This course is designed for systems analysts, operations research analysts, engineers, military planners, computer scientists, and technical managers who would like to use simulation to design and optimize real-world systems. It encompasses a full spectrum of applications, including defense, manufacturing, transportation, process reengineering, call centers, supply chains, computer and communications systems, healthcare, and services. The course presents definitive methods for developing a simulation model, ensuring its validity, choosing simulation software, selecting input probability distributions, analyzing simulation runs, and project management. A case study illustrates the step-by-step application of simulation-modeling techniques. The prerequisite for the course is a basic course in statistics, or the equivalent.
Versions of this seminar have been given for organizations such as AT&T, Boeing, Coca-Cola, GM, IBM, Intel, Lockheed Martin, Los Alamos National Lab, MITRE, NASA, NATO, Northrop Grumman, Sasol Technology (South Africa), U.S. Air Force, U.S. Army, and U.S. Navy.
Each attendee will receive the following:
The book Simulation Modeling and Analysis (
Critical
Questions That the Seminar Will Answer:
What types of problems are ideally suited for simulation?
What is a definitive overall approach for conducting a simulation study? How do you determine the best simulation software for your application? How do you decide on an appropriate level of model detail? What are the proven techniques for ensuring model validity and credibility? How can you correctly model the randomness in your system? How do you choose input probability distributions when no system data exists? How can you determine the correct length of a simulation run? How do you definitively choose a simulation warmup period? What are the 2
2
critical pitfalls in simulation modeling and how can they be avoided?
What You Will Learn:
1.
Designing and Optimizing Systems Via Simulation ModelingNew system design: ensuring that system requirements are met
Existing system modification: analyzing alternative configurations
Advantages of simulation modeling over analytic solutions
Components and logic of a simulation model
Financial benefits of simulation illustrated by examples
10 crucial steps in a sound simulation study
2.
Selecting Simulation SoftwareAdvantages of simulation software versus programming languages
General-purpose simulation packages versus application-oriented packages
Benefits of object-oriented simulation software
Overview of available software
General purpose
Manufacturing
Process reengineering and services
Call centers
Communications/computers
Healthcare
Supply chains
Animation (general purpose)
Live demonstrations of software
3. Building Valid, Credible, and Appropriately Detailed Simulation Models
Determining the level of model detail
Importance of a precise problem formulation
Involving subject-matter experts in model development
Sensitivity analyses
Iteratively increasing model complexity
Techniques for increasing model validity and credibility
Regular interaction with management
Use of a written conceptual model (“assumptions document”)
Structured walk-through of conceptual model before programming
Comparison of model and system outputs for an existing system
Animation
Statistical Techniques for Model Validation
Management’s role in the modeling process
Problem formulation
Control and commitment of resources
Approval of important model assumptions
Methods for enhancing management confidence in the model
Verification, Validation, and
Numerous real-world examples
4.
Modeling Randomness in Real-World SystemsDeciding between fitted theoretical distributions (e.g., exponential or normal) or empirical distributions when system data exist
Using triangular
ExpertFit distribution-fitting software
Computer implementation of input models
Random-number generators
Generating random values from a distribution
Modeling random equipment breakdowns
Modeling arrivals to a system
5.
Reaching Correct Decisions from Simulation Output DataCritical importance of statistics for output-data analysis
Defining experimental parameters
Determining the required number of simulation runs and their length
Specifying warmup-period duration
Estimating desired measures of performance
6. Case Study
A detailed application of simulation techniques
7.
22 Critical Pitfalls in Simulation Modeling and How to Avoid ThemModeling and validation
Simulation software
Modeling the randomness in a system
Design and analysis of simulation experiments
Public Simulation Seminars Home Page