Examples - Overview

This section discusses the included examples of the FMIFlux.jl library. You can execute them on your machine and get detailed information about all of the steps. If you require further information about the function calls, see library functions section. For more information related to the setup and simulation of an FMU see FMI.jl library.

The examples are intended for users who work in the field of first principle and/or data driven modeling and are further interested in hybrid model building. The examples show how to combine FMUs with machine learning ("NeuralFMU") and illustrates the advantages of this approach.

Examples

• Simple CS-NeuralFMU: Showing how to train a NeuralFMU in Co-Simulation-Mode.
• Simple ME-NeuralFMU: Showing how to train a NeuralFMU in Model-Exchange-Mode.

Advanced examples: Demo applications

• JuliaCon 2023: Using NeuralODEs in real life applications: An example for a NeuralODE in a real world engineering scenario.
• Modelica Conference 2021: NeuralFMUs: Showing basics on how to train a NeuralFMU (Contribution for the Modelica Conference 2021).

Workshops

Pluto workshops: Pluto based notebooks, that can easily be executed on your own Pluto-Setup.

• Scientific Machine Learning using Functional Mock-up Units: Workshop at JuliaCon 2024 (Eindhoven University, Netherlands)

Archived

• MDPI 2022: Physics-enhanced NeuralODEs in real-world applications: An example for a NeuralODE in a real world modeling scenario (Contribution in MDPI Electronics 2022).
• Growing Horizon ME-NeuralFMU: Growing horizon training technique for a ME-NeuralFMU.
• Hands-on: Hybrid Modeling using FMI: Workshop at MODPROD 2024 (Linköping University, Sweden)