Abstract
The MatlabDEVS Tbx has been developed by the research group Computational Engineering and Automation at Hochschule Wismar - University of Applied Sciences, Technology, Business and Design, Germany.
It provides an object oriented DEVS implementation within the Scientific and Technical Computation Environment Matlab®.
This Toolbox is based on the Parallel DEVS (PDEVS) extension of the Discrete EVent System Specification (DEVS) formalism and its associated abstract simulator algorithms introduced by Zeigler in 1976.
It implements a PDEVS simulator with ports and also offers an experimental status of hybrid simulation by allowing the definition of continuous variables within atomic models. Usage requires a general understanding of how DEVS algorithms and DEVS modeling works. The MatlabDEVS Tbx comes with some example models for pure discrete and hybrid scenarios and also QSS-based models.
DEVS-based approaches are relatively unknown in the engineering community. So we hope to contribute in bringing DEVS to engineers by providing it for Matlab®.
The MatalDEVS Toolbox offers the following features:
- Simulator and coordinator classes,
- Root coordinators for simulation of pure discrete or hybrid models,
- Debug modi,
- Templates for user defined atomic and coupled models,
- Modelbase containing reusable atomic and coupled models,
- Examples, including QSS and hybrid models,
- Automatic recording of states in atomic models over time during simulation,
- Methods to check plausibility of models.
The toolbox is not public domain. You can download the latest release (Vers.2.0.) of the MatlabDEVS Toolbox from here.
First, please contact the project supervisor (christina.deatcu[at]hs-wismar.de) for the required download authorisation.
If you use the Toolbox, please cite us with this paper.
Disclaimer of warranty:
The developers assume no liability for accuracy and completeness of the provided information. This refers to damages of substantial or ideational form, that are caused by the usage of the MatlabDEVS Toolbox. By reason that the MatlabDEVS Toolbox is constantly evolving there is no warranty for operability.
Major Changes in PDEVS Tbx for Matlab (Vers. 1.5. -> 2.0.) Major Changes Vers. 1.3. -> 1.5.
- Toolbox is packed as Matlab-App,
- GUI for entrance and introduction,
- Debug modes and internal handling improved
mode 0 : no messages
mode 1 : messages and fct.-calls to user defined output (e.g. std-out (default) or a .txt-file)
mode 2 : same as 1, but stepwise
mode 3 : gauges of x, y and s of atomics- Improved fault tolerant methods for setting ports, states and system parameters.
- User interface changed so that coupled models can be defined as class definitions for reuse,
- Connecting more than one outports to one inport is possible,
- Output and input messages are MATLAB cell arrays now,
- Definition of system parameters (sysparams) for atomic models, which cannot be changed during runtime to distinguish them from states (s),
- Fault tolerant methods for setting ports, states and system parameters introduced,
- States in atomic models can be automatically recorded over time during simulation by setting the observe_flag,
- DEBUG parameter can be set locally for selected model parts (is a property of simulators/coordinators),
- New Check() method for coupled models to verify that all subcomponents and all portnames used in coupling matrices are existing, that component(s) don't define self loops, and that all existing ports are connected (warning, if unconnected ports are detected).
Releases:
Vers.2.0. Sep. 2016
Vers.1.5. Dec. 2015
Vers.1.3. Nov. 2013
First, please contact the project supervisor (christina.deatcu[at]hs-wismar.de) for the required download authorisation.
Note:
The install instructions are based upon a former successful Matlab R2016a (or higher) installation. The software is tested with 32/64 bit Windows 7 and Ubuntu 12.04 but should also work with other operating systems.Unzip the downloaded file and copy it to your harddisk into a specified folder. Add the folder including any subfolders to the path variable within Matlab (Menue: File -> Set Path -> Add with subfolders).
Alternatively install as Matlab's App: goto Tab APPS --> install Apps --> choose MatlabDEVS.mlappinstall --> done ;-) For usage see the following documentation.
The MatlabDEVS Tbx comes with an HTML-documentation for all examples, all basic models and hints for modeling and simulation. You can take a look at the online-documentation here or download as ZIP-file.
Please report bugs to the project supervisor by email (christina.deatcu[at]hs-wismar.de).
Institutions of registered users (in work) of MatlabDEVS Tbx.
Last update: April 2016
- Beijing Institute of Technology, China
- Centro de Automática y Robótica, CSIC - UPM, Madrid, Spain
- CESI Centre Nord-Ouest, Mont Saint Aignan, France
- dwh GmbH, Vienna, Austria
- ETH Zürich, Switzerland
- KAIST, Department of Aerospace Engineering, Korea
- Karlsruhe Institute of Technology, Germany
- King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia
- Northrop Grumman Corporation, California, United States
- Novartis Pharma AG, Basel, Switzerland
- McMaster University, Hamilton, Canada
- TU Wien, Vienna, Austria
- University of Corsica, SPE Laboratory UMR CNRS, France
- University Duisburg-Essen, Germany
- University of South Carolina, Columbia, United States
Relevant publications to the MatlabDEVS Tbx and further research in this field.
