ICALEPCS2023 - List of Keywords (simulation)

Paper	Title	Other Keywords	Page
MO3AO06	Energy Consumption Optimisation by Using Advanced Control Algorithms	controls, operation, PLC, MMI	145
	F. Ghawash, E. Blanco Viñuela, B. Schofield CERN, Meyrin, Switzerland
	Large industries operate energy-intensive equipment and energy efficiency is an important objective when trying to optimize the final energy consumption. CERN utilizes a large amount of electrical energy to run its accelerators, detectors and test facilities, with a total yearly consumption of 1.3 TWh and peaks of about 200 MW. Final energy consumption reduction can be achieved by dedicated technical solutions and advanced automation technologies, especially those based on optimization algorithms, have revealed a crucial role not only in keeping the processes within required safety and operational conditions but also in incorporating financial factors. MBPC (Model-Based Predictive Control) is a feedback control algorithm which can naturally integrate the capability of achieving reduced energy consumption when including economic factors in the optimization formulation. This paper reports on the experience gathered when applying non-linear MBPC to some of the contributors to the electricity bill at CERN: the cooling and ventilation plants (i.e. cooling towers, chillers, and air handling units). Simulation results with cooling towers showed significant performance improvements and energy savings close to 20% over conventional heuristic solutions. The control problem formulation, the control strategy validation using a digital twin and the initial results in a real industrial plant are reported together with the experience gained implementing the algorithm in industrial controllers.
	Slides MO3AO06 [3.101 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-MO3AO06
About •	Received ※ 04 October 2023 — Revised ※ 09 October 2023 — Accepted ※ 14 November 2023 — Issued ※ 29 November 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

