Assignments

Motivated by the IPL ModeliScale initiative and part of the Glose project funded by Safran, the objective of this postdoctoral position is to investigate compile-time and run-time techniques for the distributed simulation of large cyber-physical systems. This includes Structural Analysis algorithms for partitioning large Modelica models into networks of FMI components that can be efficiently simulated on a distributed computing architecture, but also novel numerical schemes such as Quantized State Systems (QSS) methods and code-generation techniques for these methods.

Results

  • State of the art on the numerical methods used in continuous-time and hybrid co-simulation.
  • Experimentation with the Quantized State Systems (QSS) method applied to co-simulation scenario.
  • Formalization of a model-based approach for co-simulation and model exchange.

Environment

Modeling & Simulation Modelica, FMI standard, Python, Matlab.
DevOps Gitlab, Make, Docker.
Documentation MS Office, Markdown, Latex, Jupyter.