How to Use This Deck

This deck is a checklist, not a summary. Every slide is phrased as something you should be able to state or recognize without looking it up.

  • Question formats: multiple choice, true/false, matching, fill-in-the-blank, and multiple answer (“select all that apply”).
  • The Common Traps slides are built from the actual wrong answers — read them last.

Watch for the select-all questions

Chapter 1 is the one chapter whose bank uses multiple-answer items. Partial knowledge is not enough there: you must be able to rule options out as well as in, so pay attention to the “not a good use” and “not part of this phase” framings.

What This Chapter Covers

What simulation is

  • The textbook’s definition
  • Classifying systems
  • Discrete-event versus continuous
  • Descriptive and prescriptive use

How a study is run

  • The simulation methodology, phase by phase
  • Verification versus validation
  • The DEGREE problem-solving process
  • Professional rules of practice

Objectives

You should be able to:

  • Give the textbook’s definition of computer simulation
  • Classify a system as static/dynamic, deterministic/stochastic, and discrete/continuous/combined — and say which class this book focuses on
  • Order the phases of the simulation methodology and place an activity in the right one
  • Distinguish verification from validation, and name techniques supporting each
  • Expand DEGREE and describe each step
  • Recognize good uses of simulation, and the cases where it is the wrong tool
  • Name the conceptual modeling diagrams and the goals of a simulation study
  • List the professional rules of practice and the KSLCore sub-packages

What Simulation Is

Computer simulation is a numerical technique for conducting experiments on a digital computer involving logical and mathematical relationships that describe the behavior of a system over time.

Three things that definition is not:

  • A physical replica used for hands-on training
  • An analytical closed-form solution to a system of differential equations
  • A statistical regression model fitted to historical data for forecasting

Trap

A simulation is not the real system, and even a well-built one is not equivalent to direct experimentation on the real system.

Classifying Systems

Type Description
Static Time is not a significant factor; the state does not evolve over time
Dynamic The system state changes with respect to time
Deterministic Variables are not governed by underlying random processes
Stochastic Some variables are governed by underlying random processes
Combined continuous-discrete The state vector holds both continuous and discrete variables — e.g. velocity together with number of wheels deployed

This textbook focuses primarily on stochastic, dynamic, discrete systems.

The state of a system is the set of properties or variables that describe it at any time \(t\).

Discrete-Event Simulation

In a discrete-event simulation, observations of the system state are collected only at selected points in time when certain changes take place. Those selected points are called events.

  • Not continuously at every point in simulated time.
  • Not at fixed, equally spaced intervals set by an integration step size.
  • Well suited to systems such as a fast-food service counter, where the state changes at arrivals and service completions.
  • Not the method for the continuous flow of liquids — oil into a tanker — nor does it integrate differential equations over small equal intervals.

Descriptive and Prescriptive

Simulation is a predictive modeling methodology that is both descriptive and — when coupled with optimization or rigorous design evaluation — prescriptive. It is neither purely descriptive nor a pure optimization technique.

In the prescriptive picture, the alternative generator produces different scenarios (inputs) to be fed into the simulation model for evaluation. It does not generate random variates, estimate input distributions, or perform the output analysis.

Goals of a Simulation Study

Goal Meaning
Comparison Compare system alternatives and their performance measures across factors, with respect to objectives
Optimization Find the configuration that optimizes performance subject to constraints
Prediction Predict the behavior of the system at some future point in time
Investigation Learn about and gain insight into the behavior of the system given various inputs

When Simulation Is the Right Tool

Good uses — the select-all answers:

  • Understanding how complex interactions in the system affect performance
  • Understanding how randomness affects performance
  • Training people for events that would be disruptive in the real system
  • Evaluating a proposed system that does not yet exist, to ensure the design will meet specifications
  • Comparing a fixed set of design alternatives against performance goals

Not a good use: a problem that can be solved more directly by a simpler analytical technique or an exact closed-form formula. Re-driving a simulation from one day’s recorded observations is likewise not an appropriate way to simulate a month of operation.