MO4AO03	The DESY Open Source FPGA Framework	framework, FPGA, hardware, embedded	222
	Ł. Butkowski, A. Bellandi, M. Büchler, B. Dursun, C. Gümüş, N. Omidsajedi, K. Schulz DESY, Hamburg, Germany
	Modern FPGA firmware development involves integrating various intellectual properties (IP), modules written in hardware description languages (HDL), high-level synthesis (HLS), and software/hardware CPUs with embedded Linux or bare-metal applications. This process may involve multiple tools from the same or different vendors, making it complex and challenging. Additionally, scientific institutions such as DESY require long-term maintenance and reproducibility for designs that may involve multiple developers, further complicating the process. To address these challenges, we have developed an open-source FPGA firmware framework (FWK) at DESY that streamlines development, facilitates collaboration, and reduces complexity. The FWK achieves this by providing an abstraction layer, a defined structure, and guidelines to create big FPGA designs with ease. FWK also generates documentation and address maps necessary for high-level software frameworks like ChimeraTK. This paper presents an overview and the idea of the FWK.
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-MO4AO03
About •	Received ※ 05 October 2023 — Accepted ※ 13 October 2023 — Issued ※ 13 October 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TU1BCO02	Integrating System Knowledge in Unsupervised Anomaly Detection Algorithms for Simulation-Based Failure Prediction of Electronic Circuits	ISOL, electron, monitoring, radiation	249
	F. Waldhauser, H. Boukabache, D. Perrin, S. Roesler CERN, Meyrin, Switzerland M. Dazer Universität Stuttgart, Stuttgart, Germany
	Funding: This work has been sponsored by the Wolfgang Gentner Programme of the German Federal Ministry of Education and Research (grant no. 13E18CHA). Machine learning algorithms enable failure prediction of large-scale, distributed systems using historical time-series datasets. Although unsupervised learning algorithms represent a possibility to detect an evolving variety of anomalies, they do not provide links between detected data events and system failures. Additional system knowledge is required for machine learning algorithms to determine the nature of detected anomalies, which may represent either healthy system behavior or failure precursors. However, knowledge on failure behavior is expensive to obtain and might only be available upon pre-selection of anomalous system states using unsupervised algorithms. Moreover, system knowledge obtained from evaluation of system states needs to be appropriately provided to the algorithms to enable performance improvements. In this paper, we will present an approach to efficiently configure the integration of system knowledge into unsupervised anomaly detection algorithms for failure prediction. The methodology is based on simulations of failure modes of electronic circuits. Triggering system failures based on synthetically generated failure behaviors enables analysis of the detectability of failures and generation of different types of datasets containing system knowledge. In this way, the requirements for type and extend of system knowledge from different sources can be determined, and suitable algorithms allowing the integration of additional data can be identified.
	Slides TU1BCO02 [2.541 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TU1BCO02
About •	Received ※ 02 October 2023 — Accepted ※ 12 October 2023 — Issued ※ 25 October 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TU1BCO04	Laser Focal Position Correction Using FPGA-Based ML Models	controls, laser, network, FPGA	262
	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 Number DE-SC 00259037. High repetition-rate, ultrafast laser systems play a critical role in a host of modern scientific and industrial applications. We present a diagnostic and correction scheme for controlling and determining laser focal position by utilizing fast wavefront sensor measurements from multiple positions to train a focal position predictor. This predictor and additional control algorithms have been integrated into a unified control interface and FPGA-based controller on beamlines at the Bella facility at LBNL. An optics section is adjusted online to provide the desired correction to the focal position on millisecond timescales by determining corrections for an actuator in a telescope section along the beamline. Our initial proof-of-principle demonstrations leveraged pre-compiled data and pre-trained networks operating ex-situ from the laser system. A framework for generating a low-level hardware description of ML-based correction algorithms on FPGA hardware was coupled directly to the beamline using the AMD Xilinx Vitis AI toolchain in conjunction with deployment scripts. 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 Radiation25, 1877-1892 (Nov 2018).
	Slides TU1BCO04 [1.876 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TU1BCO04
About •	Received ※ 06 October 2023 — Accepted ※ 14 November 2023 — Issued ※ 18 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TU2BCO04	Accelerator Systems Cyber Security Activities at SLAC	controls, EPICS, network, operation	292
	G.R. White, A.L. Edelen SLAC, Menlo Park, California, USA
	Funding: Work supported in part by the U.S. Department of Energy under contract number DE-AC02-76SF00515. We describe four cyber security related activities of SLAC and collaborations. First, from a broad review of accelerator computing cyber and mission reliability, our analysis method, findings and outcomes. Second, lab-wide and accelerator penetration testing, in particular methods to control, coordinate, and trap, potentially hazardous scans. Third, a summary gap analysis of recent US regulatory orders from common practice at accelerators, and our plans to address these in collaboration with the US Dept. of Energy. Finally, summary attack vectors of EPICS, and technical plans to add authentication and encryption to EPICS itself.
	Slides TU2BCO04 [1.677 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TU2BCO04
About •	Received ※ 04 October 2023 — Revised ※ 13 October 2023 — Accepted ※ 15 November 2023 — Issued ※ 17 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUMBCMO13	Applications of Artificial Intelligence in Laser Accelerator Control System	laser, target, controls, experiment	372
	F.N. Li, K.C. Chen, Z. Guo, Q.Y. He, C. Lin, Q. Wang, Y. Xia, M.X. Zang PKU, Beijing, People’s Republic of China
	Funding: the National Natural Science Foundation of China (Grants No. 11975037, NO. 61631001 and No. 11921006), and the National Grand Instrument Project (No. 2019YFF01014400 and No. 2019YFF01014404). Ultra-intense laser-plasma interactions can produce TV/m acceleration gradients, making them promising for compact accelerators. Peking University is constructing a proton radiotherapy system prototype based on PW laser accelerators, but transient processes challenge stability control, critical for medical applications. This work demonstrates artificial intelligence’s (AI) application in laser accelerator control systems. To achieve micro-precision alignment between the ultra-intense laser and target, we propose an automated positioning program using the YOLO algorithm. This real-time method employs a convolutional neural network, directly predicting object locations and class probabilities from input images. It enables precise, automatic solid target alignment in about a hundred milliseconds, reducing experimental preparation time. The YOLO algorithm is also integrated into the safety interlocking system for anti-tailing, allowing quick emergency response. The intelligent control system also enables convenient, accurate beam tuning. We developed high-performance virtual accelerator software using "OpenXAL" and GPU-accelerated multi-particle beam transport simulations. The software allows real-time or custom parameter simulations and features control interfaces compatible with optimization algorithms. By designing tailored objective functions, desired beam size and distribution can be achieved in a few iterations.
	Slides TUMBCMO13 [1.162 MB]
	Poster TUMBCMO13 [1.011 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUMBCMO13
About •	Received ※ 04 October 2023 — Revised ※ 12 October 2023 — Accepted ※ 23 November 2023 — Issued ※ 23 November 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUMBCMO14	Initial Test of a Machine Learning Based SRF Cavity Active Resonance Control	cavity, controls, SRF, resonance	379
