Professional Documents
Culture Documents
Problem Formulation
Every study should begin with a problem statement.
If statement is to be provided by policymakers, it is important that it is understood clearly
by the simulation designer / the analyst.
If statement is to be developed by the analyst, it is important that the policymakers agree
with it.
There are occasions where the problem statement is reformulated as the study progresses.
Steps of simulation design
Model Conceptualization
Requires a wide range of skills:
Ability to abstract out essential features of a problem
Ability to select and modify basic assumptions that characterize the system
Ability to enrich and elaborate the model until you get a useful approximation
General guidelines:
Start with the most simple model and add layers of complexity one by one
Don’t make it too complex; just complex enough that it meets the objectives
Involve the model user – their involvement will enhance the quality of and confidence in the
model
Steps of simulation design
Data collection
The objectives of design indicate the kind of data to be collected
Constant interplay between model conceptualization and data collection
Steps of simulation design
Model translation
Most real-world systems result in models that require a great deal of information storage
and computation.
The model must be translated into a computer program.
You could write the model using a programming language [C++, Java, …] or
You could write the model in a general purpose simulation software [MATLAB, Anylogic,
…] or
You can use special purpose simulation softwares tailormade for each industry.
Steps of simulation design
Verification
Verify if the computer program written in previous step is working correctly
If the input parameters and logical structure of the model has been correctly replicated in
the program, verification is done.
For the most part, common sense is used in this step i.e., no specialized software tools are
usually required to verify.
Steps of simulation design
Validation
Validate the model by determining how accurately it is representing real-world system
Achieved by iteratively calibrating the model until it attains the desired level of accuracy
Steps of simulation design
Experimental Design
The alternatives that are to be simulated must be determined
Usually determined by runs that have been completed and analyzed
Decisions need to be made regarding:
Length of initialization period
Length of simulation runs
Number of replications to be made of each run
Steps of simulation design
Implementation
Success of implementation depends on:
how well the previous steps were carried out
how thoroughly the analyst has involved the model user / decision maker in the entire process