id stringlengths 36 36 | text stringlengths 1 1k | source stringlengths 42 191 | page int64 0 1.03k |
|---|---|---|---|
16d61943-d9fd-42c9-b2c0-9b7318201855 | erageoperationalperformance. Inparallel,thedata-driven
surrogate model was improved through an enhanced tran-
sitionmodel loss incorporating symmetry and fixed-point
constraintstoenforcephysicalconsistencyandmitigateerror
accumulation during multi-step rollouts. The offline-trained
RLagentsreachednear-targetinjectionef... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 0 |
ccd4a201-5544-405d-bc92-b419553dcbf1 | Observation Space
The design of an informative and compact observation
spaceis centralto RLagentperformance. Theobservation
must capture sufficient information about the machine state
to allow the agent to infer corrective actions, while remain-
ing low-dimensional enough for efficient training on limited
operationalda... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 0 |
d2d46c08-6c1e-4f38-8113-f6457d1a9f55 | gitudinalSchottkyspectrum,whichencodesthemomentum
distribution of the circulating beam as a function of cycle
time throughout the injection plateau. For each machine
cycle, the Schottky spectra acquired at successive injection
stepsareconcatenatedintoasingletwo-dimensionalarray
spanning frequency and cycle time. This h... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 1 |
b1c5c8eb-fd0a-4b9d-ab5b-758e59200321 | summary of the beam state at each cycle.
A key development improving the stability of offline sur-
rogatemodeltrainingwasaugmentingtheobservationspace
with time-of-flight (ToF) measurements from the Linac3-to-
LEIRtransferline. Thesecycle-boundmeasurementsexploit
pairsofcapacitiveBeamPositionMonitorsinthetransfer
line:... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 1 |
863ffc7b-bae1-4fea-a49e-c71db964ca84 | ment, the mean energy of the stripped Pb54+beam drifts,
manifested as a shift in the intercept over days to weeks.
IncludingthisobservablegivestheRLpolicyexplicitaccess
tothedominantsourceofperformancedegradation,rather
than requiring indirect inferencefrom the Schottkyspectra
alone,wherethefoilagingsignalisconvolvedwi... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 1 |
a710fec7-cbf7-4a0f-b978-2ac4465e4dc7 | spondtodifferentfoillifetimes. Thisenabledeffectivereuse
of historical data collected across multiple foil installations,
as the ToF features provide the context needed to align data
from different epochs into a coherent training set.
Action Space
As shown in [ 1], the energy distribution delivered by
Linac3canberestor... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 1 |
c6e0e93c-dc0b-4f1e-ae4f-07a2c5c4ef87 | process must also be matched to the actual beam energy by
adjustingthestartandendvaluesoftheelectrongunvoltage
program during the injection plateau. In addition, spatial
overlap between ion and electron beams at the cooler sec-
tion is controlled by four corrector magnet settings defining
thehorizontalandverticalpositi... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 1 |
2f329dad-96cd-4185-ac73-a703c9a3d88e | onacycle-by-cyclebasis,proposingadjustments 𝛿ptothese
nineparametersbasedonthecurrentobservation. Theob-
jective it maximizes is the injection efficiency, 𝑟 =𝑖injected
∑8
𝑗=1𝑖𝐿3
𝑗,
where𝑖injectedisthebeamintensityintheLEIRringatthe
peakoftheinjectionprocessand 𝑖L3
𝑗istheintensityofthe
𝑗-th pulse measured at ... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 1 |
760b658f-5c06-4c99-bdba-9d6e842c7f6a | Vertical cooler bump position [mm] -0.75 5.0
Vertical cooler bump angle [mrad] 0.26 1.65
OPERATIONAL RESULTS
The RL agent was deployed online at LEIR during the
2025 lead-ion run in a cycle-by-cycle closed loop. Two con-
figurations were tested: an agent trained from scratch on
the machine (online), and an agent pretra... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 1 |
42d69add-64cf-402a-a5b8-4f21c739bd21 | bleduringtraining,butprovedinsufficientlysample-efficient
given the 3.6 s cycle time. SAC was the only algorithm
that captured the stochasticity of the LEIR injection process
while maintaining adequate sample efficiency. To ensure
stabletraining,severalstabilizationtechniqueswereadopted17th International Particle Accel... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 1 |
bb505fd4-c798-4308-b701-830e3299693a | Figure 1: Top:Evolution of injection efficiency from Linac3 to LEIR during online training and fine-tuning. Training
curvesaresmoothedwithWelford’sonlinealgorithmwith0.95CI[ 9].Bottom:Intensitymeasuredwithabeamcurrent
transformer in the Linac3-to-LEIR transfer line.
following[ 7]. Inparticular,learningratewarm-upwasuse... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 2 |
fdcfd5a2-4a90-4020-bd35-8617b3017628 | The pretrained agent improves significantly earlier, reach-
ingabove0.20within roughly500 cycles(approximately2
hours with 1 out of 2 cycles), whereas the online agent re-
quires approximately 1500 cycles. After convergence, both
stabilize between0.20 and 0.30, with the pretrained agent
consistently at the upper end. T... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 2 |
8fac911f-f46c-43cc-8e2d-8b3fa8283a8d | the surrogate model provides a useful inductive bias guid-
ing the policy toward a better region of parameter space.
TheseresultsdemonstratethatRLholdspromiseforanau-
tonomous controller for compensating stripper foil aging at
LEIR,maintaininginjectionperformancewithoutmanual
intervention. Further long term evaluations... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 2 |
03bba5fa-e4e6-4452-8cf3-b21ef3f925d7 | Figure2: Evaluationofthepretrainedagentoverrepeated
episodes starting from randomized machine configurations.
Top:Episodelengthinnumberofiterations. Bottom:Initial
(green)andfinal(red)injectionefficiencyforeachepisode.
The shaded band indicates the operational target region.
such as beam commissioning, foil exchanges o... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 2 |
56c29f05-52ce-4e7d-97cd-e930c9185388 | ciency of ∼0.3), driven primarily by shot-by-shot variations
inLinac3delivery: ionsourcefluctuations,transportjitter,
andpulse-to-pulseenergyspreadintroducecycle-to-cycle
variability not attributable to the agent’s actions. This limits
thesurrogatemodel’sabilitytolearnaccuratedynamics,de-
gradingthepretrainedpolicyqual... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 2 |
3fe41f18-44ba-4e5f-92a9-ef398789753d | REFERENCES
[1]S. Hirlaender et al., “Energy Dependence of the Repro-
ducibility and Injection Efficiency of the LINAC3-LEIR Com-
plex”, in Proc. IPAC’19 , Melbourne, Australia, May 2019,
pp. 3188–3191.
doi:1/zero.alt4.18429/JACoW-IPAC2/zero.alt419-WEPTS/zero.alt44/zero.alt4
[2]N.Madysa,R.Alemany-Fernandez,N.Biancacci,B... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 3 |
9074e7fb-8421-42c5-808a-62f5dfac747f | doi:1/zero.alt4.18429/JACoW-IPAC2/zero.alt425-THPM1/zero.alt49
[4]C. P. Burgess et al., “Understanding disentangling in 𝛽-VAE”,
arXiv preprint, 2018.
doi:doi:1/zero.alt4.4855/zero.alt4/arXiv.18/zero.alt44./zero.alt43599
[5]T.Haarnoja,A.Zhou,P.Abbeel,andS.Levine,“SoftActor-
Critic: Off-Policy Maximum Entropy Deep Reinf... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 3 |
7a05f966-a96e-4865-b436-2cc35093cdf5 | [7]J.Bjorck,C. P.Gomes, andK. Q. Weinberger,“Is High Vari-
ance Unavoidable in RL? A Case Study in Continuous Con-
trol”, inProc. 10th Int. Conf. on Learning Representations
(ICLR’22) ,Held online, Apr.2022.
doi:1/zero.alt4.4855/zero.alt4/arXiv.211/zero.alt4.11222
[8]J.Schulman etal.,“ProximalPolicyOptimizationAlgorith... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 3 |
844637a6-8795-483d-8446-5ad1b5c99880 | ISBN: 978-3-95450-252-3 ISSN: 2673-5350 doi: 10.18429/JACoW-IPAC2026-WEP6021
MC6.D13: Instrumentation: Artificial Intelligence 2639WEP: Poster session WEP6021
Content from this work may be used under the terms of the CC BY 4.0 licence (© 2026). Any distribution of this work must maintain attribution to the author(s), t... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6021.pdf | 3 |
148963ae-4c9e-469a-ae52-b2dec1dc0b15 | LOCALIZED RESPONSE BASIS FOR DATA-EFFICIENT TRANSVERSE
BEAM DISTRIBUTION RECONSTRUCTION USING A MULTIMODE
FIBER RELAY
Q. Xu∗1,2, G. Trad2, S. Burger2, A. Gilardi2, M. Bergamaschi2, F. Roncarolo2,
H. D. Zhang1, A. Hill1, C. P. Welsch1
1University of Liverpool, Liverpool, UK; Cockcroft Institute, Daresbury, UK
2CERN, Gen... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 0 |
002828ce-f4c1-483c-bc12-8b64585d2d65 | ter an initial digital micromirror device based validation,
themethod wasassessedusing real beam data from CERN
CLEAR, where data synthesized from a raster scan basis
were used to train a Convolutional Autoencoder. The best
model using this strategy achieved 7.37% mean normalized
rootmeansquareerror(RMSE)acrossfourtran... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 0 |
926c4194-f9aa-4b48-9a3d-fb401fa8ee4d | sociated electronics [ 1]. One possible mitigation is to relay
light from a scintillating screen through a large-core mul-
timodefiber(MMF)toacameraplacedinalowradiation
area[2,3]. However,modalmixingintheMMFscrambles
the spatial intensity distribution, so the fiber output is not
directly interpretable. An inverse reco... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 0 |
8e7792ee-9487-4f3a-8e19-94f435910163 | for synthetic training data. The approach is motivated by
theapproximatelinearityofthemeasuredintensitymapping
underfixedalignmentandacquisitionconditions[ 5,6]. A
∗qiyuan.xu@liverpool.ac.ukset of measured input patterns and corresponding MMF
outputimagesis used asthe basis, and newpairedtraining
samples are generated ... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 0 |
1d6fc789-78a1-42b5-b4a3-d12208003b30 | scribedbyatransmissionmatrix(TM),whichmapsaninput
opticalfieldtoanoutputopticalfield. Inopticslaboratory
systems,suchamatrixcanbecalibratedusinganorthogonal
basis,suchasa Hadamard basis [ 7]. Foraccelerator beam
imaging context, the input is instead an intensity distribu-
tion formed on a screen, so arbitrary optical p... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 0 |
1b60891f-41d8-4c67-bfc3-b66c9bb670bb | 𝐼syn
in(𝑥,𝑦) = ∑
𝑖,𝑗𝑎𝑖𝑗𝐵in
𝑖𝑗(𝑥,𝑦),
𝐼syn
out(𝑢,𝑣) = ∑
𝑖,𝑗𝑎𝑖𝑗𝐵out
𝑖𝑗(𝑢,𝑣).(1)
Here𝐵in
𝑖𝑗istheinputsquareblockatgridposition (𝑖,𝑗),𝐵out
𝑖𝑗
istheMMFoutputimagemeasuredforthatblock,and 𝑎𝑖𝑗is
the shared intensity coefficient. The same coefficient map is