The Simulation Methodology

Phase Representative activities
Problem Formulation Define the system, establish performance metrics, build the conceptual model
Simulation Model Building Model translation, input data modeling, verification, validation
Experimental Design and Analysis Conduct preliminary and final experiments; analyze the results
Evaluate and Iterate Produce model and user documentation; iterate as needed
Implementation Execute the recommended solution and follow up on projected benefits

Verification is not part of problem formulation — it belongs to model building. That distinction is asked directly.

Verification versus Validation

Verification

Determining whether the computer program performs as intended — that the implementation is correct. Debugging the code.

Supported by, among other things, stressing the model via sensitivity analysis.

Validation

Determining whether the model adequately represents the real system.

Supported by comparing simulation output to real-system output.

They are not the same activity and are not interchangeable.

DEGREE

The problem-solving methodology, step by step:

Letter Step
D Define the problem
E Establish measures of performance
G Generate alternatives
R Rank alternatives
E Evaluate and iterate
E Execute and evaluate the solution

Conceptual Modeling Diagrams

The temporary/permanent entity distinction separates the two activity diagrams
Diagram Purpose
Context diagram A pictorial general system description, often with typical flow patterns
Activity flow diagram The process for a temporary entity flowing through the system, and its use of resources
Activity cycle diagram The activities of a permanent entity that remains in the system throughout its life
State chart A software-engineering style diagram of states and transitions

Model conceptualization does not start by writing code — diagrams and narrative descriptions come first.

Professional Rules of Practice

  • Only perform simulation studies that have clearly defined objectives and performance metrics.
  • Never accept results or decisions based on a sample size of one.
  • Use statistically valid methods to determine the sample size for experiments.
  • Take steps to mitigate initialization bias in steady-state simulations.
  • Report results with the sample size, the sample average, and a measure of sample variation.

Trap

“Always assume independence and normality without checking, to simplify the analysis” is the planted wrong answer in the select-all version of this question.

Queueing Models and the KSL

Queueing models are not generally adequate for a total analysis of complex, dynamic systems — a hospital emergency department with non-stationary arrivals and complex resource allocation policies is the textbook’s counterexample.

The KSL supports both the event-view and the process-view of simulation.

KSLCore sub-packages — another select-all item:

  • calendar
  • simulation
  • modeling
  • observers
  • utilities

Terms to Have on the Tip of Your Tongue

Term Definition
Model translation Implementing the model in computer code — timing, general procedures, and the representation of the conceptual model as a program
State The set of properties or variables describing a system at any time \(t\)
Events The selected points in time at which the system state changes
System definition statement A narrative accurately and concisely defining the system, particularly its boundaries

More Terms

Term Definition
Validation Determining whether the model adequately represents the real system
Input data modeling Analyzing and characterizing data to develop probability models for model inputs
Documentation Producing technical and user manuals to support ongoing use of the model

Common Traps

  • A simulation is not the real system and is not equivalent to experimenting on it.
  • Verification = the program works as intended. Validation = the model represents the real system. Never interchange them.
  • Verification is not a problem-formulation activity.
  • Sensitivity analysis supports verification; comparing to real output supports validation.
  • Discrete-event simulation observes at events, not continuously and not at fixed intervals.
  • Re-driving from one day of historical data is not a valid way to simulate a month.
  • Simulation is the wrong tool when a simpler analytical technique will do.
  • Conceptualization starts with diagrams and narrative, not code.
  • Queueing models alone are not adequate for a complex dynamic system.
  • On select-all items, the distractor is usually the one absolute claim — “always assume independence and normality”, “verification and validation are the same”.

Quick Self-Test

Without looking back, answer these:

  1. Give the textbook’s definition of computer simulation in one sentence.
  2. Which three-way classification does this book focus on, and what does each word rule out?
  3. Name the five phases of the simulation methodology, in order.
  4. State the difference between verification and validation, and give one supporting technique for each.
  5. Expand DEGREE.
  6. Name the four conceptual modeling diagrams and say which two are distinguished by temporary versus permanent entities.
  7. List the five professional rules of practice.
  8. Name the five KSLCore sub-packages.
⌂ Index