Masters of Business Administration

Session 2023-2024

Assignment 1

Submitted To: Submitted By:

Dr. Vijay Kumar Jain NANDINI GUPTA
HOD 1000019866
 1.Define simulation. What are the major steps
in the simulation process?
Simulation is the representation of reality through the use of models or
other devices, which will react as per reality in given circumstances.
Simulation is used system model featuring features designed to generate
the essence of reality actual operation. According to Donald G. Malcolm, a
simulation model can be defined as a which reflects the functioning of a
large-scale system of humans, machines, materials, and information
working for a period of time in a simulated real-world environment terms.
According to Naylor et al. Simulation is a numerical technique of operation
experiments on a digital computer, involving some form of mathematical
and logical the relationships required to describe the behavior and structure
of a complex real world system over an extended period of time.

There are two types of simulation, they are:

1. Analog simulation: Simulating reality in physical form (eg: children's park,

Constellations, etc.)are known as analog simulations.

2. Computer simulation: For complex managerial decision making

problems, Analog simulation may not apply. In such a situation there is a
complex system it is formulated in a mathematical model for which a
computer program has been developed. Using a high-speed computer
solves the problem. This type of simulation is known in the form of
computer simulation or system simulation.

Steps of Simulation Process:

The simulation model consists of the following components: system units,
input variables, performance measures, and functional relationships. For
example in a simulation model M/M/1 queue, server and queue are system
units, arrival rate and service rate are the input variables, average waiting time
and maximum queue length are performance measures, and
'Time in system = waiting time + service time' is an example of a functional
relationship. Almost everyone simulation software packages provide constructs
to model each of the above components. Modeling is certainly the most
important part of a simulation study. In fact, a simulation study the simulation
is only as good as the model.
Simulation modeling includes the following steps:

Step 1: Identify the problem.

Step 2: Formulate the problem.

Step 3: Collect and process actual system data.

Step 4: Design and Develop a Model.

Step 5: Validate the Model.

Step 6: Document Model for Future Use.

Step 7: Select the appropriate experimental design.

Step 8: Set up experimental conditions for the runs.

Step 9: Perform the simulation run.

Step 10: Interpret and Present Results.

Step 11: Recommend further action.

Although this is a logical sequence of steps in a simulation study, multiple

iterations on different sub-steps may be required before the objectives of a
simulation study can be achieved. Not all steps may be possible and/or
necessary. On the other hand, additional steps may have to be taken exhibited.
We will describe these steps in detail in the next sections.

Step 1: Identify the problem: Enumerate the problems with the current
system. Produce requirements for the proposed system.

Step 2: Formulate the problem: Select a system limitation, problem, or part of

it, study has to be done. Define the overall objective of the study and some
specific issues to be addressed. Performance measurement based on different
systems – define quantitative criteria configurations will be compared and
ranked. At this stage, briefly, identify the configuration of be interested and
formulate hypotheses about system performance. Set a deadline for study, i.e.,
whether the model will be used for a one-time decision (for example, capital
expenditure) or more than once a period of time on a regular basis (for
example, air traffic scheduling). Identify the end user of simulation models, for
example, corporate management versus a production supervisor. There must
be problems be prepared as accurately as possible.

Step 3: Collect and handle the unique framework information: Collect

information on framework characteristics (e.g., bandwidth), input factors for a
communication arrange, as well as performance Current framework. Recognize
sources of haphazardness in the framework, i.e., stochastic input Variables
Select an suitable input likelihood dissemination for each stochastic input
variable and gauge the comparing parameters. Program bundles for
dissemination fitting and choice incorporate Expert Fit, Best Fit, and Add-on to
a few standard measurable bundles. These helps combine goodness-of-fit tests,
for illustration, the χ2 test, Kolmogorov-Smirnov test, and Anderson-Darling
test, and parameter estimation in a user-friendly organize. Standard
dispersions, for illustration, exponential, Poisson, ordinary, hyper exponential,
etc., are simple to demonstrate and recreate. In spite of the fact that most
reenactment program bundles incorporate different conveyances as a standard
highlight, issues related to arbitrary number generators and producing arbitrary
factors from diverse conveyances are significant and ought to be tended to.
Observational disseminations are utilized when standard dispersions are not
suitable or do not fit the accessible framework information. When no
information are accessible the ternary, univariate or typical conveyance is
utilized as a to begin with estimation. See Maria and Zhang (1997) for a nitty
gritty treatment of likelihood disseminations.
Step 4: Plan and Create a Demonstrate: Create the layout and organize graph
of system (how do units stream through the framework?). Decipher these
conceptual models simulation computer program worthy frame. Confirm that
the recreation show executes as desired. Verification methods include shifting
input parameters over follows, their worthy ranges and checking the yield,
substituting constants for irregular factors and physically checking results, and
animation.
Step 5: Approve the Show: Compare show execution beneath known
conditions actual framework execution. Perform measurable theory testing and
infer models the framework was inspected by specialists. Evaluate the believe
set on the show by the conclusion user. If there is any issue, fathom it. For
major recreation considers, experienced experts advocate that structured
introduction of the show by the reenactment examiner to an audience
management and frameworks master. This not as it were guarantees that the
demonstrate assumptions correct, total and reliable, but moreover increments
certainty in the show.
Step 6: Document Model for Future Use: Document objectives, assumptions,
and input variables in detail. A simulation experiment is a test or a series of
tests in which meaningful changes occur built in input variables to a simulation
model so we can observe and identify reasons for changes in performance
measures. Number of experiments in a simulation is greater than or equal to
the number of questions asked about the study model (for example, Is there a
significant difference between the average delay in communication networks A
and B? Which network has the lowest delay: A, B, or C? How will the new
routing algorithm affect performance of network B?). The design of a
simulation experiment involves answering question: what data needs to be
obtained, in what form and how much? Following steps describe the process of
designing a simulation experiment.
Step 7: Select the appropriate experimental design: Choose a performance
measure, some inputs the variables that are likely to affect it, and the levels of
each input variable. When the number of possible configurations (the product
of the number of input variables and the level of each input variables) is large
and the simulation model is complex, common second-order design classes
including central composite, Box-Behnken, and full-factorial should be
considered. Document experimental design.

