17–22 May 2026
C.I.D
Europe/Zurich timezone

ML surrogate models and non-invasive diagnostics for real time accelerator control

MOP6335
18 May 2026, 16:00
2h
C.I.D

C.I.D

Deauville, France
Poster Presentation MC6.D13: Instrumentation: Artificial Intelligence Poster session

Speaker

Balsa Terzic (Old Dominion University)

Description

We propose to develop advanced ML models such as neural network (NN) based surrogate models to accurately represent accelerator phase space transport. These surrogate models will enable precise diagnosis and prediction of beam phase space evolution along the beamline, facilitating real-time control and optimization. The developed models will be tested using the Continuous Electron Beam Accelerator Facility (CEBAF) at JLab, providing a pathway toward ML-driven beamline control in operational accelerator environments.
Potential experiments, initially carried out at CEBAF, but later adapted for other facilities, include bunch size and length measurement using existing diagnostics and energy spectrum measurement. We will apply ML models such as Physics-Informed Neural Networks and other neural networks to design an online ML-driven diagnostics platform integrating physics knowledge and diagnostics information. We will also build a virtual beamline, train an reinforcement learning (RL) controller across varied calibration scenarios, and then transfer it to the real machine.

In which format do you inted to submit your paper? LaTeX
Preprint marking on your proceeding paper I do not wish my paper to be marked as preprint.

Author

Monika Yadav (Old Dominion University)

Co-authors

Balsa Terzic (Old Dominion University) Léon van Riesen-Haupt (École Polytechnique Fédérale de Lausanne) Dr Qianqian Su (Old Dominion University)

Presentation materials

There are no materials yet.