Keyword: network
Paper Title Other Keywords Page
MOPOST001 Performance of Automated Synchrotron Lattice Optimisation Using Genetic Algorithm lattice, dipole, synchrotron, focusing 38
 
  • X. Zhang, S.L. Sheehy
    The University of Melbourne, Melbourne, Victoria, Australia
  • S.L. Sheehy
    ANSTO, Kirrawee DC New South Wales, Australia
 
  Funding: Work supported by Australian Government Research Training Program Scholarship
Rapid advances in superconducting magnets and related accelerator technology opens many unexplored possibilities for future synchrotron designs. We present an efficient method to probe the feasible parameter space of synchrotron lattice configurations. Using this method, we can converge on a suite of optimal solutions with multiple optimisation objectives. It is a general method that can be adapted to other lattice design problems with different constraints or optimisation objectives. In this method, we tackle the lattice design problem using a multi-objective genetic algorithm. The problem is encoded by representing the components of each lattice as columns of a matrix. This new method is an improvement over the neural network based approach* in terms of computational resources. We evaluate the performance and limitations of this new method with benchmark results.
*Conference Proceedings IPAC’21, 2021. DOI:10.18429/JACoW-IPAC2021-MOPAB182
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-MOPOST001  
About • Received ※ 19 May 2022 — Revised ※ 13 June 2022 — Accepted ※ 14 June 2022 — Issue date ※ 17 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
MOPOPT002 Improvements on Sirius Beam Stability controls, operation, feedback, experiment 226
 
  • S.R. Marques, M.B. Alves, F.C. Arroyo, M.P. Calcanha, H.F. Canova, B.E. Limeira, L. Liu, R.T. Neuenschwander, A.G.C. Pereira, D.O. Tavares, F.H. de Sá
    LNLS, Campinas, Brazil
  • G.O. Brunheira, A.C.T. Cardoso, R.B. Cardoso, R. Junqueira Leão, L.R. Leão, P.H.S. Martins, Moreira, S.S. Moreira, R. Oliveira Neto, M.G. Siqueira
    CNPEM, Campinas, SP, Brazil
 
  Sirius is a Synchrotron Light Source based on a 3 GeV electron storage ring with 518 meters circumference and 250 pm.rad emittance. The facility is built and operated by the Brazilian Synchrotron Light Laboratory (LNLS), located in the CNPEM campus, in Campinas. A beam stability task force was recently created to identify and mitigate the orbit disturbances at various time scales. This work presents studies regarding ground motion (land subsidence caused by groundwater extraction), improvements in the temperature control of the storage ring (SR) tunnel air conditioning (AC) system, vibration measurements in accelerator components and the efforts concerning the reduction of the power supplies’ ripple. The fast orbit feedback implementation and other future perspectives will also be discussed.  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-MOPOPT002  
About • Received ※ 08 June 2022 — Revised ※ 14 June 2022 — Accepted ※ 15 June 2022 — Issue date ※ 17 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
MOPOPT041 Artificial Intelligence-Assisted Beam Distribution Imaging Using a Single Multimode Fiber at CERN experiment, simulation, coupling, detector 339
 
  • G. Trad, S. Burger
    CERN, Meyrin, Switzerland
 
  In the framework of developing radiation tolerant imaging detectors for transverse beam diagnostics, the use of machine learning powered imaging using optical fibers is explored for the first time at CERN. This paper presents the pioneering work of using neural networks to reconstruct the scintillating screen beam image transported from a harsh radioactive environment over a single, large-core, multimode, optical fiber. Profiting from generative modeling used in image-to-image translation, conditional adversarial networks have been trained to translate the output plane of the fiber, imaged on a CMOS camera, into the beam image imprinted on the scintillating screen. Theoretical aspects, covering the development of the dataset via geometric optics simulations, modeling the image propagation in a simplified model of an optical fiber, and its use for training the network are discussed. Finally, the experimental setups, both in the laboratory and at the CLEAR facility at CERN, used to validate the technique and evaluate its potential are highlighted.  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-MOPOPT041  
About • Received ※ 08 June 2022 — Revised ※ 14 June 2022 — Accepted ※ 15 June 2022 — Issue date ※ 19 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
MOPOPT057 Updates in Efforts to Data Science Enabled MeV Ultrafast Electron Diffraction System electron, gun, laser, experiment 397
 
  • S. Biedron, T.B. Bolin, M. Martínez-Ramón, S.I. Sosa Guitron
    UNM-ECE, Albuquerque, USA
  • M. Babzien, M.G. Fedurin, J.J. Li, M.A. Palmer
    BNL, Upton, New York, USA
  • S. Biedron
    UNM-ME, Albuquerque, New Mexico, USA
  • D. Martin, M.E. Papka
    ANL, Lemont, Illinois, USA
 
  Funding: Work supported by DOEs EPSCoR award DE-SC0021365, used resources of the Brookhaven National Laboratory’s Accelerator Test Facility and of the Argonne Leadership Computing Facility.
MeV ultrafast electron diffraction (MUED) is a pump-probe characterization technique to study ultrafast phenomena in materials with high temporal and spatial resolution. This complex instrument can be advanced into a turn-key, high-throughput tool with the aid of machine learning (ML) mechanisms and high-performance computing. The MUED instrument at the Accelerator Test Facility in Brookhaven National Laboratory was employed to test different ML approaches for both data analysis and control. We characterized different materials using MUED, mainly polycrystalline gold and single crystal Ta2NiS5. Diffraction patterns were acquired in single shot mode and convolutional neural network autoenconder models were evaluated for noise reduction and the reconstruction error was studied to identify anomalous diffraction patterns. Electron beam energy jitter was analyzed from single shot diffraction patterns to be used as a novel diagnostic tool. The MUED beamline was also simulated using VSim to construct a surrogate model for control of beam shape and energy. Progress towards ML-based controls leveraging off Argonne Leadership Computing Facility resources will also be presented.
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-MOPOPT057  
About • Received ※ 02 July 2022 — Accepted ※ 26 June 2022 — Issue date ※ 08 July 2022  
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
MOPOPT067 Electron Beam Phase Space Reconstruction From a Gas Sheet Diagnostic simulation, electron, diagnostics, experiment 414
 
  • N.M. Cook, A. Diaw, C.C. Hall
    RadiaSoft LLC, Boulder, Colorado, USA
  • G. Andonian
    RadiaBeam, Santa Monica, California, USA
  • N.P. Norvell
    UCSC, Santa Cruz, California, USA
  • M. Yadav
    The University of Liverpool, Liverpool, United Kingdom
 
  Funding: This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of High Energy Physics under Award Number DE-SC0019717.
Next generation particle accelerators craft increasingly high brightness beams to achieve physics goals for applications ranging from colliders to free electron lasers to studies of nonperturbative QED. Such rigorous requirements on total charge and shape introduce diagnostic challenges for effectively measuring bunch parameters prior to or at interaction points. We report on the simulation and training of a non-destructive beam diagnostic capable of characterizing high intensity charged particle beams. The diagnostic consists of a tailored neutral gas curtain, electrostatic microscope, and high sensitivity camera. An incident electron beam ionizes the gas curtain, while the electrostatic microscope transports generated ions to an imaging screen. Simulations of the ionization and transport process are performed using the Warp code. Then, a neural network is trained to provide accurate estimates of the initial electron beam parameters. We present initial results for a range of beam and gas curtain parameters and comment on extensibility to other beam intensity regimes.

 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-MOPOPT067  
About • Received ※ 08 June 2022 — Accepted ※ 17 June 2022 — Issue date ※ 10 July 2022  
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
MOPOTK038 BPM Analysis with Variational Autoencoders focusing, diagnostics, GPU, optics 543
 
  • C.C. Hall, J.P. Edelen, J.A. Einstein-Curtis, M.C. Kilpatrick
    RadiaSoft LLC, Boulder, Colorado, USA
 
  Funding: This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of High Energy Physics under Award Number DE-SC0021699.
In particle accelerators, beam position monitors (BPMs) are used extensively as a non-intercepting diagnostic. Significant information about beam dynamics can often be extracted from BPM measurements and used to actively tune the accelerator. However, common measurement tools, such as measurements of kicked beams, may become more difficult when very strong nonlinearities are present or when data is very noisy. In this work, we examine the use of variational autoencoders (VAEs) as a technique to extract measurements of the beam from simulated turn-by-turn BPM data. In particular, we show that VAEs may have the possibility to outperform other dimensionality reduction techniques that have historically been used to analyze such data. When the data collection period is limited, or the data is noisy, VAEs may offer significant advantages.
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-MOPOTK038  
About • Received ※ 09 June 2022 — Revised ※ 13 June 2022 — Accepted ※ 15 June 2022 — Issue date ※ 10 July 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUOXGD2 Wireless IoT in Particle Accelerators: A Proof of Concept with the IoT Radiation Monitor at CERN radiation, monitoring, electron, electronics 772
 
  • S. Danzeca, A.J. Cass, A. Masi, R. Sierra, A. Zimmaro
    CERN, Meyrin, Switzerland
 
  The Internet of Things (IoT) is an ecosystem of web-enabled "smart devices" that integrates sensors and communication hardware to collect, send and act on data acquired from the surrounding environment. Use of the IoT in particle accelerators is not new, with accelerator systems long having been connected to the network to retrieve, send and analyse data. What has been missing is the IoT concept of "smart devices" and above all wireless connectivity. We report here on the advantages of using a particular IoT technology, LoRa, for the deployment of wireless radiation monitors within the CERN particle accelerator complex. IoT Radiation Monitors have been developed as a result of growing demand for radiation measurements where standard infrastructure is not available. As a radiation-tolerant device, the IoT Radiation Monitor is a powerful "eye" for observing the real-time radiation levels in the CERN accelerators. We describe here the technologies used for the project and the various advantages their deployment offers in a particle accelerator environment. This opens up the possibility for the deployment of heterogeneous implementations that would otherwise have been impractical.  
slides icon Slides TUOXGD2 [5.797 MB]  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUOXGD2  
About • Received ※ 07 June 2022 — Revised ※ 11 June 2022 — Accepted ※ 16 June 2022 — Issue date ※ 17 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUOXGD3 6D Phase Space Diagnostics Based on Adaptively Tuned Physics-Informed Generative Convolutional Neural Networks controls, feedback, solenoid, diagnostics 776
 
  • A. Scheinker
    LANL, Los Alamos, New Mexico, USA
  • F.W. Cropp V
    UCLA, Los Angeles, USA
  • D. Filippetto
    LBNL, Berkeley, California, USA
 
  Funding: US Department of Energy, DOE Office of Science Graduate Student Research (SCGSR) contract numbers 89233218CNA000001 and DE-AC02-05CH11231 and by the NSF under Grant No. PHY-1549132.
A physics-informed generative convolutional neural network (CNN)-based 6D phase space diagnostic is presented which generates all 15 unique 2D projections (x,y), (x,y’),…, (z,E) of a charged particle beam’s 6D phase space (x,y,z,x’,y’,E)*. The CNN is trained by supervised learning over a wide range of input beam distributions, accelerator parameters, and the associated 6D beam phase spaces at multiple accelerator locations. The CNN is applied in an un-supervised adaptive manner without knowledge of the input beam distribution or accelerator parameters and is robust to their unknown time variation. Adaptive feedback automatically tunes the low-dimensional latent space of the encoder-decoder CNN to predict the 6D phase space based only on 2D (z,E) longitudinal phase space measurements from a device such as a transverse deflecting RF cavity (TCAV). This method has the potential to provide diagnostics beyond the existing state of the art at many accelerator facilities. Studies are presented for two very different accelerators: the 5-meter-long ultra-fast electron diffraction (UED) HiRES compact accelerator at LBNL and the kilometer long plasma wakefield accelerator FACET-II at SLAC.
*A. Scheinker. "Adaptive machine learning for time-varying systems: low dimensional latent space tuning." Journal of Instrumentation 16.10, 2021: P10008. https://doi.org/10.1088/1748-0221/16/10/P10008
 
slides icon Slides TUOXGD3 [3.112 MB]  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUOXGD3  
About • Received ※ 21 May 2022 — Revised ※ 13 June 2022 — Accepted ※ 15 June 2022 — Issue date ※ 16 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUPOST009 Online Correction of Laser Focal Position Using FPGA-Based ML Models laser, FPGA, controls, electron 857
 
  • J.A. Einstein-Curtis, S.J. Coleman, N.M. Cook, J.P. Edelen
    RadiaSoft LLC, Boulder, Colorado, USA
  • S.K. Barber, C.E. Berger, J. van Tilborg
    LBNL, Berkeley, California, USA
 
  Funding: This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of High Energy Physics under Award Numbers DE-SC 00259037 and DE-AC02-05CH11231.
High repetition-rate, ultrafast laser systems play a critical role in a host of modern scientific and industrial applications. We present a prototype diagnostic and correction scheme for controlling and determining laser focal position at 10 s of Hz rate by utilizing fast wavefront sensor measurements from multiple positions to train a focal position predictor. This predictor is used to determine corrections for actuators along the beamline to provide the desired correction to the focal position on millisecond timescales. Our initial proof-of-principle demonstrations leverage pre-compiled data and pre-trained networks operating ex-situ from the laser system. We then discuss the application of a high-level synthesis framework for generating a low-level hardware description of ML-based correction algorithms on FPGA hardware coupled directly to the beamline. Lastly, we consider the use of remote computing resources, such as the Sirepo scientific framework* , to actively update these correction schemes and deploy models to a production environment.
* M.S. Rakitin et al., "Sirepo: an open-source cloud-based software interface for X-ray source and optics simulations", Journal of Synchrotron Radiation 25, 1877-1892 (Nov 2018).
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUPOST009  
About • Received ※ 20 May 2022 — Revised ※ 14 June 2022 — Accepted ※ 15 June 2022 — Issue date ※ 23 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUPOST015 Commissioning and First Results of an X-Band LLRF System for TEX Test Facility at LNF-INFN LLRF, MMI, klystron, GUI 876
 
  • L. Piersanti, D. Alesini, M. Bellaveglia, S. Bini, B. Buonomo, F. Cardelli, C. Di Giulio, E. Di Pasquale, M. Diomede, L. Faillace, A. Falone, G. Franzini, A. Gallo, G. Giannetti, A. Liedl, D. Moriggi, S. Pioli, S. Quaglia, L. Sabbatini, M. Scampati, G. Scarselletta, A. Stella, S. Tocci, L. Zelinotti
    LNF-INFN, Frascati, Italy
 
  Funding: Latino is a project co-funded by Regione Lazio within POR-FESR 2014-2020 program
In the framework of LATINO project (Laboratory in Advanced Technologies for INnOvation) funded by Lazio regional government, the commissioning of the TEst stand for X-band (TEX) facility has started in 2021 at Frascati National Laboratories of INFN. Born as a collaboration with CERN to test high gradient accelerating structures, during 2022 TEX aims at feeding the first EuPRAXIA@SPARC_LAB X-band structure prototype. During 2021 the commissioning has been successfully carried out up to 48 MW. The power unit is driven by an X-band low level RF system, that employs a commercial S-band (2.856 GHz) Libera digital LLRF (manufactured by Instrumentation Technologies), with an up/down conversion stage and a reference generation and distribution system able to produce coherent frequencies for the American S-band and European X-band (11.994 GHz), both designed and realized at LNF. The performance of the system, with a particular focus on amplitude and phase resolution, together with klystron and driver amplifier jitter measurements, will be reviewed in this paper. Moreover, considerations on its suitability and main limitations in view of EuPRAXIA@SPARC_LAB project will be discussed.
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUPOST015  
About • Received ※ 20 May 2022 — Revised ※ 13 June 2022 — Accepted ※ 17 June 2022 — Issue date ※ 28 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUPOST043 A Novel Method for Detecting Unidentified Falling Object Loss Patterns in the LHC operation, Windows, ECR, machine-protect 953
 
  • L. Coyle, F. Blanc, D. Di Croce, T. Pieloni
    EPFL, Lausanne, Switzerland
  • L. Coyle, A. Lechner, D. Mirarchi, M. Solfaroli Camillocci, J. Wenninger
    CERN, Meyrin, Switzerland
 
  Understanding and mitigating particle losses in the Large Hadron Collider (LHC) is essential for both machine safety and efficient operation. Abnormal loss distributions are telltale signs of abnormal beam behaviour or incorrect machine configuration. By leveraging the advancements made in the field of Machine Learning, a novel data-driven method of detecting anomalous loss distributions during machine operation has been developed. A neural network anomaly detection model was trained to detect Unidentified Falling Object events using stable beam, Beam Loss Monitor (BLM) data acquired during the operation of the LHC. Data-driven models, such as the one presented, could lead to significant improvements in the autonomous labelling of abnormal loss distributions, ultimately bolstering the ever ongoing effort toward improving the understanding and mitigation of these events.  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUPOST043  
About • Received ※ 19 May 2022 — Revised ※ 15 June 2022 — Accepted ※ 16 June 2022 — Issue date ※ 21 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUPOST044 Fortune Telling or Physics Prediction? Deep Learning for On-Line Kicker Temperature Forecasting kicker, operation, simulation, injection 957
 
  • F.M. Velotti, M.J. Barnes, B. Goddard, I. Revuelta
    CERN, Meyrin, Switzerland
 
  The injection kicker system MKP of the Super Proton Synchrotron SPS at CERN is composed of 4 kicker tanks. The MKP-L tank provides additional kick needed to inject 26 GeV Large Hadron Collider LHC 25 ns type beams. This device has been a limiting factor for operation with high intensity, due to the magnet’s broadband beam coupling impedance and consequent beam induced heating. To optimise the usage of the SPS and avoid idle (kicker cooling) time, studies were conducted to develop a recurrent deep learning model that could predict the measured temperature evolution of the MKP-L, using the beam conditions and temperature history as input. In a second stage, the ferrite temperature is also estimated putting together the external temperature predictions from accurate thermo-mechanical simulations of the kicker magnet. In this paper, the methodology is described and details of the neural network architecture used, together with the implementation of an ad-hoc loss function, are given. The results applied to the SPS 2021 operational data are presented.  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUPOST044  
About • Received ※ 06 June 2022 — Revised ※ 14 June 2022 — Accepted ※ 16 June 2022 — Issue date ※ 18 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUPOST050 Liverpool Centre for Doctoral Training for Innovation in Data Intensive Science cathode, simulation, experiment, electron 976
 
  • C.P. Welsch
    Cockcroft Institute, Warrington, Cheshire, United Kingdom
  • C.P. Welsch
    The University of Liverpool, Liverpool, United Kingdom
 
  Funding: This new Center for Doctoral Training has received funding from the UK’s Science and Technology Facilities Council.
The Liverpool center for doctoral training for innovation in data intensive science (LIV. INNO) is an inclusive hub for training three cohorts of students in data intensive science. Starting in October 2022, each year will train about 12 PhD students in applying data skills to address cutting edge research challenges across astrophysics, nuclear, theoretical and particle physics, as well as accelerator science. This framework is expected to provide an ideal basis for driving science and innovation, as well as boosting the employability of the LIV. INNO PhD students. This contribution gives examples of the accelerator science R&D projects in the center. It includes details about research into the optimization of 3D imaging techniques and the characterization of photocathodes for accelerator applications.
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUPOST050  
About • Received ※ 05 June 2022 — Revised ※ 09 June 2022 — Accepted ※ 17 June 2022 — Issue date ※ 06 July 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUPOPT013 Twin Delayed Deep Deterministic Policy Gradient for Free-electron Laser Online Optimization FEL, electron, laser, undulator 1025
 
  • M. Cai, C. Feng, L. Tu, Z.T. Zhao, Z.H. Zhu
    SINAP, Shanghai, People’s Republic of China
  • C. Feng, K.Q. Zhang, Z.T. Zhao
    SSRF, Shanghai, People’s Republic of China
  • D. Gu
    SARI-CAS, Pudong, Shanghai, People’s Republic of China
 
  X-ray free-electron lasers (FEL) have contributed to many frontier applications of nanoscale science which benefit from its extraordinary properties. During FEL commissioning, the beam status optimization especially orbit correction is particularly significant for FEL amplification. For example, the deviation between beam orbit and the magnetic center of undulator can affect the interaction between the electron beam and the FEL pulse. Usually, FEL commissioning requires a lot of effort for multi-dimensional parameters optimization in a time-varying system. Therefore, advanced algorithms are needed to facilitate the commissioning procedure. In this paper, we propose an online method to optimize the FEL power and transverse coherence by using a twin delayed deep deterministic policy gradient (TD3) algorithm. The algorithm exhibits more stable learning convergence and improves learning performance because the overestimation bias of policy gradient methods is suppressed.  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUPOPT013  
About • Received ※ 17 May 2022 — Revised ※ 14 June 2022 — Accepted ※ 15 June 2022 — Issue date ※ 22 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUPOPT058 A Machine Learning Approach to Electron Orbit Control at the 1.5 GeV Synchrotron Light Source DELTA storage-ring, synchrotron, controls, electron 1137
 
  • D. Schirmer
    DELTA, Dortmund, Germany
 
  Machine learning (ML) methods have found their application in a wide range of particle accelerator control tasks. Among other possible use cases, neural networks (NNs) can also be utilized for automated beam position control (orbit correction). ML studies on this topic, which were initially based on simulations, were successfully transferred to real accelerator operation at the 1.5-GeV electron storage ring of the DELTA accelerator facility. For this purpose, classical fully connected multi-layer feed-forward NNs were trained by supervised learning on measured orbit data to apply local and global beam position corrections. The supervised NN training was carried out with various conjugate gradient backpropagation learning algorithms. Afterwards, the ML-based orbit correction performance was compared with a conventional, numerical-based computing method. Here, the ML-based approach showed a competitive orbit correction quality in a fewer number of correction steps.  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUPOPT058  
About • Received ※ 20 May 2022 — Accepted ※ 16 June 2022 — Issue date ※ 25 June 2022  
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUPOPT062 A Data-Driven Anomaly Detection on SRF Cavities at the European XFEL cavity, FEL, SRF, experiment 1152
 
  • A. Sulc, A. Eichler, T. Wilksen
    DESY, Hamburg, Germany
 
  Funding: This work was supported by HamburgX grant LFF-HHX-03 to the Center for Data and Computing in Natural Sciences (CDCS) from the Hamburg Ministry of Science, Research, Equalities and Districts.
The European XFEL is currently operating with hundreds of superconducting radio frequency cavities. To be able to minimize the downtimes, prevention of failures on the SRF cavities is crucial. In this paper, we propose an anomaly detection approach based on a neural network model to predict occurrences of breakdowns on the SRF cavities based on a model trained on historical data. We used our existing anomaly detection infrastructure to get a subset of the stored data labeled as faulty. We experimented with different training losses to maximally profit from the available data and trained a recurrent neural network that can predict a failure from a series of pulses. The proposed model is using a tailored architecture with recurrent neural units and takes into account the sequential nature of the problem which can generalize and predict a variety of failures that we have been experiencing in operation.
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUPOPT062  
About • Received ※ 17 May 2022 — Revised ※ 14 June 2022 — Accepted ※ 15 June 2022 — Issue date ※ 24 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUPOPT070 Surrogate Modelling of the FLUTE Low-Energy Section simulation, gun, electron, controls 1182
 
  • C. Xu, E. Bründermann, A.-S. Müller, A. Santamaria Garcia, J. Schäfer
    KIT, Karlsruhe, Germany
 
  Funding: Supported by the Helmholtz Association (Autonomous Accelerator, ZT-I-PF-5-6) and the DFG-funded Doctoral School "Karlsruhe School of Elementary and Astroparticle Physics: Science and Technology".
Numerical beam dynamics simulations are essential tools in the study and design of particle accelerators, but they can be prohibitively slow for online prediction during operation or for systematic evaluations of new parameter settings. Machine learning-based surrogate models of the accelerator provide much faster predictions of the beam properties and can serve as a virtual diagnostic or to augment data for reinforcement learning training. In this paper, we present the first results on training a surrogate model for the low-energy section at the Ferninfrarot Linac- und Test-Experiment (FLUTE).
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUPOPT070  
About • Received ※ 30 May 2022 — Accepted ※ 15 June 2022 — Issue date ※ 05 July 2022  
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUPOTK055 One Year of Operation of the New Wideband RF System of the Proton Synchrotron Booster MMI, operation, cavity, controls 1344
 
  • G.G. Gnemmi, S. Energico, M. Haase, M.M. Paoluzzi, C. Rossi
    CERN, Meyrin, Switzerland
 
  Within the LHC Injectors Upgrade project, the PS Booster(PSB) has been upgraded. Both the injection (160 MeV) and extraction (2 GeV) energies have been increased, bringing also changes in the injection beam revolution frequency, the maximum revolution frequency, and the beam intensity. To meet the requirements of the High Luminosity LHC a new RF system has been designed, based on the wideband frequency characteristics of Finemet® Magnetic Alloy and solid-state amplifiers. The wideband frequency response (1 MHz to 18 MHz) covers all the required frequency schemes in the PSB, allowing multi-harmonics operation. The system is based on a cellular configuration in which each cell provides a fraction of the total RF voltage. The new RF system has been installed in 3 locations replacing the old systems. The installation has been performed during 2019/2020, while the commissioning started later in 2020 and relevant results for the physics have been already observed. This paper describes the new RF chain, the results achieved and the issues that occurred during this year of operation, together with the changes made to the system to improve performance and reliability.  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUPOTK055  
About • Received ※ 02 June 2022 — Revised ※ 14 June 2022 — Accepted ※ 15 June 2022 — Issue date ※ 28 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUPOTK061 Prospects to Apply Machine Learning to Optimize the Operation of the Crystal Collimation System at the LHC collimation, operation, collider, hadron 1362
 
  • M. D’Andrea, G. Azzopardi, M. Di Castro, E. Matheson, D. Mirarchi, S. Redaelli, G. Valentino
    CERN, Meyrin, Switzerland
  • G. Ricci
    Sapienza University of Rome, Rome, Italy
 
  Funding: Research supported by the HL-LHC project.
Crystal collimation relies on the use of bent crystals to coherently deflect halo particles onto dedicated collimator absorbers. This scheme is planned to be used at the LHC to improve the betatron cleaning efficiency with high-intensity ion beams. Only particles with impinging angles below 2.5 urad relative to the crystalline planes can be efficiently channeled at the LHC nominal top energy of 7 Z TeV. For this reason, crystals must be kept in optimal alignment with respect to the circulating beam envelope to maximize the efficiency of the channeling process. Given the small angular acceptance, achieving optimal channeling conditions is particularly challenging. Furthermore, the different phases of the LHC operational cycle involve important dynamic changes of the local orbit and optics, requiring an optimized control of position and angle of the crystals relative to the beam. To this end, the possibility to apply machine learning to the alignment of the crystals, in a dedicated setup and in standard operation, is considered. In this paper, possible solutions for automatic adaptation to the changing beam parameters are highlighted and plans for the LHC ion runs starting in 2022 are discussed.
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUPOTK061  
About • Received ※ 07 June 2022 — Revised ※ 12 June 2022 — Accepted ※ 21 June 2022 — Issue date ※ 24 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
TUPOMS054 Data Augmentation for Breakdown Prediction in CLIC RF Cavities operation, cavity, experiment, ECR 1553
 
  • H.S. Bovbjerg, M. Shen, Z.H. Tan
    Aalborg University, Aalborg, Denmark
  • A. Apollonio, H.S. Bovbjerg, T. Cartier-Michaud, W.L. Millar, C. Obermair, D. Wollmann
    CERN, Meyrin, Switzerland
  • C. Obermair
    TUG, Graz, Austria
 
