PharoDEVS

An Pharo based implementation of Discrete Event System Specification - formalism for modelling and simulation by B. Ziegler.

Loading project

Install Pharo latest stable version e.g. from command line:
wget -O- https://get.pharo.org/64 | bash

Run in Pharo Playground:

Metacello new
 baseline: 'PharoDEVS';
 repository: 'github://ApptiveGrid/PharoDEVS:main/src';
 load

Running examples

TODO

History

Current implementation is Pharo port and continuation of original SmallDEVS library implemented in Squeak 4.3 at Department of Intelligent Systems on FIT BUT (namely doc. Ing. Vladimír Janoušek Ph.D., Ing. Elöd Kironský).

Overview of changes

This is list of changes compared to original SmallDEVS implementation:

• removed all squeak and repository/proto object code
• cleaned core and made class prefixes etc. consistent
• removed all SIXX and storage dependencies
• added a project baseline
• initial unit tests added, fixed bugs/inconsistencies
• added github actions to run tests on commits
• added a simple traffic system simulation that enabled me to understand
• removed all Transcript based logging and introduced events
• inspector extension for a simulation logger can be used to have a decent log of everything happened

TODOs and next steps, open questions

• separate roles of solver, model and coordinator
• clarify DEVSnDESS code that is there
• parallel stuff looks too naive and needs improvement
• Bloc based UI?

---

Key Differences between DES and DEVS

• DES is more straightforward and is widely used for simpler systems where events are the primary drivers of change.
• DEVS is a more rigorous and flexible framework, suitable for complex systems requiring modularity and hierarchical organization.

DES (Discrete Event Simulation) and DEVS (Discrete Event System Specification) are both approaches to modeling and simulation, but they have different focuses and structures:

Aspect	DES	DEVS
Focus	Event scheduling and processing.	Modular, hierarchical system specification.
Time Management	Event-driven (time jumps between events).	Event-driven with explicit time advance functions.
Modeling Approach	Informal, system-specific implementations.	Formal, mathematical framework.
Structure	Flat, centralized event list.	Hierarchical composition of atomic and coupled models.
Reusability	Limited modularity, typically system-specific.	High modularity and reusability.
Applications	Queueing, logistics, manufacturing systems.	Complex, hierarchical systems like distributed simulations.

Discrete Event Simulation (DES)

• Concept:
  • DES is a simulation approach where the state of the system changes only at specific points in time, triggered by events.
  • Time advances by jumping from one event to the next (event-driven).
• Characteristics:
  • Events are instantaneous and cause state transitions.
  • Typically used to model queueing systems, manufacturing processes, and other systems where changes occur at discrete points (e.g., arrival of a customer, completion of a task).
  • Often implemented using event lists, where events are processed chronologically.
• Example Applications:
  • Customer service systems (banks, call centers).
  • Production lines in factories.
  • Computer network packet transfers.

---

DEVS (Discrete Event System Specification)

• Concept:
  • DEVS is a formal modeling framework for discrete event systems, developed by Bernard Zeigler.
  • It provides a modular and hierarchical way to specify systems using a well-defined mathematical structure.
• Characteristics:
  • Consists of two main types of models:
    • Atomic Models: The basic building blocks defining state variables, inputs, outputs, internal and external transitions, and time advances.
    • Coupled Models: Compositions of multiple atomic and/or coupled models, defining how they interact.
  • Supports modularity and reusability, making it easier to build complex systems from simpler components.
  • Formalism allows rigorous analysis, validation, and verification of the models.
• Example Applications:
  • Complex system simulations, such as transportation systems, military simulations, and distributed systems.
  • Hierarchical and modular systems, e.g., smart cities, IoT systems.