AI for Real-Life and Full-Scale Power Grid Operations using EnerGNN – A Hands-On Approach

Do you want to learn how to train a Graph Neural Network on real power grid data in under an hour?

Then welcome to this EnerGNN workshop, hosted by RTE, the French Transmission System Operator!

During the webinar, you will discover the EnerGNN library and train your first model on an electrical network through an interactive Python notebook.

EnerGNN is an open source project hosted by LF Energy, which provides a framework for applying Graph Neural Networks to energy networks. Among other features, EnerGNN offers:

• A generic, domain-agnostic graph representation, designed to handle power grids as well as other network infrastructures;
• Ready-to-use GNN architectures based on message passing, able to scale to real-world transmission networks;
• Seamless data import from PowSyBl, so you can turn your grid models into training datasets in a few lines of code.

The library is written in Python on top of JAX and Flax, benefiting from just-in-time compilation and GPU acceleration.

What you will learn during the webinar:

• For beginners: discover how machine learning can be applied to power systems through a quick general presentation and the notebook, without needing any deep learning expertise;
• For advanced developers and researchers: master the tools to build, train and evaluate your own GNN models on your real-world grid scenarios.

Speakers:
Balthazar Donon, Research Scientist
Maria Roman

Organizer: Stephan Gross, Community Director, Europe

Webinar Host: RTE

Link: https://ocgroups.dev/lf-energy/group/htc7btu/event/abaz4k3