usedtosynthesizebothinputandoutput... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 0 |
15184df2-4945-4ea0-a47d-8e8868bb7426 | ISBN: 978-3-95450-252-3 ISSN: 2673-5350 doi: 10.18429/JACoW-IPAC2026-WEP6024
2640 MC6.D13: Instrumentation: Artificial IntelligenceWEP6024 WEP: Poster session
Content from this work may be used under the terms of the CC BY 4.0 licence (© 2026). Any distribution of this work must maintain attribution to the author(s), t... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 0 |
f1decf06-5b8c-49ef-acd5-6abe514a1e7b | ViewportBeam Pipe
e Beam-
NPBSScintillation Screen
CMOS
Camera 1CMOS
Camera 2
MMF
Ø1500 µm
(15 m)OBJ 1
OBJ 2L1
L2
Transverse
Beam
DistributionMMF Output
Distribution
QFD880 QDD515
DHJ/DVJ840…
ChromoxFigure1: SchematicoftheMMFrelayconceptfortransverse
beamdistributionimaging. Thebeamlineimplementationis
shown; fo... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 1 |
2a028405-539f-4f29-b01a-12fc9c8adcba | output was imaged by a second CMOS camera to record
the corresponding MMF image. Data were collected in two
stages. First,a 32×32positionbasis wasmeasuredbydis-
playingonesquareblockatatimeontheDMD,scannedrow
by row. The block intensity was set high enough to improve
thesignal-to-noiseratiowhileremainingbelowsaturation... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 1 |
149b57ed-29d0-4e74-a414-7eadd3d3df04 | generateddistributionwasdownsampledtothesame 32×32
gridasthepositionbasis,givingthecoefficientmap 𝑎𝑖𝑗. The
coefficient map was then applied directly to the measured
basis pairs by index matching, giving paired synthetic input
and MMF output images for CAE training. Using synthetic
training data generated from the 10... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 1 |
9bbea1c2-5c85-44ea-bbd5-960b12b8b854 | practicalapproximationwasto usea focusedbeamspot on
the Chromox screen as the scanning unit. The raster-scanTable1: CLEARBeamParametersUsedforDataCollection
Beam configuration Specifications
Particle type 𝑒−
Beam energy 192–199 MeV
Repetition rate 10 Hz
Bunches per pulse 1–25
Total pulse charge ∼500 pC (15 bunches)
sp... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 1 |
7b5ce3a9-1974-41b0-af34-6f4ee848e228 | independent evaluation data.
The optical relay followed the same concept as described
aboveandshowninFig. 1. Themaindifferencewasthatthe
MMFlengthwasincreasedto15m,routingthelightfrom
theCLEARbeamlinetotheupstairsshieldedcameraarea.
In total, 1503 raster scan samples were used to form the
basis. A separate random scan ... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 1 |
c96ff9d4-096e-411e-86f7-d670d87da7a3 | wasfirstreducedtoa 23×23coefficientmap. Thiswascho-
sen as a practical compromise between the measured beam
spotsizeandavoidingintensitysaturationinthesynthesized
images. Each raster scan sample was assigned a centroid
positiononthescreen. Foreachcoefficientinthemap,the
nearest raster scan sample was selected according... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 1 |
f46dff83-00c7-42f6-a7fe-841bb7879080 | For comparison, a CAE was trained directly on the raster
scansamples,andalinearTMbaselinewasbuiltfromthe
samerasterscanbasis. Arandomscan referenceCAEwas
also trained using the non-test random scan pairs.
For the best model trained with synthesized data from
beam image priors, representative CLEAR test samples are
show... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 1 |
bce5befd-9ac1-4c11-9375-8c8471fc5d14 | Width
overestimatedReference ReconstructionMulti-peak
partialTypical
caseWidth
underestimatedOff-centre
beam
∣Δμ∣=2.65%
∣Δσ∣=0.92%∣Δμ∣=1.52%
∣Δσ∣=2.10%
∣Δμ∣=1.27%
∣Δσ∣=1.36%∣Δμ∣=0.62%
∣Δσ∣=0.65%∣Δμ∣=0.76%
∣Δσ∣=0.74%Figure 2: Representative CLEAR test set reconstructions
usingtheCAEtrainedwithbeamimagepriorsyn... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 2 |
bca67e30-4bc9-49cc-966a-77d2a2fea55f | row: normalizedcentroidandwidtherrorhistogramswith
Gaussianfits. Bottomrow: predictedversusreferencecen-
troidandwidthvalues,withdasheddiagonallinesindicating
ideal agreement.
The test set statistics for the model trained with beam
image prior synthesis are shown in Fig. 3. The width errors
are slightlyshifted towardpo... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 2 |
647875af-5859-4dbb-becb-fadbf6edcf76 | basis methods, beam image prior synthesis gives the best
& '%
* & + ! % #% "!
) *+!) *#*#-!
) *+!) *#* *+!) * &
*#*
) *+!) *#* *+!) * &
*#*'+) #&#&"
+
'++'%$ !$+) #&#&" + *',) !)',($ !$%' !$+.(!
')% $#*!
& '%... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 2 |
a1e63f11-1995-431d-934c-dd611da6321c | performsworse,likelybecauseitscoefficientmapsdonot
fully match the beam distribution and need further tuning.
CONCLUSION
ThisstudyfirstusedaDMDsetuptovalidatetheposition
basissynthesismethodfortheMMFrelay,andthenextended
the idea to real beam data from the CLEAR beamline. The
results suggest that a raster scan basis ca... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 2 |
6b606281-65c6-49b5-8ee6-02f3ba2ce5a0 | Theremaininglimitationsarethatthesyntheticcoefficient
mapsmustmatchthetargetbeamdistribution,andthatthe
finite Gaussian width of the raster scan beam spot limits the
achievablespatialresolution. Futureworkwillinvestigate
improved SGM tuning with reduced dependence on mea-
sured beam priors, and optical basis calibratio... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 2 |
f52d13b5-d22c-40e1-88f9-d5d85db4a1f9 | ties at the University of Liverpool, UK.
REFERENCES
[1]F.Roncarolo etal.,“ReviewofCERNbeaminstrumentation
for fixed target experiments”, in Proc. IPAC’23 , Venice, Italy,
May2023, pp. 4625–4628.
doi:1/zero.alt4.18429/JACoW-IPAC2/zero.alt423-THPL/zero.alt48/zero.alt417th International Particle Accelerator Conference, De... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 2 |
a9854183-a236-4941-9f3a-c5dd8aa2d689 | [2]G. Trad and S. Burger, “Artificial Intelligence-Assisted Beam
Distribution Imaging Using a Single Multimode Fiber at
CERN”, in Proc. IPAC’22 , Bangkok, Thailand, Jun. 2022,
pp. 339–342.
doi:1/zero.alt4.18429/JACoW-IPAC2/zero.alt422-MOPOPT/zero.alt441
[3]Q. Xuet al., “Estimation of beam transverse parameters
through ... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 3 |
4cadba64-627e-4dc0-87bc-0340559cbdf7 | opticaltransmissionmatrixofamultimodefibre”, LightSci.Appl., vol. 10, Art. no. 88, 2021.
doi:1/zero.alt4.1/zero.alt438/s41377-/zero.alt421-/zero.alt4/zero.alt4514-9
[6]B. Rahmani, D. Loterie, G. Konstantinou, D. Psaltis, and C.
Moser, “Multimode optical fiber transmission with a deep
learning network”, Light Sci. Appl.... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 3 |
ebde1986-40bb-4193-a1a6-1d203b099c21 | Welsch, “Development of radiation-tolerant beam imaging via
multimode fiber and synthetic data-driven machine learning”,
Phys.Rev.Accel.Beams ,vol.29, no. 3, p. 032801, 2026.
doi:1/zero.alt4.11/zero.alt43/wddb-m37l17th International Particle Accelerator Conference, Deauville, France JACoW Publishing
ISBN: 978-3-95450-2... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6024.pdf | 3 |
c5ded7f6-c087-446b-a730-32640b131bfd | IMPROVING PEAK PREDICTION IN MEBT SURROGATE MODELS
USING FOURIER FEATURE MAPPING∗
J. Cha†1, X. Liu1, J. Thiyagalingam1, K. Baker2
1STFC, Swindon, United Kingdom
2ISIS Neutron and Muon Source, Didcot, United Kingdom
Abstract
Accurate prediction of localised beam features such as
peak loss and rapid envelope oscillations... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 0 |
a34745ba-07cd-4684-8c8a-e9a78cd10ffd | frequency variations more effectively. Wetrain twosurro-
gate models on MEBT simulation data, one using Fourier
feature mapping and the other without, and compare their
performanceincapturingpeakregions. Resultsshowthat
Fourier feature mapping significantly improves peak predic-
tionaccuracywhilemaintainingcomparablepe... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 0 |
779a8719-d0e3-4fd9-a2df-0531c3e51826 | experience. While particle tracking simulations provide ac-
curate predictions during the design phase, their compu-
tational cost limits their use in real-time operation, where
beam characterisation is further constrained by limited and
often disruptive diagnostics.
Beamlossisakeyoperationaldiagnostic,providingan
indi... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 0 |
1a26eabf-8e2a-4988-9789-410fdf3c3453 | applicationssuchasonlineoptimisationandcontrol. Neural
network-based surrogate models have therefore been
∗ThisworkwassupportedbytheSTFCFacilitiesFundthroughtheAda
LovelaceCentre.
†jaehoon.cha@stfc.ac.ukadopted to approximate these simulations with reduced
computational cost [ 5,6].
Despite their success as universal f... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 0 |
ed84f65c-7059-4059-872f-bac228497a97 | spatialvariations. Itisevaluatedonasurrogatemodelofthe
MediumEnergyBeamTransport(MEBT)attheISISNeu-
tronandMuonSource[ 9],demonstratingimprovedpeakpre-
dictionaccuracy whilemaintainingperformanceon global
beam trends.
MEBT MODEL
To investigate the impact of Fourier feature mapping
on high-frequency beam prediction, we ... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 0 |
121d9c21-d926-4c42-a82e-a4c4c30e6156 | to 96%.
TheMEBTconsistsofeightquadrupolesfortransverse
focusingandfourrebunchingcavitiesforlongitudinalcon-
trol. The beam dynamics depend on the field strengths of
theseelementsandtheincomingbeamconditions,including
beam size, position, and current.
