AI-ASSIST is a research project focused on improving power-grid analysis and control, with a strong emphasis on real-time security assessment and grid stability estimation.
The project applies advanced AI methods to evaluate how safe and stable a power system is under current operating conditions. By combining live measurements with a large historical database, AI-ASSIST uses pattern-recognition techniques to match current states with known operating scenarios. This enables accurate risk assessment and system-behavior prediction, helping grid operators make informed decisions.
AI-ASSIST is a joint initiative of:
- SensorLab at the Jožef Stefan Institute
- Laboratory of Electric Power Supply at the Faculty of Electrical Engineering, University of Ljubljana
- ELES, the Slovenian transmission system operator
This repository contains code, data, and analyses related to the AI-ASSIST project.
At this stage, the public analyses focus primarily on the IEEE 39-bus test case, with selected results included in this repository. Research on the Slovenian power grid is ongoing and will be incorporated as appropriate.
Use the following steps to set up the environment and reproduce the IEEE 39-bus and ELES pipeline:
- Create a virtual environment:
python -m venv .venv- Activate the virtual environment:
source .venv/bin/activate- Create a local environment file:
cp .env.example .env-
Update the variables in
.envfor your environment (API keys, paths, hosts, ports, etc.). -
Install project dependencies:
# see Makefile
pip install -e .- Prepare the dataset in the dataset config directory:
# required flow: config/<dataset>/dvc.yml
# in this repository, dataset configs are under configs/<dataset>/dvc.yaml
cd configs/bus39
dvc repro- Start the service with Docker Compose:
docker compose up --buildPipeline configuration path:
config/<dataset>/dvc.yml(required flow)configs/bus39/dvc.yaml(path used in this repository)
├── configs/ # Configuration files for experiments and pipelines
│ └── bus39/ # DVC pipeline configuration for IEEE 39-bus workflow
├── data/ # Public datasets used for experiments
│ └── bus39/ # IEEE 39-bus related data
├── reports/ # Jupyter notebooks with analyses and reports
├── scripts/ # Utility and data processing scripts
│ └── bus39/transform.py # Data transformation for IEEE 39-bus workflow
├── src/ # Source code (metrics, preprocessing, utilities)
└── README.md # Project documentation
-
IEEE 39-Bus System Includes power-flow simulations, machine learning-based stability analysis, and visualization.
-
Slovenian Grid Network (in progress) Due to data sensitivity, this part of the analysis is not included in the public repository. Access to related data may be available upon request in the future.
The AI-ASSIST project is funded by the Slovenian Research and Innovation Agency (ARIS) under Grant Agreement No. L2-50053.
- SensorLab, Jožef Stefan Institute
- Laboratory of Electric Power Supply, Faculty of Electrical Engineering, University of Ljubljana
- ELES, Slovenian transmission system operator
This project is licensed under the MIT License. See LICENSE for details.
For project-related inquiries, please refer to: