QMT634 – SIMULATION

Topic 1: Introduction to Simulation

What is Simulation?

1. Simulation is the process of building a model of a system or a decision problem

and experimenting with the model to obtain insight into the system’s behavior
or to assist in solving the decision problem.

2. The two key elements of this definition are model and experiment.

3. To simulate means that to try to duplicate the features, appearance and

characteristics of a real system.

4. The idea behind simulation is to imitate a real-world situation with a simulation

model. The experiment using simulation model will not affect the real operations.

5. The term simulation model means a model that has been adapted to be analyzed
with the use of simulation.

6. For situations in which a problem does not meet, the assumptions required by
standard analytical modeling approaches, simulation could be the best approach to
modeling and solving problem.

Applications of Simulation

1. Analyzing the operations of:

 Passenger flows at an airport terminal.

 Patient care at an outpatient department.
 Call center operations.
 Product development.
 Staffing in a manufacturing system.
 Work flow in a financial services firm.

2. Application of simulation to business and industrial problems are increasing over the
past 30 years, this is because to the three developments:

i) Fast and easy to use computers, available to most people employed.

ii) Efficient software is available.
iii) Number of students who study simulation has increased.

1
 QMT634 – SIMULATION

Decision and Decision Models

1. Almost any time a decision is made, a model is used to aid the decision maker.

2. Decisions can take many forms.

Example:

Business decisions – to determine ware house capacities, production levels, staff

hiring levels, or other parameters of a production system; selecting or rejecting
certain investment opportunities; determining whether to increase the prime
interest rate in a financial context; selecting a plan to market a new product; or
determining how much money to retain for a self insurance plan.

Engineering decisions – whether to accept a design for a new automobile

suspension, or how many processors to put in a newly designed computer system.

Decision in statistics – to estimate mean cost per year for college education in
Malaysia using a random sample of college students in Malaysia.

* To make this decision, we need to use statistical model.

3. In operations research, we use mathematical models to make decisions, so that

any time a decision is made with the use of mathematical model, the work can be
classified as operations research.

4. Operations research also referred to as quantitative analysis, management

science or decision science.

Important Terms

a) System

 A system is a set of interacting components or entities.

 The components operate together to achieve a common goal or objective.
 When the components work together, the system can perform tasks and
accomplish objectives that individual components cannot.

Example:
Consider the system for stocking and selling goods at a grocery store.
- the arriving customers,
- the checkout counters and attendants,
- a customer service attendant,
- the queues or waiting lines for customers,
- the baskets for carrying groceries, and
- the store with its stock.

2
 QMT634 – SIMULATION

 Other systems?
Hospital – patients, rooms, personnel and equipment
Air traffic control – radar detectors, planes, runways and controllers
Telecommunications – messages, communication network and receiving
stations
Manufacturing – machine, inventories, items produced

 Entities in a system may be physical entities or conceptual entities.

 Most systems in the real world are highly complex, so it is useful to divide the
systems into subsystems.

 A subsystem is a part of a system that performs a specific task for the entire
system.

Example:
A hospital is divided into subsystems – emergency room, surgical suite,
intensive care unit, maternity ward, medical wards, X-ray department and
laboratory.

b) Model

 A model is an abstract and simplified representation of a system.

 Models may be: iconic (look-alike the real entity), analog (explains a
phenomenon by reference to some other occurrence) e.g. thermometer,
schematic (drawing and pictures) e.g. map and mathematical
(mathematical relationship).

 Simulation model include all; iconic, analog, schematic and

mathematical model.
Example: Model of the checkout system

Queue Checkout counter

Customer
arrives Customer
leaves

 Models are prescriptive; they determine the optimal policy (eg. Linear
Programming).
Models are descriptive; they describe relationships and provide information
for evaluation (eg. Queuing Model).

 Models may also be deterministic or probabilistic.

Deterministic model – all input variables are known and have constant
values (not involved chance and risk) (eg. EOQ model). Probabilistic
(Stochastic) model – at least one input variable is uncertain and with value

3
 QMT634 – SIMULATION

that varies. Some data are described by probabilistic distributions (eg.

Monte Carlo Simulation).

 Models may be discrete or continuous. In mathematical programming,

refers to the types of variables in the model. It may also refer to how
model variables change over time.

 Simulation models are descriptive; they simply evaluate measures of

performance or the behavior of a system for a specific set of inputs.

 Simulation models in this course are under probabilistic models. Some

common examples of data that exhibit random behavior are consumer
sales, machine operating times until failure, customer service times and
project activity completion times.

 Simulation models can be either discrete or continuous. For most

systems, events occur at discrete points in time (customer arrivals, start
times of production jobs, departures of ships from ports and so on.
Simulating operations such as an oil refinery, we must incorporate variables
such as temperature, pressure and material flow rates, these occur
continuously over time.

 Simulation models in this course are under discrete simulations.

 Developing models to provide information for making decisions, the term

decision model is frequently used. The model helps in making decisions by
allowing us to understand how the system behaves and by predicting how
the system will respond to various decisions. From various decisions the
decision maker will choose the best.

c) Parameter

 A parameter is defined to be any numerical characteristics of a model or

system. It is a number that describes something about the model.