Beam transport is simulated using the ASTRA particle
trackingcode[ 4]... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 0 |
cea1a678-6db4-4ab2-b055-e306cb1ae497 | tive quantities relevant for machine protection and optimisa-
tion. As the MEBT is not yet operational, simulation-based
modelling provides the onlymeans to study this behaviour,
motivating the development of fast surrogate models.17th International Particle Accelerator Conference, Deauville, France JACoW Publishing
IS... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 0 |
270b85d6-9bdc-4416-af4c-f1b41f9c08cc | Atotalof10,000simulationsweregeneratedbysampling
theinputparameter spacearound the designsettings of the
MEBT. Among these, 4,193 were discarded due to unstable
beam transport with excessive loss, leaving 5,807 simula-
tions. Thedatasetwassplitintotraining,validation,andtest
sets of size 4,065, 871, and 871 respectivel... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 1 |
3ac7eec6-daa6-49d5-a332-4909a896f529 | 𝑗𝑣)]𝑚
𝑗=1(1)
where𝑚is the embedding dimension, 𝑏𝑗defines the fre-
quency components, and 𝑎𝑗= 1/𝑗𝑝controls the ampli-
tude scaling, with 𝑝determining the contribution of higher-
frequency components.
This transformation projects the input into a higher-
dimensionalspaceofsinusoidalbasisfunctions,enablingim-
... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 1 |
a62a3516-0877-492a-adb5-7c8db1795b13 | rametersandincomingbeamconditions,togetherwiththe
spatialcoordinatealongthebeamline. IntheFourierfeature
model,therawspatialcoordinateisreplacedbyitsFourier
feature embedding defined in Eq. ( 1).
The network contains three hidden layers with 128,
256, and 128 neurons, respectively, each followed by a
LeakyReLU activati... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 1 |
38b3f3fd-0a3e-4aed-bed0-fa13d495a719 | differing only in how the spatial coordinate is represented.
Training and Data Processing
Allinputvariablesarestandardisedusingthemeanand
standarddeviationcomputedfromthetrainingset. Forthe
outputs, the first 10 values are standardised in the sameway, while the beam loss and reconstructed cumulative loss
remain in thei... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 1 |
75b91efa-4edd-4a03-b445-e3554e78f165 | particular, we focus on the reconstruction of sparse beam
lossprofiles,wherestandardneuralnetworkstendtosmooth
or suppress sharp peaks. Models without Fourier features
arecomparedagainstmodelswithembeddingdimensions
𝑚 = 25,50, and100.
Quantitative and Qualitative Comparison
The quantitative performance of the models i... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 1 |
4570c3a3-d9a5-4bba-a563-b4227c3223b4 | differencesaremostevidentinpeakreconstructionandcu-
mulative errors. Although the baseline model achieves a
moderate MSE, this metric is misleading due to the spar-
sityofthelossprofile,wheremostvaluesareclosetozero.
AsshowninFig. 1foraregioncontaininghigh-frequency
variations ( 𝑧 = 75–175), the baseline model fails t... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 1 |
336a5806-37df-40fb-beca-d4d7cbf31149 | machine optimisation and operation.
Embedding Dimension Ablation
Theeffectofembeddingdimensiononpeakreconstruc-
tion error is summarised in Fig. 2. A clear monotonic de-
crease is observed as 𝑚increases, indicating that larger17th International Particle Accelerator Conference, Deauville, France JACoW Publishing
ISBN: ... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 1 |
561431ed-63ff-474a-bdbf-d53c1403af50 | Figure1: Comparisonofpredictedsparselossprofilesfor
representativetestexamples. Thebaselinemodelpredictsan
almostconstantresponse, whileFourierfeatureembeddings
enable accurate reconstruction of sharp loss peaks.
Fourier embeddings improve the reconstruction of localised
peaks and mitigate the low-frequency bias of the... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 2 |
68d382f4-a9f9-4b76-a08e-d4e6d13e1303 | tures in beam profiles. While standard neural networks
trained with mean squared error tend to produce overlysmooth outputs, Fourier features enable accurate recon-
structionofsharpvariations. Performanceconsistentlyim-
proveswith increasing embedding dimension, with FF-100
achieving the best results across all metrics... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 2 |
894b692d-f286-436e-a5c5-a0bf54255c0c | library”,Nucl.Instrum.MethodsPhys.Res.A ,vol.558,no.1,
pp. 356–359, 2006. doi:1/zero.alt4.1/zero.alt416/j.nima.2/zero.alt4/zero.alt45.11./zero.alt4/zero.alt41
[3]M.Borland,“elegant:aflexibleSDDS-compliantcodeforac-
celeratorsimulation”,ArgonneNationalLaboratory,Lemont,
IL, USA, Rep. LS-287, 2000. doi:1/zero.alt4.2172/7... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 2 |
fa2edee4-443e-4901-8016-46983e079236 | p. 145001, 2023.
doi:1/zero.alt4.11/zero.alt43/PhysRevLett.13/zero.alt4.145/zero.alt4/zero.alt41
[7]R.Basri,M.Galun,A.Geifman,D.Jacobs,Y.Kasten,and
S.Kritchman,“Frequencybiasinneuralnetworksforinput
of non-uniform density”, in Proc. ICML’20 , Vienna, Austria,
pp. 685–694, Jul. 2020.
[8]N.Rahaman etal.,“On the spectral ... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 2 |
6d842f14-3b03-4f25-800c-c1fab9be2ebc | 2019. doi:1/zero.alt4.1/zero.alt463/1.5127263
[11]C.E.Mayes etal.,“LightsourceUnifiedModelingEnviron-
ment (LUME), a start-to-end simulation ecosystem”,in Proc.