Step 8: Set up experimental conditions for the runs: Solve the question of
obtaining precision information and the most information from each run.
Determine if the system is stable (performance measurement does not change
over time) or non-stationary (performance measurement changes
over time). Generally, in stable systems, the reaction behavior remains
constant the variable is of interest. Detect if there is a terminating or non-
terminating simulation run suitable. Select run length. Select appropriate
initial conditions (for example, empty and inactive, five customers in queue at
a time 0). If necessary, select the length of the warm-up period. Set number of
independent runs – each run uses a different random number stream and
same initial conditions – by considering the output data sample size. Sample
size should be large sufficient to provide the necessary confidence (at least 3-5
runs for each configuration).
performance measurement estimation. Alternatively, use normal random
numbers to compare alternative configuration using a separate random
number stream for each sample process in one configuration. Identify the
output data most likely to be correlated.

Step 9: Perform the simulation run: Execute the run as per steps 7-8 above.
Most simulation package provides run statistics (mean, standard deviation,
minimum value, maximum value) on performance measurements, for
example, wait time (non-periodic persistent statistics), inventory on hand
(time continuous data). Let the average waiting time in the M/M/1 queue
observed from n runs be N 2 1 W ..., , W , W . It is important to understand that
the average waiting time W is a random the purpose of variable and output
analysis is to estimate and quantify the true mean of W its variability. Despite
the fact that there are no data collection errors in the simulation, the
underlying model is completely known, and replication and configuration are
controlled by the user, the simulation results are difficult to interpret. An
observation may be due to system characteristics or just a coincidental
incident. Generally, statistical inference can assess the significance of any
observed phenomena, but most statistical inference techniques assume
independence, uniformly distributed data. Most types of simulation data are
auto correlated, and therefore, tend to does not satisfy this assumption.
Analysis of simulation output data involves the following steps.

Step 10: Interpret and Present Results: Calculate numerical estimates (e.g.,
mean, confidence intervals) of desired performance measurements for each
configuration of interest. To get confidence interval for the mean of
autocorrelated data, the technique of batch mean can be used. In batch, the
original contiguous data set from one run is replaced with a smaller data set
containing the means of contiguous batches of original observations. This
notion that batch means are independent may not always be true; increasing
the total sample size and increasing the batch length may help. Test
hypotheses about system performance. Construction graphical display of
output data (for example, pie chart, histogram). Document results and
conclusions.

Step 11: Recommend further action: This may include further experiments
increase accuracy and reduce estimator bias to perform sensitivity analysis etc.
2. Why is computer necessary in conducting a real
world simulation?
The purpose of using computers in simulation is to model and predict the
behavior of complex systems such as physical systems, biological systems, and
social systems. Computer simulations can be used to test hypotheses, design
new products, and make predictions about how a system will behave in the
real world.
An example of simulation used in the real world is in the field of weather
forecasting. Weather forecast models use computer simulations to predict the
behavior of the atmosphere. These simulations take into account factors such
as temperature, pressure, wind speed, and precipitation to create a model of
the atmosphere. This model is then used to make predictions about future
weather conditions. These predictions are used by meteorologists to
make forecasts, which can help people prepare for severe weather and make
decisions about things like travel and agriculture.
Another example of simulation being used in the real world is in medicine and
health care. For example, simulation is used to test new drugs and medical
procedures before using them on human patients. They can also help train
medical professionals such as surgeons by providing a realistic, virtual
environment in which to practice.
In short, simulation allows us to test ideas and scenarios in a controlled, virtual
environment, which can provide valuable insights and predictions about how a
system will behave in the real world. They are used in engineering, science,
medicine and many other fields.

Thank You