	F.Y. Wang, J. Cruz SLAC, Menlo Park, California, USA
	We’ll introduce a high precision active motion controller based on machine learning (ML) technology and electric piezo actuator. The controller will be used for SRF cavity active resonance control, where a data-driven model for system motion dynamics will be developed first, and a model predictive controller (MPC) will be built accordingly. Simulation results as well as initial test results with real SRF cavities will be presented in the paper.
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUMBCMO14
About •	Received ※ 03 October 2023 — Revised ※ 14 November 2023 — Accepted ※ 27 November 2023 — Issued ※ 09 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPDP004	System Identification via ARX Model and Control Design for a Granite Bench at Sirius/LNLS	controls, experiment, feedback, acceleration	479
	J.P.S. Furtado, I.E. Santos, T.R. Silva Soares LNLS, Campinas, Brazil
	Modern 4th generation synchrotron facilities demand mechanical systems and hardware capable of fine position control, improving the performance of experiments at the beamlines. In this context, granite benches are widely used to position systems such as optical elements and magnetos, due to its capacity of insulating interferences from the ground. This work aims to identify the transfer function that describes the motion of the granite bench at the EMA Beamline (Extreme conditions Methods of Analysis) and then design the control gains to reach an acceptable motion performance in the simulation environment before embedding the configuration into the real system, followed by the validation at the beamline. This improvement avoids undesired behaviour in the hardware or in the mechanism when designing the controller. The bench, weighting 1.2 tons, is responsible by carrying a coil, weighting 1.8 tons, which objective is to apply a 3 T magnetic field to the sample that receives the beam provided by the electrons accelerator. The system identification method applied in this paper is based on the auto-regressive model with exogenous inputs (ARX). The standard servo control loop of the Omron Delta Tau Power Brick controller and the identified plant were simulated in Simulink in order find the control parameters. This paper shows the results and comparison of the simulations and the final validation of the hardware performance over the real system.
	Poster TUPDP004 [0.720 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUPDP004
About •	Received ※ 06 October 2023 — Accepted ※ 28 November 2023 — Issued ※ 17 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPDP098	Automatic Conditioning of High Voltage Pulsed Magnets	kicker, controls, PLC, vacuum	780
	C.A. Lolliot, M.J. Barnes, N. Magnin, S. Pavis, P. Van Trappen CERN, Meyrin, Switzerland C. Monier INSA Lyon, Villeurbanne, France
	Fast pulsed kicker magnets are used across the various accelerators of CERN complex to inject and extract the beam. These kicker magnets, powered by high voltage pulsed generators and under vacuum, are prone to electrical breakdown during the pulse. To prepare the kicker magnet for reliable operation, or in case an electrical breakdown occurred, a conditioning is necessary: the magnet is pulsed gradually increasing the pulse voltage and length up to a value beyond operational conditions. This is a lengthy process that requires kicker experts on site to manually control the pulse voltage and length, and monitor the vacuum activity. For the start of LHC operation, a first automatic conditioning system was deployed on injection kicker magnet (MKI). Configurable voltage and pulse length ramps are automatically performed by the controller. In case abnormal vacuum activity occurs, the voltage is reduced and then the process continues. Based on this experience, a standardised algorithm has been developed, adding new features such as logarithmic ramp, or simulation of the whole conditioning cycle with test of reaction to vacuum activity. This new automatic conditioning system was deployed on several kicker systems across various CERN accelerators, allowing smoother conditioning, and great reduction on manpower. It also offers the possibility for further automate kicker system operation, starting automatically a magnet conditioning when needed without intervention of kickers experts, similarly as what was deployed for SPS Beam Dump System.
	Poster TUPDP098 [0.328 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUPDP098
About •	Received ※ 06 October 2023 — Revised ※ 21 October 2023 — Accepted ※ 05 December 2023 — Issued ※ 17 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPDP109	Tickit: An Event-Based Multi-Device Simulation Framework	framework, controls, hardware, EPICS	823
	A. Emery, T.M. Cobb, C.A. Forrester, G. O’Donnell DLS, Oxfordshire, United Kingdom
	Tickit is an event-based multi-device simulation framework providing configuration and orchestration of complex simulations. It was developed at Diamond Light Source in order to overcome limitations presented to us by some of our existing hardware simulations. With the Tickit framework, simulations can be addressed with a compositional approach. It allows devices to be simulated individually while still maintaining the interconnected behaviour exhibited by their hardware counterparts. This is achieved by modelling the interactions between devices, such as electronic signals. Devices can be collated into larger simulated systems providing a layer of simulated hardware against which to test the full stack of Data Acquisition and Controls tools. We aim to use this framework to extend the scope and improve the interoperability of our simulations; enabling us to further improve the testing of current systems and providing a preferential platform to assist in development of the new Acquisition and Controls tools.
	Poster TUPDP109 [0.703 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUPDP109
About •	Received ※ 29 September 2023 — Revised ※ 21 October 2023 — Accepted ※ 04 December 2023 — Issued ※ 18 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPDP115	Machine Learning for Compact Industrial Accelerators	cavity, controls, industrial-accelerators, network	846
	J.P. Edelen, J.A. Einstein-Curtis, M.J. Henderson, M.C. Kilpatrick RadiaSoft LLC, Boulder, Colorado, USA J.A. Diaz Cruz, A.L. Edelen SLAC, Menlo Park, California, USA
	Funding: This material is based upon work supported by the DOE Accelerator R&D and Production under Award Number DE-SC0023641. The industrial and medical accelerator industry is an ever-growing field with advancements in accelerator technology enabling its adoption for new applications. As the complexity of industrial accelerators grows so does the need for more sophisticated control systems to regulate their operation. Moreover, the environment for industrial and medical accelerators is often harsh and noisy as opposed to the more controlled environment of a laboratory-based machine. This environment makes control more challenging. Additionally, instrumentation for industrial accelerators is limited making it difficult at times to identify and diagnose problems when they occur. RadiaSoft has partnered with SLAC to develop new machine learning methods for control and anomaly detection for industrial accelerators. Our approach is to develop our methods using simulation models followed by testing on experimental systems. Here we present initial results using simulations of a room temperature s-band system.
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUPDP115
About •	Received ※ 06 October 2023 — Accepted ※ 05 December 2023 — Issued ※ 18 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUSDSC02	Integrating Online Analysis with Experiments to Improve X-Ray Light Source Operations	experiment, interface, real-time, framework	921
	N.M. Cook, E.G. Carlin, J.A. Einstein-Curtis, R. Nagler, R. O’Rourke RadiaSoft LLC, Boulder, Colorado, USA A.M. Barbour, M. Rakitin, L. Wiegart, H. Wijesinghe BNL, Upton, New York, USA
