Jochem Snuverink

Dr. Jochem Snuverink

orcid

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
28 June 2022 in Physical Review Accelerators and Beams
Verified
Journal Article
Open access
PDF file
JKJohannes Kirschner
Johannes Kirschner
MMMojmir Mutný
Mojmir Mutný
AKAndreas Krause
Andreas Krause
JVJaime Maria Coello De Portugal - Martinez Vazquez
Jaime Maria Coello De Portugal - Martinez Vazquez
NHNicole Hiller
Nicole Hiller
JSJochem Snuverink
Jochem Snuverink
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 in each iteration, such as 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 facilities of the PSI: (a) the SwissFEL and (b) HIPA. We report promising experimental results on both machines, tuning up to 16 parameters subject to 224 constraints.

12 March 2021 in Information
Verified
Journal Article
Open access
PDF file
SLSichen Li
Sichen Li
MZMélissa Zacharias
Mélissa Zacharias
JSJochem Snuverink
Jochem Snuverink
JVJaime Maria Coello De Portugal - Martinez Vazquez
Jaime Maria Coello De Portugal - Martinez Vazquez
FPFernando Perez-Cruz
Fernando Perez-Cruz
AAAndreas Adelmann
Andreas Adelmann
The beam interruptions (interlocks) of particle accelerators, despite being necessary safety measures, lead to abrupt operational changes and a substantial loss of beam time. A novel time series classification approach is applied to decrease beam time loss in the High-Intensity Proton Accelerator complex by forecasting interlock events. The forecasting is performed through binary classification of windows of multivariate time series. The time series are transformed into Recurrence Plots which are then classified by a Convolutional Neural Network, which not only captures the inner structure of the time series but also uses the advances of image classification techniques. Our best-performing interlock-to-stable classifier reaches an Area under the ROC Curve value of 0.71±0.01 compared to 0.65±0.01 of a Random Forest model, and it can potentially reduce the beam time loss by 0.5±0.2 s per interlock.

22 October 2022 in Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
Verified
Journal Article
Open access
MTM. Haj Tahar
M. Haj Tahar
DKD. Kiselev
D. Kiselev
AKA. Knecht
A. Knecht
DLD. Laube
D. Laube
DRD. Reggiani
D. Reggiani
JSJochem Snuverink
Jochem Snuverink

20 February 2024 in IEEE Transactions on Applied Superconductivity
Verified
Journal Article
Closed access
RRR. Riccioli
R. Riccioli
CCC. Calzolaio
C. Calzolaio
AGA. Gabard
A. Gabard
PKPeter-Raymond Kettle
Peter-Raymond Kettle
DKD. Kiselev
D. Kiselev
EVEremey Valetov
Eremey Valetov
The Isotope and Muon Production using Advanced-Cyclotron and Target Technology Project (IMPACT) foresees the introduction of two new target stations and three new beamlines: one for radionuclide production and two for surface muon production. The latter forms the project, High-Intensity Muon Beams (HIMB), which plans to increase the muon rate from the current world record of 10 8 µ + /s up to 10 10 µ + /s. This work presents an overview of the future HIMB beamlines focusing on the magnet designs that have been developed to ensure increased muon rate production and transmission. Specific radiation-hard resistive coils, based on mineral insulation, are required in this case due to the proximity to the target station. A high muon capture and transmission efficiency requires solenoid-like magnets, as well as dipole magnets and crossed-field separators to select the desired particles, while suppressing unwanted background particles. The radiation-hard capture solenoid plays the most important role in the whole beamline since it must provide a high capture efficiency. Beam optics studies provided the on-axis field profile necessary for optimizing the size and shape of the capture solenoid. Therefore, the article will also elucidate on these solenoid design strategies for achieving the desired capture efficiency.

24 June 2019 in Physical Review Accelerators and Beams
Verified
Journal Article
Open access
PDF file
MFMatthias Frey
Matthias Frey
JSJochem Snuverink
Jochem Snuverink
CBChristian Baumgarten
Christian Baumgarten
AAAndreas Adelmann
Andreas Adelmann
The usage of numerical models to study the evolution of particle beams is an essential step in the design process of particle accelerators. However, uncertainties of input quantities such as the beam energy and magnetic field lead to simulation results that do not fully agree with measurements; hence, the final machine will behave slightly differently than the simulations. In the case of cyclotrons, such discrepancies affect the overall turn pattern or may even alter the number of turns in the machine. Inaccuracies at the PSI Ring cyclotron facility that may harm the isochronism are compensated by additional magnetic fields provided by 18 trim coils. These are often absent from simulations or their implementation is very simplistic. In this paper, a newly developed realistic trim coil model within the particle accelerator framework OPAL is presented that was used to match the turn pattern of the PSI Ring cyclotron. Because of the high-dimensional search space consisting of 48 design variables (simulation input parameters) and 182 objectives (i.e., turns), the simulation and measurement cannot be matched in a straightforward manner. Instead, an evolutionary multiobjective optimization with a population size of more than 8000 individuals per generation together with a local search approach were applied that reduced the maximum absolute error to 4.5 mm over all 182 turns.

3 December 2025 in arXiv
Verified
Preprint
Open access
PDF file
MTM. Haj Tahar
M. Haj Tahar
WJW. Joho
W. Joho
ESE. Solodko
E. Solodko
MBM. Bocchio
M. Bocchio
SMS. Marquie
S. Marquie
MSM. Schneider
M. Schneider
Reliable operation of high-power proton cyclotrons is a critical requirement for Accelerator Driven Systems (ADS) and other large-scale applications. Beam tuning in such machines is traditionally performed manually, a process that can be slow, non-optimal, and difficult to execute in the presence of faults or changing conditions. To address this, we developed and deployed a machine learning (ML) based tuning framework on the Injector 2 cyclotron at PSI, chosen as an ideal testbed for high-power operation. The system combined a tailored reinforcement learning algorithm with real-time diagnostics and control, and incorporated accelerator-physics inspired adaptations such as an overshoot strategy that reduced magnetic field settling times by nearly a factor of six. Over an extensive 12-day operational test campaign, relatively long in the context of real-time ML experiments, the ML agent successfully tuned the machine across multiple operating points, achieving convergence within hours and maintaining stable beam extraction with reduced losses. Beyond initial tuning, the system was also operated in evaluation mode overnight, where it autonomously monitored and corrected the machine to compensate for drifts, demonstrating robustness and long-term stability. Crucially, the learned policy generalized reliably from low-current training to higher-current operation, underscoring its scalability. These results constitute the first demonstration of ML-assisted tuning on a high-power cyclotron, with direct relevance to ADS-class drivers.