[BTeam] Tuning Particle Accelerators with Safety Constraints using Bayesian Optimization

[Jay Benesch]

https://arxiv.org/abs/2203.13968

Tuning Particle Accelerators with Safety Constraints using Bayesian 
Optimization
Johannes Kirschner, Jaime Coello de Portugal, Jochem Snuverink, Nicole 
Hiller, Mojmir Mutný, Andreas Krause

     Tuning machine parameters of particle accelerators is a repetitive 
and time-consuming task, that is challenging to automate. While many 
off-the-shelf optimization algorithms are available, in practice their 
use is limited because most methods do not account for safety-critical 
constraints that apply to each iteration, including loss signals or 
step-size limitations. One notable exception is safe Bayesian 
optimization, which is a data-driven tuning approach for global 
optimization with noisy feedback. We propose and evaluate a step 
size-limited variant of safe Bayesian optimization on two research 
faculties of the Paul Scherrer Institut (PSI): a) the Swiss Free 
Electron Laser (SwissFEL) and b) the High-Intensity Proton Accelerator 
(HIPA). We report promising experimental results on both machines, 
tuning up to 16 parameters subject to more than 200 constraints.

Subjects: 	Accelerator Physics (physics.acc-ph); Machine Learning (cs.LG)
Cite as: 	arXiv:2203.13968 [physics.acc-ph]

The first paragraph in section III.A

The different versions of the BO algorithm with con-
straints have been tested in several dedicated experi-
ments at HIPA. The target of these experiments has been
to reduce the overall beam losses around the machine by
minimizing a target signal defined as a weighted sum of
about 60 beam loss monitors (or loss related monitors)
spread across the machine. This weighted sum reflects
that the beam loss monitors are not distributed equidis-
tantly along the machine and that losses at lower energies
cause less activation.