AN INTEGRATED MODELING, SIMULATION AND EXPERIMENTATION ENVIRONMENT IN PYTHON']PYTHON BASED ON SES/MB AND DEVS

This paper proposes a Python-based infrastructure for studying the characteristics and behavior of families of systems. The infrastructure allows automatic execution of simulation experiments with varying system structures as well as with varying parameter sets in different simulators. Possible system structures and parametrizations are defined using an extended System Entity Structure (SES). Based on rules one specific system configuration is derived. In connection with a model base (MB) an executable model for multiple simulation platforms can be created. This paper focuses on the specification and generation of Discrete Event System Specification (DEVS) models for the DEVS simulator DEVSimPy. The derived model is executed in the DEVSimPy simulator. Depending on the results the subsequent derivation and generation of a model is controlled reactively. The proposed infrastructure is illustrated using an example that includes various abstraction hierarchies and time granularity levels.