IPAC’21,Campinas, SP,Brazil, pp. 4212–4215, May2021.
doi:1/zero.alt4.18429/JACoW-IPAC2/zero.alt421-THPAB217
[12]M. Tancik et al., “Fourier features let network... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6027.pdf | 2 |
f02a5b5b-db0c-4edb-bfac-7a6abab14435 | RATE CAPABILITIES OF RADIATION MONITORS
BASED ON DIAMOND DETECTORS∗
J. Melbinger, M. Bacak, D. Divya, D. Hainz, E. Jericha, TU Wien, Vienna, Austria
T. Fragner, E. Griesmayer, C. Weiss, CIVIDEC Instrumentation, Vienna, Austria
H. Frais-Koelbl, Fachhochschule Wiener Neustadt, Wiener Neustadt, Austria
Abstract
Diamond de... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 0 |
7b60dd65-aa14-4e90-af1d-60d5f9da3eb8 | Low-noise CIVIDEC charge-sensitive spectroscopic ampli-
fiers with FWHM of 10ns,30ns,50ns and 180ns were
compared. Thecorrelationbetweencountratesandthemea-
sured current was studied.
INTRODUCTION
This work presents concurrent counting (AC) and current
(DC)measurementsforadiamonddetectorwitha 500MBq
90Sr𝛽-source. Inco... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 0 |
c469af8f-2e39-4d31-800c-825ab6b48096 | pulses,orstochasticwith ΔtfollowingaPoissondistribution.
MEASUREMENT SETUP
The measurement was conducted at the TRIGA Center
of TU Wien located at the Atominstitut in Vienna, Aus-
tria [1]. A CIVIDEC B1-HV Diamond Beam Loss De-
tector[2]wasused,housinga 4mm×4mm single-crystal
chemical vapour deposition (sCVD) diamond s... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 0 |
442bc35d-1662-4f5a-a561-ac520bfeaa86 | scope mode to check for the occurrence of pile-up events.
Aschematicofthesetupandread-outdiagramisshownin
Fig.1.
∗julian.melbinger@cividec.at
sCVD
90Sr
500 MBqRemote stage 1 mm 95 mm
-400V
AC signal
DC signalFigure 1: Schematic of the measurement setup with the AC
and DC path shown.
The500MBq90Srsourcehasadiameterof 10... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 0 |
df8802c1-b3ff-4552-9a0d-27621e5b5d25 | back.
Figure 2: Alignment of the source placeholder with the
center of the diamond detector in the background.
The distance between the source and the detector was
stepwise varied from d = 1mmto95mm , with30sAC and
DC measurements at each position.
AC Coupled Amplifiers
Fourdifferentcapacitativecoupledchargesensitiveam... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 0 |
20e59cbb-4986-4ca0-8bad-799d09fd93d9 | ISBN: 978-3-95450-252-3 ISSN: 2673-5350 doi: 10.18429/JACoW-IPAC2026-WEP6030
MC6.T03: Beam Diagnostics and Instrumentation 2647WEP: Poster session WEP6030
Content from this work may be used under the terms of the CC BY 4.0 licence (© 2026). Any distribution of this work must maintain attribution to the author(s), title... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 0 |
03919a73-fdb8-4faa-a2c3-ff311b150652 | fiershaveanundershootof 1%. Thisaccumulateswithan
increasingpulserate,resultinginashiftofthebaselineto-
wardsincreasinglynegativevalues. Whenusinganabsolute
triggerthisincreasestheeffectivetriggerlevel,resultingin
anunderestimationofthecountrate. ThePBWoftheam-
plifiers is defined as the time from the start of the puls... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 1 |
9658f0eb-8dfb-4f40-b624-811fd85c9575 | tothesignalwithinthePBWoftheamplifier,apile-upevent
occurs[7]. Thisresultsinanunderestimationoftheparticle
flux because of missed pulses.
DC Coupled Amplifiers
A CIVIDEC C8 Electrometer Amplifier [ 8] was used for
the DC measurements. It has a dynamic range from 1pA
to10mAwithabandwidthofupto 10kHz,dependingon
thesensi... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 1 |
743c66fd-2efe-4c19-994d-b656194b18b4 | 𝑃𝐵𝑊.
RESULTS
Figure3shows the measured current as function of the
distance between the90Sr source and the diamond detector
from the real-time measurements. The measurement shows
thevaryingcurrent,andconsequentlycount-rateasfunction
of the distance to the90Sr source.
Oscilloscope Measurement
Before taking real-time m... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 1 |
4bb2c5fe-a071-47eb-9351-76b2060a7cea | involving three pulses. In a real-time measurement, one
signalisdetectedforsuchapile-upevent,insteadofthree,
withthebenefitofprovidingcontinuousdatawithoutpost
processing.
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Time [ns]50
050100150200250300Amplitude [mV]
© CIVIDEC4000 4200 4400 4600050100
PBWCxL-50
Shift... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 1 |
69d039da-4d7c-4b51-80d6-d9bf9e3526c0 | Thus,small Δtarealsopresentatlowparticlefluences. The
resultingmaximumfrequency 𝑓𝑃todetectapercentageof
thenumberofpulses( 𝜂)iscalculatedforagivenPBWusing
Eq. (1):
𝑓𝑃= −ln(𝜂)
𝑃𝐵𝑊. (1)
ForaC6-10amplifierwithaPBWof 30nsthedeterminis-
tic count rate is 𝑓𝐷=33.3MHz . For𝜂=0.9the correspond-
ing maximum Poisson c... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 1 |
76a790f0-8dfe-408d-8111-f6d025fbf529 | ≈90%ofpulsesareseparatedbyatleastthePBW.Figure. 6
shows the Δt distribution for a measurement at d = 95mm.
109
108
107
106
105
104
103
t [s]
100101102103104105106107Counts
© CIVIDECDAQ
CVD diamondC6-10CxL-30CxL-50CxL-180Measurement
Poisson
Figure 6: The Δt distribution with the PBW of the four
amplifiers shown.
The dis... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 2 |
b5253dee-2789-4d1e-bf15-754b2e785a81 | Current [nA]01234567Count rate [MHz]
© CIVIDECC6-10 | fP
CxL-30 | fP
CxL-50 | fP
CxL-180 | fPC6-10
CxL-30
CxL-50
CxL-180
Figure 7: The deadtime corrected count rate of each ampli-
fier as function of the measured current.
