From simple physical systems to full production lines, numerical models could be used to minimize downtime and to optimize performances. In this article, the system of interest is the SPIRAL2 (Syst\`eme de Production d'Ions RAdioactifs en Ligne de 2e g\'en\'eration) particles accelerator cryogenic system. This paper illustrates three totally different applications based on a SPIRAL2 cryostat model: optimal controller synthesis, virtual sensor synthesis and anomaly detection. The tow firsts applications have been deployed on the system. Experimental results are used to illustrate the benefits of such applications.
Cite
@article{arxiv.2103.10299,
title = {SPIRAL2 Cryomodules Models: a Gateway to Process Control and Machine Learning},
author = {Adrien Vassal and Adnan Ghribi and François Millet and François Bonne and Patrick Bonnay and Pierre-Emmanuel Bernaudin},
journal= {arXiv preprint arXiv:2103.10299},
year = {2021}
}