2012 C. Deatcu, T. Pawletta: A Qualitative Comparison of Two Hybrid DEVS Approaches. In: Simulation Notes Europe (SNE), Vol. 22(1), ARGESIM/ASIM Pub. TU Vienna, Austria, April 2012, Pages 15-24. (Print ISSN 2305-9974, Online ISSN 2306-0271, doi:10.11128/sne.22.tn.10107)
2011 C. Deatcu, T. Pawletta, T. Schwatinski: Hybride DEVS - Qualitativer Vergleich zweier konträrer Ansätze. In: Proc. of ASIM STS/GMSS Workshop 2011, Krefeld, Germany, February 24-15, 2011, 12 pages
T. Schwatinski, T. Pawletta: A Solution to ARGESIM Benchmark C17 "SIR-type Epidemic" Using a Quantization-Based Approximation and a HPP-LGCA Approach in a DEVS Environment in Matlab. In: Simulation Notes Europe (SNE), Vol. 21(1), ARGESIM/ASIM Pub. TU Vienna, Austria, April 2011, Pages 57-60. (Print ISSN 2305-9974, Online ISSN 2306-0271, doi:10.11128/sne.21.bn17.10053)
2010 O. Hagendorf, T. Pawletta, C. Deatcu, R. Larek: An Approach for Combined Simulation Based Parameter and Structure Optimization Using Evolutionary Algorithms. In: Proc. of 7th EUROSIM Congress on Modelling & Simulation, Prague, Czech Republic, September 6-10, 2010, Volume 2, Full Papers (CD)
T. Schwatinski, T. Pawletta: An Advanced Simulation Approach for Parallel DEVS with Ports. In: Proc. of Spring Simulation Multiconference 2010, Book 4 - Symposium on Theory of Modeling & Simulation -- DEVS, Orlando/Florida, USA, April 11-15, 2010, 132-139
2009 C. Deatcu, T. Pawletta, O. Hagendorf, B. Lampe: Considering Workpieces as Integral Parts of a DEVS Model. In: Proc. of 21st European Modeling & Simulation Symposium, Puerto de la Cruz, Tenerife, Spain, September 23-25, 2009, Volume 1, 27-35
O. Hagendorf, T. Pawletta, C. Deatcu: Extended Dynamic Structure DEVS. In: Proc. of 21st European Modeling & Simulation Symposium, Puerto de la Cruz, Tenerife, Spain, September 23-25, 2009, Volume 1, 36-45
C. Deatcu, T. Pawletta: Towards Dynamic Structure Hybrid DEVS for Scientific and Technical Computing Environments. In: Proc. of Mathematical Modelling MATHMOD 2009 (Argesim Report No. 35), Ed. I. Troch and F. Breitenecker, Vienna, Austria, February 11-13, 2009, ARGESIM-Verlag, Vienna, Austria, 2009, 2716-2719
2006
(cite THIS)T. Pawletta, C. Deatcu, O. Hagendorf, S. Pawletta, G. Colquhoun: DEVS-Based Modeling and Simulation in Scientific and Technical Computing Environments. In: Proc. of DEVS Integrative M&S Symposium (DEVS'06) - Part of the 2006 Spring Simulation Multiconference (SpringSim'06), D. Hamilton (Ed.), Huntsville/AL, USA, April 2-6, 2006, 151-158
2005 O. Hagendorf, T. Pawletta, S. Pawletta, G. Colquhoun: Comparison 16 Restaurant Business Dynamics - A MatlabDEVS based Approach. In: Simulation News Europe (SNE), Issue 44/45, December 2005, page 58 Publisher Eurosim/ARGESIM, Technical University Vienna
2004 T. Pawletta, S. Pawletta: A DEVS-based simulation approach for structure variable hybrid systems using high accuracy integration methods. In: Proc. of CSM2004 - Conference on Conceptual Modeling and Simulation, Part of the Mediterranean Modelling Multiconference (I3M), Genova, Italy, October 28-31 2004, Vol.1, 368-373
2003 C. Deatcu, T. Pawletta: An Object-oriented Solution to ARGESIM Comparison «C6 Emergency Department» with MATLAB-DEVS2. In: Simulation News Europe, 38/39(2003), page 56
C. Deatcu: Development of an Object-Orientated DEVS-Simulators with MATLAB, Diploma Thesis, Wismar, January 2003
2002 Pawletta, T.; Lampe, B.; Pawletta, S.; Drewelow, W.: A DEVS-Based Approach for Modeling and Simulation of Hybrid Variable Structure Systems. In: Modeling, Anlysis, and Design of Hybrid Systems. Engel S., Frehse G., Schnieder E. (Ed.), Lecture Notes in Control and Information Sciences 279, Springer, 2002, 107-129
1999 Schildmann, P.; Pawletta, T.; Lampe, B.; Drewelow, W.: Eine Simulationsmethode für hybride Systeme mit Strukturdynamik und deren Umsetzung. In Proc. 2. Wismarer Automatisierungssymposium, Wismar, 1999