	Funding: This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research under Award Number DE-SC00215553. The design, execution, and analysis of light source experiments requires the use of sophisticated simulation, controls and data management tools. Existing workflows require significant specialization to accommodate specific beamline operations and data pre-processing steps necessary for more intensive analysis. Recent efforts to address these needs at the National Synchrotron Light Source II (NSLS-II) have resulted in the creation of the Bluesky data collection framework, an open-source library for coordinating experimental control and data collection. Bluesky provides high level abstraction of experimental procedures and instrument readouts to encapsulate generic workflows. We present a prototype data analysis platform for integrating data collection with real time analysis at the beamline. Our application leverages Bluesky in combination with a flexible run engine to execute user configurable Python-based analyses with customizable queueing and resource management. We discuss initial demonstrations to support X-ray photon correlation spectroscopy experiments and future efforts to expand the platform’s features.
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUSDSC02
About •	Received ※ 06 October 2023 — Revised ※ 22 October 2023 — Accepted ※ 11 December 2023 — Issued ※ 14 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THMBCMO17	FAIR Data of Physical and Digital Beamlines	experiment, software, controls, GUI	1231
	G. Günther, O. Mannix, O. Ruslan, S. Vadilonga HZB, Berlin, Germany
	Simulations play a crucial role in instrument design, as a digital precursor of a real-world object they contain a comprehensive description of the setup. Unfortunately, this digital representation is often neglected once the real instrument is fully commissioned. To preserve the symbiosis of simulated and real-world instrument beyond commissioning we connect the two worlds through the instrument control software. The instrument control simultaneously starts measurements and simulations, receives feedback from both, and directs (meta)data to a NeXus file - a standard format in photon and neutron science. The instrument section of the produced NeXus file is enriched with detailed simulation parameters where the current state of the instrument is reflected by including real motor positions such as incorporating the actual aperture of a slit system. As a result, the enriched instrument description increases the reusability of experimental data in sense of the FAIR principles. The data is ready to be exploited by machine-learning techniques, such as for predictive maintenance applications as it is possible to perform simulations of a measurement directly from the NeXus file. The realization at the Aquarius beamline * at Bessy II in connection with the Ray-UI simulation software and RayPyNG API * serves as a prototype for a more general application. * https://www.helmholtz-berlin.de/forschung/oe/wi/optik-strahlrohre/projekte/aquarius_en.html https://doi.org/10.1063/1.5084665 * https://pypi.org/project/raypyng
	Slides THMBCMO17 [0.632 MB]
	Poster THMBCMO17 [0.828 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-THMBCMO17
About •	Received ※ 06 October 2023 — Accepted ※ 11 December 2023 — Issued ※ 14 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THMBCMO18	Advancements in Beamline Digital Twin at BESSYII	operation, software, experiment, MMI	1236
	S. Vadilonga, G. Günther, S. Kazarski, R. Ovsyannikov, S.S. Sachse, W. Smith HZB, Berlin, Germany
	This presentation reports on the status of beamline digital twins at BESSY II. To provide a comprehensive beamline simulation experience we have leveraged BESSY II’s x-ray tracing program, RAY-UI[], widely used for beamline design and commissioning and best adapted to the requirements of our soft X-ray source BESSY II. We created a Python API, RayPyNG, capable to convert our library of beamline configuration files produced by RAY-UI into Python objects[]. This allows to embed beamline simulation into Bluesky[], our experimental controls software ecosystem. All optical elements are mapped directly into the Bluesky device abstraction (Ophyd). Thus beamline operators can run simulations and operate real systems by a common interface, allowing to directly compare theory predictions with real-time results[*]. We will discuss the relevance of this digital twin for process tuning in terms of enhanced beamline performance and streamlined operations. We will shortly discuss alternatives to RAY-UI like other software packages and ML/AI surrogate models. []https://doi.org/10.1063/1.5084665 []https://raypyng.readthedocs.io/ []https://doi.org/10.1080/08940886.2019.1608121 [***]https://raypyng-bluesky.readthedocs.io/
	Slides THMBCMO18 [0.333 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-THMBCMO18
About •	Received ※ 06 October 2023 — Accepted ※ 11 December 2023 — Issued ※ 16 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THPDP060	Beam Instrumentation Simulation in Python	electron, electronics, instrumentation, framework	1454
	M. Gonzalez-Berges, D. Alves, A. Boccardi, V. Chariton, I. Degl’Innocenti, S. Jackson, J. Martínez Samblas CERN, Meyrin, Switzerland
	The design of acquisition electronics for particle accelerator systems relies on simulations in various domains. System level simulation frameworks can integrate the results of specific tools with analytical models and stochastic analysis. This allows the designer to estimate the performance of different architectures, compare the results, and ultimately optimize the design. These simulation frameworks are often made of custom scripts for specific designs, which are hard to share or reuse. Adopting a standard interface for modular components can address these issues. Also, providing a graphical interface where these components can be easily configured, connected and the results visualised, eases the creation of simulations. This paper identifies which characteristics ISPy (Instrumentation Simulation in Python) should fulfill as a simulation framework. It subsequently proposes a standard format for signal-processing simulation modules. Existing environments which allow script integration and an intuitive graphical interface have then been evaluated and the KNIME Analytics Platform was the proposed solution. Additionally, the need to handle parameter sweeps for any parameter of the simulation, and the need for a bespoke visualisation tool will be discussed. Python has been chosen for all of these developments due to its flexibility and its wide adoption in the scientific community. The ensuing performance of the tool will also be discussed.
	Poster THPDP060 [2.931 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-THPDP060
About •	Received ※ 07 October 2023 — Accepted ※ 08 December 2023 — Issued ※ 12 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THPDP062	Controls Optimization for Energy Efficient Cooling and Ventilation at CERN	controls, operation, ECR, software	1465
	D. Monteiro, R. Barillère, N. Bunijevac, I. Rühl CERN, Meyrin, Switzerland
	Cooling and air conditioning systems play a vital role for the operation of the accelerators and experimental complex of the European Organization for Nuclear Research (CERN). Without them, critical accelerator machinery would not operate reliably as many machines require a fine controlled thermodynamic environment. These operation conditions come with a significant energy consumption: about 12% (75 GWh) of electricity consumed by the Large Hadron Collider (LHC) during a regular run period is devoted to cooling and air conditioning. To align with global CERN objectives of minimizing its impact on the environment, the Cooling and Ventilation (CV) group, within the Engineering Department (EN), has been developing several initiatives focused on energy savings. A particular effort is led by the automation and controls section which has been looking at how controls and automation strategies can be optimized without requiring costly hardware changes. This paper addresses several projects of this nature, by presenting their methodology and results achieved to date. Some of them are particularly promising as real measurements revealed that electricity consumption was more than halved after implementation. Due to the pertinence of this effort in the current context of energy crisis, the paper also draws a careful reflection on how it is planned to be further pursued to provide more energy-efficient cooling and ventilation services at CERN.
	Poster THPDP062 [7.056 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-THPDP062
About •	Received ※ 05 October 2023 — Accepted ※ 08 December 2023 — Issued ※ 10 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

