile: University of Bio Bio.
TAHA, H. A. (2012).INVESTIGACION DE [Link]: PEARSON.
WINSTON, W.L. (2005). OPERATIONS RESEARCH APPLICATION
[Link]: THOMSON.
Carlos Zapata, L. C. (May 24, 2004). THE SIMULATION METHOD
MONTECARLO. Pereira, Colombia .
Chingaté Ávila, V.M. (2012). Improvement strategies in the system of
waiting lines at Bancamía S.A. using discrete simulation.
Bogota, Colombia.
Guerrero, M. A. (2012). Discrete event simulation of the logistics chain
export of commodities. Chilean Journal of Engineering, 257-258-
259-260-261-262. Obtained from [Link]
pid=S0718-33052014000200011&script=sci_arttext&tlng=en.
Huizar perez, M.T.(14 de Diciembre de 2009). Metodos de montecarlo.
Tijuana, Mexico.
Quijada, J. (2008 ). DISEÑO DE LA OPERACION DE UN CALL CENTER DE
TOURIST INFORMATION, THROUGH DISCRETE SIMULATION. CASE
APPLIED GUATEMALA. Mexico City, Mexico.
Vargas, J. M. (December 16, 2013). Scielo. Obtained from Scielo:
[Link]
07642014000400019&script=sci_arttext
[Link]
[Link]
Science Direct
Virtual pro
Engineering plus
 INTRODUCTION
The growing capacity of computers and the immense research in
the field of Computer Science provides new tools for
support the decision-making process in various disciplines and areas of
design and management of the industry. Simulation is one of the most important tools
important and more interdisciplinary. In short, we can say that the
simulation is performed when the computer pretends to be a store, an airplane or a
grocery market. The user defines the structure of the system they want
similar. A run of the corresponding simulation program tells you which
it will be the dynamic behavior of your company or of the machine that is
designing. This way we can see the forecasts for demand and utility of
our product, or to see when a mechanism might fail under the conditions
adverse conditions of the environment where it will operate.

In highly developed countries, simulation is a tool.

principal in decision-making processes, in business management
and the production planning. Additionally, Simulation is increasingly
"user-friendly" that does not require being a specialist in
computing.
 Discrete Simulation

Laura Rojas
Carlos Rebolledo Ortega

Queueing theory
Ileana Pérez

Santiago de Cali
San Buenaventura University
2015