FMI in the Cloud

This repository contains the materials and instructions for the tutorial FMI in the Cloud at the 14th International Modelica Conference 2021.

Prerequesites

• a Google account to run the Python Notebook on Google Colab
• a clone or local copy of this repository
• a GitHub account to fork this repository
• a Conda environment with FMPy to run the Jupyter Notebook locally and to run the Web App
• a Python IDE or text editor to write the Web App

To create the Conda environment

• install Miniconda (if you don't already have Conda installed)

• create a new Conda environment

conda create -n fmi-tutorial -c conda-forge python=3.9 fmpy notebook ipywidgets

• activate the environment

conda activate fmi-tutorial

Part 1: Create a Jupyter Notebook

Option 1: run the Jupyter Notebook on the cloud

• open Heater.ipynb on Google Colab

Option 2: run the Jupyter Notebook locally

• change into the directory where you downloaded or cloned this repository and run

jupyter Heater.ipynb

Part 2: Create a Web App

Resources and inspirations:

• Dash Bootstrap Components
• Plotly Python Examples
• Bootstrap CSS Framework
• generic FMPy Web App
• FMU Check (source code)
• FMI-based simulation workflows based on open source and commercial tools (Thu, 13:50 h)

Exercises:

• change the default simulation time to 50 s
• change the result variables
• add an additional parameter
• add the description from the modelDescription.xml below the image
• create a custom input signal
• deploy the web app

Part 3: Hackathon and Q&A

Build your own Jupyter Notebooks and Web Apps and ask any questions about FMPy.

• explore the test FMUs in the FMI Cross-Check
• fork this repository and share your results!