FR2AO02	A Digital Twin for Neutron Instruments	neutron, software, experiment, detector	1626
	S. Nourbakhsh, Y. Le Goc, P. Mutti ILL, Grenoble, France
	Data from virtual experiments are becoming an extremely valuable asset for research infrastructures in a multitude of aspects and different actors: for instrument scientists to develop and optimise current and future instruments; for training external users in the usage of the instrument control system; for scientists in studying, quantifying and reducing instrumental effects on acquired data. Furthermore large sets of simulated data are also a necessary ingredient for the development of surrogate models for faster and more accurate simulation, reduction and analysis of the data. The development of a digital twin of an instrument can answer such different needs with a single unified approach wrapping in a user-friendly envelop the knowledge about the instrument physical description, the specific of the simulation packages and their interaction, and the high performing computing setup. In this article we will present the general architecture of the digital twin prototype developed at the ILL in the framework of the PANOSC European project in close collaboration with other research facilities (ESS and EuXFel). The communication patterns (based on ZQM) and interaction between the control system (NOMAD), simulation software (McStas), instrument description and configuration, process management (CAMEO) will be detailed. The adoption of FAIR principles for data formats and policies in combination with open-source software make it a sustainable project both for development and maintenance in the mid and long-term.
	Slides FR2AO02 [1.245 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-FR2AO02
About •	Received ※ 31 October 2023 — Revised ※ 02 November 2023 — Accepted ※ 05 December 2023 — Issued ※ 07 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

