10–13 Jun 2024
Parkhotel Bad Zurzach, Switzerland
Europe/Zurich timezone

A physic-guided recurrent machine learning model for long-time prediction of quench dynamics

12 Jun 2024, 14:00
30m
Paul Scherrer Institute (PSI)

Paul Scherrer Institute (PSI)

Speakers

Prof. Frédéric Sirois (Polytechnique Montreal) Nicolas Breault (Polytechnique Montréal)

Description

This work presents a PGRML model for predicting superconducting tape temperature profiles during a quench. The model accurately forecasts the nonlinear dynamics involved in a drastically reduced computational time with respect to finite element models.

Topic Innovative methods and tools for modelling large-scale HTS systems

Primary authors

Christian Lacroix (Polytechnique Montréal) Prof. Frédéric Sirois (Polytechnique Montreal) Nicolas Breault (Polytechnique Montréal)

Presentation materials