  One of the primary limitations on the achievable accelerating gradient in normal-conducting accelerator cavities is the occurrence of vacuum arcs, also known as RF breakdowns. A recent study on experimental data from the CLIC XBOX2 test stand at CERN proposes the use of supervised machine learning methods for predicting RF breakdowns. As RF breakdowns occur relatively infrequently during operation, the majority of the data was instead comprised of non-breakdown pulses. This phenomenon is known in the field of machine learning as class imbalance and is problematic for the training of the models. This paper proposes the use of data augmentation methods to generate synthetic data to counteract this problem. Different data augmentation methods like random transformations and pattern mixing are applied to the experimental data from the XBOX2 test stand, and their efficiency is compared.  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-TUPOMS054  
About • Received ※ 08 June 2022 — Revised ※ 12 June 2022 — Accepted ※ 13 June 2022 — Issue date ※ 15 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
WEINGD1 Industry and Accelerator Science, Technology, and Engineering - the Need to Integrate (Building Bridges) electron, laser, radiation, MMI 1644
 
  • R. Geometrante
    KYMA, Trieste, Italy
  • S. Biedron
    Element Aero, Chicago, USA
  • E. Braidotti
    CAEN ELS srl, Trieste, Italy
  • J.M.A. Priem
    VDL ETG, Eindhoven, The Netherlands
  • J.C. Rugsancharoenphol
    FTI, Bangkok, Thailand
  • S.L. Sheehy
    The University of Melbourne, Melbourne, Victoria, Australia
  • M. Vretenar
    CERN, Meyrin, Switzerland
 
  Abstract  
slides icon Slides WEINGD1 [36.079 MB]  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-WEINGD1  
About • Received ※ 05 July 2022 — Accepted ※ 04 July 2022 — Issue date ※ 05 July 2022  
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
WEPOPT008 Supervised Machine Learning for Local Coupling Sources Detection in the LHC coupling, quadrupole, optics, simulation 1842
 
  • F. Soubelet, T.H.B. Persson, R. Tomás García
    CERN, Meyrin, Switzerland
  • Ö. Apsimon, C.P. Welsch
    Cockcroft Institute, Warrington, Cheshire, United Kingdom
 
  Funding: This research is supported by the LIV. DAT Center for Doctoral Training, STFC and the European Organization for Nuclear Research
Local interaction region (IR) linear coupling in the LHC has been shown to have a negative impact on beam size and luminosity, making its accurate correction for Run 3 and beyond a necessity. In view of determining corrections, supervised machine learning has been applied to the detection of linear coupling sources, showing promising results in simulations. An evaluation of different applied models is given, followed by the presentation of further possible application concepts for linear coupling corrections using machine learning.
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-WEPOPT008  
About • Received ※ 03 June 2022 — Revised ※ 15 June 2022 — Accepted ※ 16 June 2022 — Issue date ※ 29 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
THIYGD1 White Rabbit Based Beam-Synchronous Timing Systems for SHINE timing, FEL, electron, FPGA 2415
 
  • Y.B. Yan, G.H. Chen, Q.W. Xiao, P.X. Yu
    SSRF, Shanghai, People’s Republic of China
  • G.H. Gong
    Tsinghua University, Beijing, People’s Republic of China
  • J.L. Gu, Z.Y. Jiang, L. Zhao
    USTC, Hefei, Anhui, People’s Republic of China
  • Y.M. Ye
    TUB, Beijing, People’s Republic of China
 
  Shanghai HIgh repetition rate XFEL aNd Extreme light facility (SHINE) is under construction. SHINE requires precise distribution and synchronization of the 1.003086 MHz timing signals over a long distance of about 3.1 km. Two prototype systems were developed, both containing three functions: beam-synchronous trigger signal distribution, random-event trigger signal distribution and data exchange between nodes. The frequency of the beam-synchronous trigger signal can be divided according to the accelerator operation mode. Each output pulse can be configured for different fill modes. A prototype system was designed based on a customized clock frequency point (64.197530 MHz). Another prototype system was designed based on the standard White Rabbit protocol. The DDS (Direct Digital Synthesis) and D flip-flops (DFFs) are adopted for RF signal transfer and pulse configuration. The details of the timing system design, laboratory test results will be reported in this paper.  
slides icon Slides THIYGD1 [5.582 MB]  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-THIYGD1  
About • Received ※ 29 May 2022 — Revised ※ 10 June 2022 — Accepted ※ 15 June 2022 — Issue date ※ 17 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
THPOPT008 Beam Orbit Shift Due to BPM Thermal Deformation Using Machine Learning storage-ring, synchrotron, vacuum, feedback 2577
 
  • K.M. Chen, M. Hosaka, F.Y. Wang, G. Wang, Z. Wang, W. Xu
    USTC/NSRL, Hefei, Anhui, People’s Republic of China
  • L. Guo
    Nagoya University, Nagoya, Japan
 