FR2AO04	A Physics-Based Simulator to Facilitate Reinforcement Learning in the RHIC Accelerator Complex	cavity, controls, booster, diagnostics	1630
	L.K. Nguyen, K.A. Brown, M.R. Costanzo, Y. Gao, M. Harvey, J.P. Jamilkowski, J. Morris, V. Schoefer BNL, Upton, New York, USA
	Funding: Work supported by Brookhaven Science Associates, LLC under Contract No. DE-SC0012704 with the U.S. Department of Energy. The successful use of machine learning (ML) in particle accelerators has greatly expanded in recent years; however, the realities of operations often mean very limited machine availability for ML development, impeding its progress in many cases. This paper presents a framework for exploiting physics-based simulations, coupled with real machine data structure, to facilitate the investigation and implementation of reinforcement learning (RL) algorithms, using the longitudinal bunch-merge process in the Booster and Alternating Gradient Synchrotron (AGS) at Brookhaven National Laboratory (BNL) as examples. Here, an initial fake wall current monitor (WCM) signal is fed through a noisy physics-based model simulating the behavior of bunches in the accelerator under given RF parameters and external perturbations between WCM samples; the resulting output becomes the input for the RL algorithm and subsequent pass through the simulated ring, whose RF parameters have been modified by the RL algorithm. This process continues until an optimal policy for the RF bunch merge gymnastics has been learned for injecting bunches with the required intensity and emittance into the Relativistic Heavy Ion Collider (RHIC), according to the physics model. Robustness of the RL algorithm can be evaluated by introducing other drifts and noisy scenarios before the algorithm is deployed and final optimization occurs in the field.
	Slides FR2AO04 [2.694 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-FR2AO04
About •	Received ※ 04 October 2023 — Accepted ※ 05 December 2023 — Issued ※ 16 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

