SysID: System identification

This is a collection of Jupyter notebooks and scripts to demonstrate the capablities of an opensource system identification package SIPPY. A comparison study of MATLAB system identification toolbox vs SIPPY can be found here

Installation

1. Install Miniconda.

2. Open conda prompt and install Jupyter using the following command.

conda install -c conda-forge jupyterlab

3. Create control conda environment using the following command. You can use my yml file from here.

conda env create -f environment.yml

4. Verify that the new environment was installed correctly.

conda env list

This should return something similar as given below:

base                  *  C:\Miniconda3
control                  C:\Miniconda3\envs\control

5. Activate the new environment.

conda activate control

6. To use Jupyter in conda env run the following command.

python -m ipykernel install --user --name=control

7. Clone SIPPY from Github.

• Colne the SIPPY from Github by running the following command (You should have git installed).

git clone "https://github.com/CPCLAB-UNIPI/SIPPY"

• Else, download it from Github, then extract the zip file to a local folder.

8. To install SIPPY to control env, open conda prompt and cd (change directory command) to the SIPPY folder then run the following commands.

conda activate control

Then

python setup.py install

9. To test, follow the below stpes.
   • cd to Example folder folder within SIPPY folder. Then run the following command. Then

python SS.py

This should show a trend of prediction after running a N4SID (As configured in the script).

10. Clone my SysID

• Colne the SIPPY from Github by running the following command (You should have git installed).

git clone "https://github.com/jamestjsp/sysid"

• Else, download it from Github, then extract the zip file to a local folder.

Usage

1. Open conda prompt then activate control env.

conda activate control

2. cd to SysID folder then, launch Jupyter lab by running following command.

jupyter lab

This should open Jupyter Lab in your browser then open notebook folder and try test the files model_id_democol.ipynb and model_id_uptream_june_dataset_HP_section.ipynb

Documents and reference

SIPPY document assumes that you have a good understanding about system identification. When I started with this library I struggled a little and Aspentech whitepaper helped me to pickup this useful library.

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

License

MIT