  Stabilizing beam orbit is critical for advanced synchrotron radiation light sources. The beam orbit can be affected by many sources. To maintain a good orbit stability, global orbit feedback systems (OFB) has been widely used. However, the BPM thermal deformation would lead to BPM misreading, which can not be handled by OFB. Usually, extra diagnostics, such as position transducers, is needed to measure the deformation dependency of BPM readings. Here, an alternative approach by using the machine operation historic data, including BPM temperature, insertion device (ID) gaps and corrector currents, is presented. It is demonstrated at Hefei Light Source (HLS). The average orbit shift due to BPM thermal deformation is about 34.5 microns/degree Celsius (horizontal) and 20.0 microns/degree Celsius (vertical).  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-THPOPT008  
About • Received ※ 19 May 2022 — Revised ※ 14 June 2022 — Accepted ※ 15 June 2022 — Issue date ※ 19 June 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
THPOTK061 Machine Learning Approach to Temporal Pulse Shaping for the Photoinjector Laser at CLARA laser, target, experiment, electron 2917
 
  • A.E. Pollard, D.J. Dunning, W.A. Okell, E.W. Snedden
    STFC/DL/ASTeC, Daresbury, Warrington, Cheshire, United Kingdom
 
  The temporal profile of the electron bunch is of critical importance in accelerator areas such as free-electron lasers and novel acceleration. In FELs, it strongly influences factors including efficiency and the profile of the photon pulse generated for user experiments, while in novel acceleration techniques it contributes to enhanced interaction of the witness beam with the driving electric field. Work is in progress at the CLARA facility at Daresbury Laboratory on temporal shaping of the ultraviolet photoinjector laser, using a fused-silica acousto-optic modulator. Generating a user-defined (programmable) time-domain target profile requires finding the corresponding spectral phase configuration of the shaper; this is a non-trivial problem for complex pulse shapes. Physically informed machine learning models have shown great promise in learning complex relationships in physical systems, and so we apply machine learning techniques here to learn the relationships between the spectral phase and the target temporal intensity profiles. Our machine learning model extends the range of available photoinjector laser pulse shapes by allowing users to achieve physically realisable configurations for arbitrary temporal pulse shapes.  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-THPOTK061  
About • Received ※ 30 May 2022 — Revised ※ 15 June 2022 — Accepted ※ 01 July 2022 — Issue date ※ 03 July 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
THPOMS032 Advances in the Optimization of Medical Accelerators proton, medical-accelerators, FEL, detector 3030
 
  • C.P. Welsch
    Cockcroft Institute, Warrington, Cheshire, United Kingdom
  • C.P. Welsch
    The University of Liverpool, Liverpool, United Kingdom
 
  Funding: This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska Curie grant agreement No 675265.
Between 2016 and 2020, 15 Fellows have carried out collaborative research within the 4 M€ Optimization of Medical Accelerators (OMA) EU-funded innovative train-ing network. Based at universities, research and clinical facilities, as well as industry partners in several European countries, the Fellows have successfully developed a range of beam and patient imaging techniques, improved biological and physical models in Monte Carlo codes, and also helped improve the design of existing and future clinical facilities. This contribution presents three selected OMA research highlights: the use of Medipix3 for dosimetry and real-time beam monitoring, studies into the technical challenges for FLASH proton therapy, recognized by the European Journal of Medical Physics’ 2021 Galileo Gali-lei Award, and research into novel monitors for in-vivo dosimetry that emerged on the back of the OMA network.
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-THPOMS032  
About • Received ※ 05 June 2022 — Revised ※ 15 June 2022 — Accepted ※ 16 June 2022 — Issue date ※ 02 July 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
THPOMS048 Challenge Based Innovation "Accelerators for the Environment" FEM, background, HOM 3077
 
  • N. Delerue
    Université Paris-Saclay, CNRS/IN2P3, IJCLab, Orsay, France
  • P. Burrows
    JAI, Oxford, United Kingdom
  • R. Holland, L. Rinolfi
    ESI, Archamps, France
  • E. Métral, M. Vretenar
    CERN, Meyrin, Switzerland
 
  Funding: This project has received funding from the European Union’s Horizon 2020 Research and Innovation programme under Grant Agreement No 101004730.
We present an initiative to foster new ideas about the applications of accelerators to the Environment. Called "Challenge Based Innovation" this initiative will gather four teams each of six master-level students each coming from different academic backgrounds. As part of the EU-funded I.FAST project (Innovation Fostering in Accelerator Science and Technology), they will gather during 10 days in Archamps near CERN to receive high level lectures on accelerators and the environment and to brainstorm on possible new applications of accelerators for the environment. At the end of the gathering, they will present their project at CERN to a jury made of experts.
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-THPOMS048  
About • Received ※ 09 June 2022 — Revised ※ 10 June 2022 — Accepted ※ 20 June 2022 — Issue date ※ 01 July 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)