SIE 431 - Simulation Modeling and Analysis

SIE 431 - Simulation Modeling and Analysis (3) Discrete event simulation, model development, statistical design and analysis of simulation experiments, variance reduction, random variate generation, Monte Carlo simulation. 1.5 ES, 1.5 ED, P, 330R, 330L, CR, 321. May be convened with 531.

Law and Kelton, "Simulation Modeling and Analysis", Second Edition, McGraw Hill, 1991.

L.W. Schruben, "Graphical Simulation Modeling and Analysis Using Sigma for Windows", Boyd and Fraser Publishing, 1995.

References: None

Frank Ciarallo, Assistant Professor of Systems and Industrial Engineering

1. Knowledge of probability theory and statistics
2. Introductory programming experience
3. Comfortable with working on a Windows based PC

None

Understand how to take a problem, develop a simulation based model for it, analyze the results, and present the results convincingly.

1. Understand the power and limitations of discrete event simulation.
2. Understand how to formulate an appropriate and correct discrete event simulation model of a system at an appropriate level of detail.
3. Understand how to implement a simulation model in the SIGMA language.
4. Apply standard statistical techniques in analyzing the output data from a simulation program. Students should be able to correctly incorporate ideas such as expected value, variance and independence, in their analysis.
5. Ability to report on the results of a simulation model in the context of the original problem and with an appropriate representation of the uncertainty in the result.

1. Definitions and Examples
2. Monte Carlo Simulation
3. Event Scheduling vs Process Modeling Approaches to Simulation
4. Review of Probability and Statistics
5. Event Based Models using SIGMA: Events, Scheduling Relationships, Managing the events list, Variables, Assignments
6. Running Experiments and Analyzing Output: Collecting Statistics, Plots, Confidence Intervals, Terminating and Steady State Analysis, Multiple Replications, Batching within a Run, Warmup and Run Lengths, Comparing Alternatives Variance Reduction
7. Input Distribution Modeling
8. Random Variate Generation

Class Requirements:

1. Three hours of lecture per week.
2. Four simulation projects for which students are given a problem and must develop a simulation model, generate data using the model, analyze the data and then report on the results in the context of the original problem.
3. Several (4-5) written assignments.
4. One midterm exam and a final exam.

1. Students use SIGMA simulation software to develop the simulation models.
2. Spreadsheets and statistics software are used to analyze data.
3. Spreadsheets and word processing are used in preparation of reports.

Laboratory Projects: None

1. Two exams.
2. Performance on the simulation projects.
3. Performance on the homework.

Contribution to program objectives: Goals 1, 2, 3, 4, 5

Prepared by: Frank Ciarallo   Date: November 26, 1997