The count rate as function of the measured current shows
a non-linear behaviour because of the ele... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 2 |
7af3fed6-7245-4648-870a-4da4ce8edc4c | ducted, using a diamond detector and a moveable 500MBq90Sr beta source. Four CIVIDEC charge sensitive ampli-
fiers with a FWHM between 10ns and 180ns were used
toinvestigateanddiscusstheirrespectivemaximumcount
rate for deterministic and stochastic measurements. Table 2
summarises the maximum count rates.
Table 2: The ... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 2 |
20e7f0a6-6bef-4377-95ef-ce0c2c53115a | approachprovidesreliableparticlefluxmeasurementsand
overcomes the respective limitations from the counting and
current measurement.
REFERENCES
[1]TRIGA Center | TU Wien, https://www.tuwien.at/en/
research/facilities/trigacenter
[2]CIVIDECInstrumentation- DiamondBeamLossDetector,
https://cividec.at/detectors-B1-HV.html
... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 2 |
5680d543-54f0-49a3-99f5-369757753f2a | [7]M. Arkani and G. Raisali, “Measurement of dead time by
timeintervaldistributionmethod”, Nucl.Instrum.Methods
Phys. Res. A , vol. 774, pp. 151–158, Feb. 2015.
doi:1/zero.alt4.1/zero.alt416/j.nima.2/zero.alt414.11./zero.alt469
[8]CIVIDEC Instrumentation - C8 Electrometer Amplifier,
https://cividec.at/electronics-C8-Na... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 2 |
7d81c80a-7f95-4ce0-bab1-b00f77515a32 | ISBN: 978-3-95450-252-3 ISSN: 2673-5350 doi: 10.18429/JACoW-IPAC2026-WEP6030
MC6.T03: Beam Diagnostics and Instrumentation 2649WEP: Poster session WEP6030
Content from this work may be used under the terms of the CC BY 4.0 licence (© 2026). Any distribution of this work must maintain attribution to the author(s), title... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6030.pdf | 2 |
a7d1f1cd-ae93-4a15-bb69-61e9404cd696 | EXPERIMENTAL EVALUATION OF A DIGITIZER DESIGNED FOR ACCT
BEAM CURRENT TRANSFORMERS
D. Coves†, F. Stulle, E. Touzain, Bergoz Instrumentation, Saint -Genis -Pouilly, France
P. Peier , F. Frei , METAS, Wabern bei Bern, Switzerland
L. Soutter, C. Bailat, CHUV , Lausanne, Switzerland
Abstract
To integrate the si... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 0 |
ac2b8b5e-9e8f-4841-9bb5-561207201b7d | waveform acquisition s preserving the quality of the analog
signal . Additionally, it allows controlling the ACCT con-
figuration. Communication is implemented over Ethernet
through a simple command protocol and a n EPICS soft
IOC is provided.
The digitizer was tested with two ACCT s installed at the
CHUV Oria... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 0 |
fc98d75f-7e31-4a3b-aa17-c0d231ae3c7c | range version or with three digitally selectable measure-
ment ranges . The current waveform is provided as an ana-
log output signal.
Various electronic s standards are used for acquisition
and control across particle accelerators, including VME-
bus, CompactPCI , and MicroTCA, each providing high
performance bu... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 0 |
39e9b3d2-b4d7-4f9b-8bf5-a81f8a1deab1 | ACCT electronics range control and on- board data pro-
cessing, including real -time conversion of voltage wave-
forms into beam current and pulse charge. This consoli-
dates functionalities typically distributed across multiple
hardware modules into a single device . Modular Digital Solution
The Modular Digital Solu... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 0 |
441a495c-e965-4c51-ba0b-95e514ee7d10 | Figure 1: MDS -ACCT in i ts tabletop chassis.
The MDS -ACCT is based on a 16 -bit analog -to-digital
converter (ADC) operating at 160 MS/s. The digitized
samples are streamed to the programmable logic (PL) of a Zynq UltraScale+ system -on-chip (SoC), where they are
processed and stored in multiple circular buffer... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 0 |
e5aebcfa-7e99-4890-b988-2189bc3ed98c | trigger number ...), which are transmitted over Gigabit
Ethernet.
In addition, long- term sustainability is enhanced by the
ability to update both the PL and the PS, as the firmware is
stored on a MicroSD card that can be replaced on -site. This
therefore allows for the future implementation of addi-
tional funct... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 0 |
06652a77-791f-41f6-a4e6-e3c10da0379e | ISBN: 978-3-95450-252-3 ISSN: 2673-5350 doi: 10.18429/JACoW-IPAC2026-WEP6031
2650 MC6.T03: Beam Diagnostics and InstrumentationWEP6031 WEP: Poster session
Content from this work may be used under the terms of the CC BY 4.0 licence (© 2026). Any distribution of this work must maintain attribution to the author(s), title... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 0 |
7d39555c-6ff0-4314-b589-4cc1d3853666 | with the eRT6, two ACCTs have been installed at the exit
of the accelerator. Both use the same sensor type (ACCT -
S-055), but electronics with different measurement ranges ,
+1 V / 10 mA and +1 V / 300 mA (ACCT -E-RM-10mA -
50R-3M and ACCT -E-RM-300mA -50R-3M).
Output signals of the ACCTs electronics are digit... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 1 |
37781660-86a7-43f4-a5d0-4c695db6ebd8 | ates a 22 MeV Microtron type electron accelerator from
Scanditronix [5] for research and reference measurements
in radiotherapy. The accelerator produces electron pulses
of 3 µs length and pulse currents of up to 100 mA. Beam
energy reaches from 5 to 22 MeV and the repetition rate
can be varied from 1 to 250 H... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 1 |
86f7f467-f3be-4f34-9271-c72554797744 | ics.
