Presentation is loading. Please wait.

Presentation is loading. Please wait.

McGill University School of Computer Science Ph.D. Candidate in the Modelling, Simulation and Design Lab Eugene Syriani.

Similar presentations


Presentation on theme: "McGill University School of Computer Science Ph.D. Candidate in the Modelling, Simulation and Design Lab Eugene Syriani."— Presentation transcript:

1 McGill University School of Computer Science Ph.D. Candidate in the Modelling, Simulation and Design Lab Eugene Syriani

2 OVERVIEW  Motivation  Language  Semantic Domain  Constructs  Applications  Modelling “everything”  Examples  ¿What’s next?  Scaleability 2

3 MOTIVATION Model Transformation  Controlled Graph Rewriting 3

4 SEMANTIC DOMAIN The Discrete Event System Specification [1] (DEVS) formalism Highly compositional simulation framework DEVS: – Blocks – Ports – Events Semantics: Parallel composition of blocks 4 [1] Zeigler B.P., Multifacetted Modelling and Discrete Event Simulation/ Academic Press, London (1984)

5 DEVS META-MODEL 5

6 S t  (s) (s,0) s s'' s' Y t tyty y1y1 (s)  int (s) X x1x1 txtx t txtx (s,e)  ext ((s,e),x) (s 0,0) s0s0 0 ATOMIC DEVS

7 COUPLED DEVS 7 C1 A1 A2 C2

8 MoTiF MAPPED ONTO DEVS [2] Blocks – Atomic: encapsulate a graph rewriting rule – Coupled: encapsulate a set of rules, scoping Events – Inport: receive the host graph – Outport(s): send the transformed graph 8 [2] Syriani E. and Vangheluwe H.: Programmed Graph Rewriting with DEVS. AGTIVE 2007, LNCS (2008)

9 MOTIF META-MODEL 9

10 LANGUAGE CONSTRUCTS AtomicRule (ARule): transformation rule application ForAllRule (FRule): apply rule on all matches (parallel independent) StarRule (SRule): apply rule as long as possible TransactionalRule (XRule): control backtracking enabled Selector: At most 1 rule is applied Synchronizer: synchronize (merge) “threads” of rule applications CoupledRule (CRule): parallel composition of inner models 10

11 EXAMPLES (1) 11

12 EXAMPLES (2) (Selection Pattern) Back-tracking 12

13 EXAMPLES (3) Synchronization Pattern 13

14 WORKFLOW 14

15 MODEL THE TRANSFORMATION SYSTEM [3] 15 APPLICATIONS: [3] Syriani E. and Vangheluwe H.: Programmed Graph Rewriting with Time for Simulation-based Design. ICMT 2007, LNCS 5063, pp (2008)

16 MODELLING & SIMULATION-BASED DESIGN [3] 16 APPLICATIONS: Finding the optimal game speed [3] Syriani E. and Vangheluwe H.: Programmed Graph Rewriting with Time for Simulation-based Design. ICMT 2007, LNCS 5063, pp (2008)

17 REAL-TIME TRANSFORMATIONS (GAMES) 17 APPLICATIONS:

18 ANTWORLD SIMULATION CASE STUDY [4] 18 APPLICATIONS: [4] Eugene Syriani and Hans Vangheluwe. Using MoTif for the AntWorld Simulation Tool Contest. In: Arend Rensik and Pieter Van Ghorp (eds.) Workshop on Graph-Based Tools (GraBaTS 2008). Best Live Solution (GUI). Leicester (2008).

19 CD 2 RDBMS CASE STUDY 19 APPLICATIONS:

20 ESSENTIAL PRIMITIVE PROPERTIES 20

21 WHAT’S NEXT: SCALEABILITY Higher-order transformations 21

22 WHAT’S NEXT: SCALEABILITY Declarative bi-directional transformations 22

23 WHAT’S NEXT: SCALEABILITY Efficiency 23 Matching Implementation Distributed

24 REFERENCES Upcoming SoSym article Eugene Syriani and Hans Vangheluwe. Using MoTif for the AntWorld Simulation Tool Contest. In: Arend Rensik and Pieter Van Ghorp (eds.) Workshop on Graph-Based Tools (GraBaTS 2008). Best Live Solution (GUI). Leicester (2008). Eugene Syriani and Hans Vangheluwe. Programmed Graph Rewriting with Time for Simulation-Based Design. In Alfonso Pierantonio, Antonio Vallecillo, Jean Bézivin, Jeff Gray (eds.): ICMT LNCS, vol. 5063, pp Springer-Verlag, Zürich (2008). Eugene Syriani. Programmed Graph Rewriting: MoTif, Presentation for COMP 763: Modelling and Simulation Based Design (2008). Eugene Syriani and Hans Vangheluwe: Programmed Graph Rewriting with DEVS. In A. Schürr, M. Nagl, and A. Zündorf (Eds.): AGTIVE LNCS, vol.5088, pp Springer-Verlag, Kassel (2008). 24


Download ppt "McGill University School of Computer Science Ph.D. Candidate in the Modelling, Simulation and Design Lab Eugene Syriani."

Similar presentations


Ads by Google