 There are two types of parameters:

Input parameter – any parameter whose value is required as part of the
model specification.
Example:
- The number of counters or servers.
- The mean inter-arrival time between customers.
- The service time distribution parameters.

Output parameter – any parameter whose value is determined by the

system and its input parameters. Specifies some measure of the system’s
performance.
Example:
- The mean number of customers waiting in the system.
- The mean time spent waiting in the queue.

4
 QMT634 – SIMULATION

- The utilization of checkout counters.

 The relationship between input parameters, the model and output parameters
can be depicted as in the following diagram.

Operating policies
Model Output parameters
Input parameters

 Output parameters are the answers that we want the model to provide to
aid us in decision making.

 We must specify how the model relates the output parameters to the input
parameters. There are two basic ways to do this:

- Use mathematical analysis

- Use simulation

 A mathematical analysis will yield formulas and computational procedure to

produce an exact value of the model’s performance. Compute exact value.

 A simulation, will yield a sample observations that can be used to compute a

confidence interval for the performance measure, and therefore to estimate
the value of the performance measure. Simulation cannot compute the
exact value.

5
 QMT634 – SIMULATION

When Simulations Are Used

1. The early years (1950s-1960s)

 Very expensive, specialized tool to use by large corporations (steel and
aerospace).
 Required big computers, special training.
 Mostly in FORTRAN language.
 Processing cost as high as $600 to $1000 per hour.

2. The formative years (1970s-early 1980s)

 Computers got faster, cheaper.
 Value of simulation more widely recognized, although most of the company quite
large.
 Simulation was seldom considered until there was a disaster.
 Often used to clean up “disasters” in auto, aerospace industries.
• Car plant; heavy demand for certain model.
• Assembly line underperforming.
• Simulated, problem identified.
• But demand had dried up — the result was too late.
 Simulation also found in academic area as a standard part of industrial
engineering and operations research curriculum.

3. The recent past (late 1980s-1990s)

 Introduction to personal computer and animation.
 Wider acceptance across more areas
• Traditional manufacturing applications
• Services
• Health care
• “Business processes”
 Still mostly in large firms.

4. The present
 Began to mature during the 1990s.
 Many smaller firms accept the tool.
 Better animation, greater ease of use, faster computers, easy integration with
other packages and the emergence of simulators have all helped simulation
become a standard tool in many companies.
 Most managers admit that simulation can add value to their enterprise.

Types of Simulation Models

1. Gaming simulation – involve the interaction of one or more persons with the
simulation program (example: video games, war games, flight simulators and etc.).
2. Static simulation (Monte Carlo simulation) – sampling observations and
transforming them according to formulas or rules that compose the model.
Repeated independently many times.
3. Dynamic system simulation – observe the behavior of system models over time
(example: inventory, queuing and etc.).
 Dynamic system simulation – divide by two subareas, continuous simulation and
discrete simulation.

6
 QMT634 – SIMULATION

 Continuous simulation – state variables/attributes change continuously as time

progresses (eg. Model of water flow in a river).
 Discrete simulations – allow state variables/attributes to change only at discrete
points in time. (eg. Manufacturing systems that produce parts).
System quantities change are called events – discrete event simulation

The Simulation Process

1. Develop a conceptual model of the system or problem under study. This step
begins with understanding and defining the problem, identifying the goals and
objectives of the study, determining the important input variables and defining output
measures.

2. Build the simulation model. This includes developing appropriate formulas,

equations, formulating the model design, collecting any necessary data and take into
consideration the constraints involve, determining the probability distributions of
uncertain variables and constructing a format for recording the results.

3. Verify and validate the model.

Verification refers to the process of ensuring that the model is free from logical
errors.

Validation ensures that the model behaves the same as the real system or problem.

4. Design experiments using the model. We should design a complete set of

experiments that we intend to conduct in order to improve the system’s performance.

5. Perform the experiments and analyze the results. Run the appropriate
simulations to obtain the information required to make an informed decision.

Advantages of simulation

1. It is relatively straightforward and flexible.

2. Using software – make simulation model very easy to develop.

3. It can be used to analyze large and complex real world situations that cannot be
solved by conventional quantitative analysis models.

4. Simulation allows what-if? types of questions. This enables the decision maker
evaluate / consider more than one alternative before choosing the best one.

5. Simulation does not interfere with the real-world system. With simulation,
experiments are done with the model, not on the system itself.

6. Simulation allows us to study the interactive effect of several variables to determine

which one is important.

7
 QMT634 – SIMULATION

7. The effects of changes in variables over many months / years can be obtained by
computer simulation in a short time.

Disadvantages of simulation

1. Developing a good simulation model is often a long and complicated process

especially for large and complex problems.

2. Simulation does not give us the optimal solution. It is a trial and error approach that
produces different solutions in repeated runs.

3. The user must generate all of the conditions and constraints for solutions that they
want to examine.

4. Each simulation model is unique. Its solutions and inferences are not usually
transferable to other problems.

Problems

1. For the following systems, identify their subsystems, the input parameters and
output parameters.
a) A restaurant.
b) A university.
c) A bank.
d) A supermarket.
e) A petrol station.

2. Give three examples of simulation applications.

3. Explain the difference between deterministic and stochastics model.