Speaker
Description
Beam commissioning of slow extracted beams from the CERN Super Proton Synchrotron (SPS) to the North Area experimental targets requires trajectory control through multiple transfer lines using corrector magnets—a process that traditionally demands significant expert intervention. Previous work demonstrated the feasibility of applying reinforcement learning (RL) for automated trajectory correction based on secondary emission monitor (SEM) split-foil intensity measurements, successfully centering the beam on target under nominal conditions. However, this approach fails when the beam is lost or its position exceeds the SEM's active surface, and when the corrector magnets' polarities are not known; common sources of uncertainty during commissioning.
We present an extended multi-stage optimization scheme that addresses these critical limitations by automating beam threading when the trajectory exceeds the SEMs' acceptance, systematically identifying corrector magnet polarity configurations, and optimizing the impact angle to maximize beam intensity at the fixed-target stations, measured by scintillators arranged around the target. The threading algorithm employs quasi-random search combined with Bayesian optimization (BO) to center the beam in the SEMs, before handing over to the RL controller. The automated polarity determination uses online system identification to resolve sign ambiguities in the correctors, eliminating a common source of commissioning delays when using RL or other dedicated steering algorithms. Finally, BO is used to optimize the position of the movable SEM monitors at the targets' locations, maximizing target intensity.
| In which format do you inted to submit your paper? | LaTeX |
|---|---|
| Preprint marking on your proceeding paper | I wish my paper to be marked as preprint. |