FR2AO05	Python Library for Simulated Commissioning of Storage-Ring Accelerators	MMI, lattice, storage-ring, closed-orbit	1637
	L. Malina, I.V. Agapov, J. Keil, E.S.H. Musa, B. Veglia DESY, Hamburg, Germany N. Carmignani, L.R. Carver, L. Hoummi, S.M. Liuzzo, T.P. Perron, S.M. White ESRF, Grenoble, France T. Hellert LBNL, Berkeley, California, USA
	Simulations of the commissioning procedure became vital to the storage-ring lattice design process. The achievable tolerances on lattice imperfections, such as equipment misalignments or magnet gradient errors, would, without correction, prohibit reaching the design parameters. We present a Python library which includes an extensive set of error sources in the accelerator lattice and provides a variety of correction algorithms to commission a storage ring. The underlying beam dynamics simulations are performed with pyAT. This project builds upon previous works and expands them in the direction of realistic control room experience and software maintainability. The performance is demonstrated using example commissioning studies, and further development plans are discussed.
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-FR2AO05
About •	Received ※ 06 October 2023 — Revised ※ 27 October 2023 — Accepted ※ 05 December 2023 — Issued ※ 19 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

Paper

Title

Other Keywords

Page

MO3AO06

Energy Consumption Optimisation by Using Advanced Control Algorithms

controls, operation, PLC, MMI

145

F. Ghawash, E. Blanco Viñuela, B. Schofield
CERN, Meyrin, Switzerland

Large industries operate energy-intensive equipment and energy efficiency is an important objective when trying to optimize the final energy consumption. CERN utilizes a large amount of electrical energy to run its accelerators, detectors and test facilities, with a total yearly consumption of 1.3 TWh and peaks of about 200 MW. Final energy consumption reduction can be achieved by dedicated technical solutions and advanced automation technologies, especially those based on optimization algorithms, have revealed a crucial role not only in keeping the processes within required safety and operational conditions but also in incorporating financial factors. MBPC (Model-Based Predictive Control) is a feedback control algorithm which can naturally integrate the capability of achieving reduced energy consumption when including economic factors in the optimization formulation. This paper reports on the experience gathered when applying non-linear MBPC to some of the contributors to the electricity bill at CERN: the cooling and ventilation plants (i.e. cooling towers, chillers, and air handling units). Simulation results with cooling towers showed significant performance improvements and energy savings close to 20% over conventional heuristic solutions. The control problem formulation, the control strategy validation using a digital twin and the initial results in a real industrial plant are reported together with the experience gained implementing the algorithm in industrial controllers.

Slides MO3AO06 [3.101 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-MO3AO06

About •

Received ※ 04 October 2023 — Revised ※ 09 October 2023 — Accepted ※ 14 November 2023 — Issued ※ 29 November 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

MO4AO03

The DESY Open Source FPGA Framework

framework, FPGA, hardware, embedded

222

Ł. Butkowski, A. Bellandi, M. Büchler, B. Dursun, C. Gümüş, N. Omidsajedi, K. Schulz
DESY, Hamburg, Germany

Modern FPGA firmware development involves integrating various intellectual properties (IP), modules written in hardware description languages (HDL), high-level synthesis (HLS), and software/hardware CPUs with embedded Linux or bare-metal applications. This process may involve multiple tools from the same or different vendors, making it complex and challenging. Additionally, scientific institutions such as DESY require long-term maintenance and reproducibility for designs that may involve multiple developers, further complicating the process. To address these challenges, we have developed an open-source FPGA firmware framework (FWK) at DESY that streamlines development, facilitates collaboration, and reduces complexity. The FWK achieves this by providing an abstraction layer, a defined structure, and guidelines to create big FPGA designs with ease. FWK also generates documentation and address maps necessary for high-level software frameworks like ChimeraTK. This paper presents an overview and the idea of the FWK.

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-MO4AO03

About •

Received ※ 05 October 2023 — Accepted ※ 13 October 2023 — Issued ※ 13 October 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TU1BCO02

Integrating System Knowledge in Unsupervised Anomaly Detection Algorithms for Simulation-Based Failure Prediction of Electronic Circuits

ISOL, electron, monitoring, radiation

249

F. Waldhauser, H. Boukabache, D. Perrin, S. Roesler
CERN, Meyrin, Switzerland
M. Dazer
Universität Stuttgart, Stuttgart, Germany

