modelica-tutorial

A tutorial repo for modeling, simulation and optimization in Modelica/FMU environments

Syllabus

Section 1 - Introduction on Modelica

Section 2 - Debugging Modelica Models

Section 3 - Advanced Applications

This section talks about how to use Modelia/FMU models to perform different advanced applications, including design optimization (e.g., system sizing, model calibration etc), simulation restart from a given states, pertubation experiment (e.g., system excitation for system identification) and optimal control (e.g., model predictive control of a target system) etc..

Section 3.0 Some Basics

states

tunable parameters

Modelica compilers

FMU

Section 3.1 Modelica-based Design Optimization

Section 3.2 Modelica-based Simulation Restart

Section 3.3 Modelica-based Pertubation Experiment

Section 3.4 Modelica-based Optimal Control

perfect model

imperfect model

Section 3.5 FMU-based Design Optimization

Section 3.6 FMU-based Simulation Restart

Section 3.7 FMU-based Pertubation Experiment

Section 3.8 FMU-based Optimal Control

perfect model

imperfect model

Section 3.9 FMU-based Co-simulation among Different Simulators

Software Installation

The testing examples are mostly written in Python, and the physical models are written in Modelica. Two integrated software environment are provided here for tutorial purposes. One is a containerized docker environment that contains jModelica for Modelica modeling and simulation, and a Python 3.8 scripting envrionment. The other is a Python 3.8 conda environment, which contains only Python and no Modelica tools. This environment is only designed for examples depending on Dymola or Dymola FMU. Dymola is similar to jModelica but a commercial tool.

Docker Environment

The docker contains an integrated Modelica modeling and simulation environment in Python based on jModelica.

Install prequirements

Docker: https://docs.docker.com/get-docker/.

Make:
```
 for windows, use cygwin from https://www.cygwin.com/
```
Activate docker. Just open the installed Docker desktop.
Make an image for tutorial purposes by opening a terminal, directing to the top folder modelica-tutorial, and typing:
```
make build
```
Check the image after build by typing:
```
docker image ls
```
If successfully build, this should return a list of images on the local computer with an image repository named modelica_tutorial.
Test the Docker environment by directing to any sub-folders and type:

on windows:
```
XXXX.bat
```
on Mac/Ubuntu:
```
bash XXXX.sh
```

Conda Environment

The library dependency has been configured and listed in the modelica_tutoral.yml file.

Make sure conda is installed in the local computer by opening your favorite terminals and typing:

conda info

If sucessfully installed, similar messages as follows should be returned:

 active environment : base
 active env location : C:\tools\Anaconda2
 shell level : 1
 user config file : C:\Users\yangyang.fu\.condarc
 populated config files :
 conda version : 4.8.3
 conda-build version : 3.18.9
 python version : 2.7.16.final.0
 virtual packages : __cuda=11.3
 base environment : C:\tools\Anaconda2  (writable)
 channel URLs : https://repo.anaconda.com/pkgs/main/win-64
                     https://repo.anaconda.com/pkgs/main/noarch
                     https://repo.anaconda.com/pkgs/r/win-64
                     https://repo.anaconda.com/pkgs/r/noarch
                     https://repo.anaconda.com/pkgs/msys2/win-64
                     https://repo.anaconda.com/pkgs/msys2/noarch
 package cache : C:\tools\Anaconda2\pkgs
                 C:\Users\yangyang.fu\.conda\pkgs
                 C:\Users\yangyang.fu\AppData\Local\conda\conda\pkgs
 envs directories : C:\tools\Anaconda2\envs
                     C:\Users\yangyang.fu\.conda\envs
                     C:\Users\yangyang.fu\AppData\Local\conda\conda\envs
 platform : win-64
 user-agent : conda/4.8.3 requests/2.22.0 CPython/2.7.16 Windows/10 Windows/10.0.19041
 administrator : False
 netrc file : None
 offline mode : False

If conda is not sucessfully installed, please check https://conda.io/projects/conda/en/latest/user-guide/install/index.html.

Configure the provided environment in modelica_tutorial.yml by directing to the top folder modelica-tutorial and typing:
```
conda env create -f modelica_tutorial.yml
```
Check the installation by typing:
```
conda env list
```
If installed correctly, users should be able to see a conda environment named modelica_tutorial being returned.
Activate modelica_tutoral virtual environment if the environment has been successfully installed. Type in the terminal:
```
conda activate modelica_tutoral
```
Test example codes. Now the users can direct to any sub-folders to test the example scripts written in Python by typing
```
python xxxx.py
```
If the scripts call FMUs, then the FMUs has to be regenerated on the local computer using Dymola.

Contact

Yangyang Fu

fuyy2008@gmail.com

YangyangFu/modelica-tutorial