# Chapter 8 Applications of Simulation Modeling

LEARNING OBJECTIVES

• To be able to apply simulation to analyze multiple design alternatives in a statistically valid manner.

• To be able to understand the issues in developing and applying simulation to real systems.

• To be able to perform experiments and analysis on practical simulation models.

Chapter 1 presented a set of general steps for problem solving called DEGREE. Those steps were expanded into a methodology for problem solving within the context of simulation. As a reminder, the primary steps for a simulation study can be summarized as follows:

1. Problem Formulation

1. Define the system and the problem

2. Establish performance metrics

3. Build conceptual model

4. Document modeling assumptions

2. Simulation Model Building

1. Model translation

2. Input data modeling

3. Verification and validation

3. Experimental Design and Analysis

1. Preliminary runs

2. Final experiments

3. Analysis of results

4. Evaluate and Iterate

5. Documentation

6. Implementation of results

To some extent, our study of simulation has followed these general steps. Chapter 2 introduced a basic set of modeling questions and approaches that are designed to assist with step 1:

• What is the system? What information is known by the system?

• What are the required performance measures?

• What are the entities? What information must be recorded or remembered for each entity? How should the entities be introduced into the system?

• What are the resources that are used by the entities? Which entities use which resources and how?

• What are the process flows? Sketch the process or make activity flow diagrams.

• Develop pseudo-code for the model.

Answering these questions can be extremely helpful in developing a conceptual understanding of a problem in preparation for the simulation model building steps of the methodology.

Chapter 4 provided the primary modeling constructs to enable the translation of a conceptual model to a simulation model within Arena. In addition, Chapters 6 and 7 delved deeper into a variety of modeling situations to round out the tool set of modeling concepts and constructs that you can bring to bear on a problem.

Since simulation models involve randomness, Appendix B showed how input distributions can be formed and presented the major methods for obtaining random values to be used within a simulation. Since random inputs imply that the outputs from the simulation will also be random, Chapters 3 and 5 showed how to analyze the statistical aspects of simulation. However, there is one topic that we have yet to discuss that is useful when performing realistic simulation studies, comparing multiple alternatives or scenarios. Once you have a practical grasp of how to compare multiple system configurations in a valid manner, you will be all set to analyze realistic models. To prepare you for this topic and the rest of this chapter, we will revisit the LOTR Makers model of Chapter 4 in order to apply these new concepts within a familiar situation. Then, we will apply the techniques with a new modeling context on a larger, more realistic system. Let’s get started with the topic of comparing multiple systems.