The combination of an ACCT and such an oscilloscope
offers impressive flexibility thanks to the provided soft-
ware package and the easy setup. Further, the high band-
width allows for direct visualization of the temporal shape
of the electron pulses. Its draw backs include limited verti-
cal resolution, non -n... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 1 |
d8ca8db5-1a94-4c2f-a22c-c331c488a5a5 | operated by small teams for whom ease of integration and
rapid deployment of digitizers are critical factors.
Qualitative Test C ampaign at CHUV
At the CHUV Oriatron Linac, the signal from the ACCT
electronics was split using a power divider, allowing sim-
ultaneous acquisition with the MDS -ACCT while main-
ta... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 1 |
d0874f62-0d4c-401e-8e01-0549c833f04e | rect system operation . One of the pulses is shown in
Fig. 2, alongside a typical pulse acquired by the
CHUV control system a few weeks before this test
campaign.
2. Progressive increase of the repetition rate up to the Oriatron limit (250 Hz). The MDS -ACCT was capa-
ble of operating at this frequency .
3. Lo... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 1 |
50aff61a-04d2-4afe-a2ba-9a58a1434f72 | the MDS -ACCT at a repetition frequency of 25 Hz. Wave-
forms were time -aligned and offsets were adjusted.
During long acquisition sequences , missing UDP /IP
packets were observed at the receiver side. However, in-spection of the received meta data showed matching trigger
numbering and packet numbering, even i... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 1 |
368dfa05-37ec-4a5a-94c3-96d3fcce3650 | For subsequent tests, the Ethernet cable was replaced
with a better shielded one, and the Python application was
replaced by a compiled C program. After these modifica-
tions, no further packet loss was observed.
Comparative Test C ampaign at METAS
As for the CHUV campaign, the installation of the MDS -
ACCT was... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 2 |
87eb2b56-c75b-44a8-8e21-94662ada6c0e | Figure 4: Same pulse acquired with the METAS PicoScope
(red) and the MDS -ACCT (blue). The lower plot shows a
zoom of the top region .
As shown in Fig. 4, a good agreement is observed be-
tween the two digitizers, with a lower noise level for the
MDS -ACCT, even though both digitizers have a similar
bandwidth. ... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 2 |
5abd10f6-7057-4d99-93d3-b1cf671b5239 | Second, although the PicoScope offers adequate resolu-
tion and signal -to-noise ratio for high -amplitude signals, its
performance becomes more limited for low -amplitude sig-
nals unless the input range is adjusted. In practice, range switching is not always possible because , in some acceler-
ator facilities, une... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 2 |
b9d45976-8599-4612-ad29-0257aa4940b7 | The test bench's software architecture was built around
EPICS , as it has already been done at other facilities [ 6].
Custom Python- based device drivers were programmed for
all test and measurement devices . This approach enabled a
clear separation between hardware interfaces and higher -
level control logic, whil... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 2 |
b05c30ae-e9b7-498a-aed6-b3dcc0d79db8 | urations of 50 Ω, ±1 V full scale and high- impedance ,
±10 V full scale . The SNR was computed from RMS meas-
urements as −20 𝑙𝑙𝑙𝑙𝑙𝑙
10�𝑅𝑅𝑅𝑅𝑅𝑅𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑅𝑅𝑅𝑅𝑅𝑅𝑛𝑛𝑛𝑛𝑠𝑠𝑛𝑛𝑠𝑠𝑠𝑠 ⁄ � where the
noise floor was measured with no input signal and the ref-erence signal corresponds to a f... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 2 |
959bef5f-fd2b-4526-b299-d83936a207ea | ers has been developed and experimentally validated. The
system integrates signal acquisition and processing as well
as measurement range control into a single platform, with
a focus on performance and straightforward integration
into accelerator control environments.
Tests performed at the CHUV Oriatron Linac ... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 2 |
ff72c457-7cf8-4a39-bd3e-0b66cada2b66 | 17th International Particle Accelerator Conference, Deauville, France JACoW Publishing
ISBN: 978-3-95450-252-3 ISSN: 2673-5350 doi: 10.18429/JACoW-IPAC2026-WEP6031
2652 MC6.T03: Beam Diagnostics and InstrumentationWEP6031 WEP: Poster session
Content from this work may be used under the terms of the CC BY 4.0 licence (©... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 2 |
b4bb3e3f-66d4-47f6-a80c-330f005cbe9a | REFERENCES
[1] B. Sharp, “The induction type beam monitor for the PS: He-
reward transformer” , CERN MPS/Int -CO-62-15, 1962 ,
https://cds.cern.ch/record/1068123
[2] N. Hubert, “Overview of Standards for Beam Instrumentation
and Control”, in Proc. of IP AC'12 , Busan, Korea, May 2016,
pp. 3139-3144 .
doi:10... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 3 |
c36f7381-6552-470b-b8f8-f76c8b171b0b | Medical Physics , vol. 49, no. 3, pp. 1831-1838 , 2022.
doi:10.1002/mp.15474
[5] F. Frei, S. Vörös, M . Lüthi and P. Peier , “M22 – the electron
accelerator for calibration and research in radiation dosimetry at METAS ”, Nucl. Instrum. Methods Phys. Res. A, vol. 1077 ,
p. 170588 , 2025 .
doi:10.1016/j.nima.... | D:\Daily_ihep\Conferences\IPAC2026\WEP\WEP6031.pdf | 3 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.