An Eclipse-Based Graphical Modeling Tool for Discrete Event Simulation.
ABSTRACT
We propose DEVS-Driven Modeling Language (DDML), a graphical notation for DEVS [1] modeling and an Eclipse-based graphical editor, Eclipse-DDML.
DDML attempts to bridge the gap between expert modelers and domain experts making it easy to model systems, and capture the static, dynamic, and functional aspects of a system.
At the same time, it unifies C- DEVS and P-DEVS models. DDML integrates excellent modeling concepts from powerful formalisms and glues them in one unique consistent framework.
Eclipse-DDML provides enhanced graphical editing; further simplifying model construction and promoting good modeling practices.
Integration with eclipse simplifies software development, installation, and updates. This also makes the editor extensible.
TABLE OF CONTENTS
Acknowledgement …….. iii
Abstract ….. iv
List of Figures …….. vii
Chapter 1. Introduction
1.1. Motivation ……… 1
1.2. Objective …. 1
1.3. Structure of the Work …….. 2
Chapter 2. Discrete Event System Specification (DEVS)
2.1. The DEVS Formalism ……… 3
2.1.1. Parallel DEVS …………… 5
2.1.2. DEVS simulation algorithms …….. 6
2.2. Tools for DEVS ………… 8
Chapter 3. DEVS Driven Modelling Language (DDML)
3.1. DDML Processes ……. 11
3.1.1. Relation to Classical DEVS Theory ……….. 12
3.1.2. Relation to Parallel DEVS Theory ………… 13
3.2. DDML States and States Transition …… 13
3.2.1. Relation to Classical DEVS Theory ……… 16
3.2.2. Relation to Parallel DEVS Theory ……. 17
3.3. Modeling a Traffic Light System Using DDML. 17
Chapter 4. Building Graphical Editors with Eclipse
4.1. Overview of Eclipse ………….. 21
4.2. Eclipse Plugin Environment …….. 22
4.3. Eclipse Modeling Framework (EMF) …24
4.4. Graphical Editing Framework (GEF) …… 25
4.5. Graphical Modeling Framework (GMF) …….. 26
4.5.1. Runtime Framework ……… 26
4.5.2. Tooling Framework ……26
4.5.3. Graphical Definition Model ……. 27
4.5.4. Tooling Definition Model ……. 27
4.5.5. Mapping Definition Model …………. 28
4.5.6. Generator Model ………… 28
Chapter 5. Architecture of Eclipse-DDML Graphical Editor
5.1. The DDML Coupled Model Editor …….. 31
5.1.1. Graphical Definition (DDML_Model.gmfgraph) . 31
5.1.2. DDML Tooling Definition ……… 35
5.1.3. DDML Mapping Definition …35
5.1.4. DDML Generator Model …….. 38
5.2. The DDML Atomic Model Editor ….. 39
5.2.1. Atomic Model Graphical Definition …….. 39
5.2.2. Atomic Model Tooling Definition ….. 44
5.2.3. Atomic Model Mapping Definition ……… 44
5.2.4. Atomic Model Generator Model ……. 45
5.3. Creating the DDML UI Plugin ……………. 46
Chapter 6. Using the Eclipse-DDML Graphical Editor
6.1. The DDML Coupled Model Editor ……… 47
6.1.1. The Menu Bar and Tool Bar ……. 47
6.1.2. The Project Explorer ………. 49
6.1.3. The Outline View ……… 49
6.1.4. The Palette …………… 50
6.1.5. The Drawing Workspace ….. 50
6.1.6. The Properties View ………. 52
6.2. The DDML Atomic Model Editor …. 54
Chapter 7. Conclusion .. 56
References … 57
INTRODUCTION
To model is to abstract from reality a description of a dynamic system. Modeling serves as a language for describing systems at some level of abstraction, or additionally, at multiple levels of abstraction. Modeling is a way of thinking and reasoning about systems.
Several methodologies and tools have been developed for modeling of dynamic systems. Most of these tools require the modeler to be an expert in programming and/or mathematics.
Verifying models built with these tools with domain experts is difficult because a wide knowledge gap exists between the domain expert and the expert modeler.
Also, modeling and simulation activities are wide apart. Modeling involves developing multiple levels of abstractions of a system and capturing these abstractions with algorithms that represent the static, dynamic and functional aspects of the system under study.
The primary issues in simulation are timing aspects identification and time management. A generic approach is recommended to integrate advanced modeling into generic simulation methodologies.
Hence, both simulation and software engineering domain expertise should be integrated in the modeling and simulation process. And since there is an underlying simulation operational semantic, there is no need for paradigm/formalism transformation.
In order to realize this solution, an intermediate level of abstraction has to be adopted, which is high enough to be generalized (and accessible to a wide community) and low enough to reduce complexity of code synthesis.
This representation needs to express the structural and behavioral characteristics described by declarative and functional models and this must be inherently coherent.
REFERENCES
Zeigler, B.; Praehofer, H; Kim, T. 2000. ―Theory of Modeling and Simulation‖. 2nd Edition, Academic Press.
Adegoke A. 2010. ―Efficient Object Oriented Implementations for the DEVS Formalism‖. M.Sc. Thesis, Computer Science Stream, African University of Science and Technology, Abuja, Nigeria.
Christen, G., A. Dobniewski, and G. Wainer. 2004. ―Modeling state-based DEVS models CD++‖. Proceedings of MGA, Advanced Simulation Technologies Conference 2004, Arlington, VA, USA.
Kidisyuk, K., and G. Wainer. 2007. ―CD++Modeler: A graphical viewer for DEVS models‖. Technical report SCE-017, Ottawa, ON, Canada.
Matias, B.; G. Wainer; R. Castro; 2010. ―Advanced IDE for Modeling and Simulation of Discrete Event Systems‖. Proceedings of 2010 Symposium on Theory of Modeling and Simulation, DEVS’10. Orlando, FL. 2010.
Nutaro, J. ADEVS. URL: http://www.ornl.gov/~1qn/adevs/index.html. Accessed: November 1, 2010.
Kim, K. H., Y. R. Seong, T. G. Kim, and K. H. Park. 1996. ―Distributed simulation of hierarchical DEVS models: Hierarchical scheduling locally and time warp globally‖. Transactions of the SCS 13 (3): 135–154.
Be the first to comment