AI techniques and strategies for Linac tuning at LNL

WECDC01
Oct 22, 2025, 5:00 PM
20m
room C (conference center)

room C

conference center

Contributed Oral Presentation WGB:Beam Dynamics in Linacs WECDC WGB contribute oral

Speaker

Michele Comunian (Istituto Nazionale di Fisica Nucleare)

Description

Based on the operational experience on the superconducting heavy ion linac ALPI, we introduced Machine Learning techniques to improve the performances and speed up the tuning of the ion linac at LNL.
The technique proved to be particularly effective to overcome issues on the longitudinal matching.
The algorithm has been adapted to be compatible with high intensity linacs, in terms of transverse and longitudinal acceptance, available diagnostics and machine protection prescriptions.
After a detailed preparatory simulation work, the algorithm has been tested during the 2025 commissioning shift of the ESS linac.
In this paper the results obtained in the ALPI and ESS campaigns.

I have read and accept the Privacy Policy Statement Yes

Authors

Francesco Grespan (Istituto Nazionale di Fisica Nucleare) Luca Bellan (Istituto Nazionale di Fisica Nucleare)

Co-authors

Andrea Pisent (Istituto Nazionale di Fisica Nucleare) Damiano Bortolato (Istituto Nazionale di Fisica Nucleare) Domenic Nicosia (European Spallation Source) Enrico Fagotti (Istituto Nazionale di Fisica Nucleare) Maurizio Montis (Istituto Nazionale di Fisica Nucleare) Michele Comunian (Istituto Nazionale di Fisica Nucleare) Natalia Milas (European Spallation Source) Ryoichi Miyamoto (European Spallation Source) Ysabella Kassandra Ong (Istituto Nazionale di Fisica Nucleare)

Presentation materials

There are no materials yet.