Funding: This work has been sponsored by the Wolfgang Gentner Programme of the German Federal Ministry of Education and Research (grant no. 13E18CHA).
Machine learning algorithms enable failure prediction of large-scale, distributed systems using historical time-series datasets. Although unsupervised learning algorithms represent a possibility to detect an evolving variety of anomalies, they do not provide links between detected data events and system failures. Additional system knowledge is required for machine learning algorithms to determine the nature of detected anomalies, which may represent either healthy system behavior or failure precursors. However, knowledge on failure behavior is expensive to obtain and might only be available upon pre-selection of anomalous system states using unsupervised algorithms. Moreover, system knowledge obtained from evaluation of system states needs to be appropriately provided to the algorithms to enable performance improvements. In this paper, we will present an approach to efficiently configure the integration of system knowledge into unsupervised anomaly detection algorithms for failure prediction. The methodology is based on simulations of failure modes of electronic circuits. Triggering system failures based on synthetically generated failure behaviors enables analysis of the detectability of failures and generation of different types of datasets containing system knowledge. In this way, the requirements for type and extend of system knowledge from different sources can be determined, and suitable algorithms allowing the integration of additional data can be identified.

Slides TU1BCO02 [2.541 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TU1BCO02

About •

Received ※ 02 October 2023 — Accepted ※ 12 October 2023 — Issued ※ 25 October 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TU1BCO04

Laser Focal Position Correction Using FPGA-Based ML Models

controls, laser, network, FPGA

262

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 Number DE-SC 00259037.
High repetition-rate, ultrafast laser systems play a critical role in a host of modern scientific and industrial applications. We present a diagnostic and correction scheme for controlling and determining laser focal position by utilizing fast wavefront sensor measurements from multiple positions to train a focal position predictor. This predictor and additional control algorithms have been integrated into a unified control interface and FPGA-based controller on beamlines at the Bella facility at LBNL. An optics section is adjusted online to provide the desired correction to the focal position on millisecond timescales by determining corrections for an actuator in a telescope section along the beamline. Our initial proof-of-principle demonstrations leveraged pre-compiled data and pre-trained networks operating ex-situ from the laser system. A framework for generating a low-level hardware description of ML-based correction algorithms on FPGA hardware was coupled directly to the beamline using the AMD Xilinx Vitis AI toolchain in conjunction with deployment scripts. 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 Radiation25, 1877-1892 (Nov 2018).

Slides TU1BCO04 [1.876 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TU1BCO04

About •

Received ※ 06 October 2023 — Accepted ※ 14 November 2023 — Issued ※ 18 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TU2BCO04

Accelerator Systems Cyber Security Activities at SLAC

controls, EPICS, network, operation

292

G.R. White, A.L. Edelen
SLAC, Menlo Park, California, USA

Funding: Work supported in part by the U.S. Department of Energy under contract number DE-AC02-76SF00515.
We describe four cyber security related activities of SLAC and collaborations. First, from a broad review of accelerator computing cyber and mission reliability, our analysis method, findings and outcomes. Second, lab-wide and accelerator penetration testing, in particular methods to control, coordinate, and trap, potentially hazardous scans. Third, a summary gap analysis of recent US regulatory orders from common practice at accelerators, and our plans to address these in collaboration with the US Dept. of Energy. Finally, summary attack vectors of EPICS, and technical plans to add authentication and encryption to EPICS itself.

Slides TU2BCO04 [1.677 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TU2BCO04

About •

Received ※ 04 October 2023 — Revised ※ 13 October 2023 — Accepted ※ 15 November 2023 — Issued ※ 17 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUMBCMO13

Applications of Artificial Intelligence in Laser Accelerator Control System

laser, target, controls, experiment

372

F.N. Li, K.C. Chen, Z. Guo, Q.Y. He, C. Lin, Q. Wang, Y. Xia, M.X. Zang
PKU, Beijing, People’s Republic of China

Funding: the National Natural Science Foundation of China (Grants No. 11975037, NO. 61631001 and No. 11921006), and the National Grand Instrument Project (No. 2019YFF01014400 and No. 2019YFF01014404).
Ultra-intense laser-plasma interactions can produce TV/m acceleration gradients, making them promising for compact accelerators. Peking University is constructing a proton radiotherapy system prototype based on PW laser accelerators, but transient processes challenge stability control, critical for medical applications. This work demonstrates artificial intelligence’s (AI) application in laser accelerator control systems. To achieve micro-precision alignment between the ultra-intense laser and target, we propose an automated positioning program using the YOLO algorithm. This real-time method employs a convolutional neural network, directly predicting object locations and class probabilities from input images. It enables precise, automatic solid target alignment in about a hundred milliseconds, reducing experimental preparation time. The YOLO algorithm is also integrated into the safety interlocking system for anti-tailing, allowing quick emergency response. The intelligent control system also enables convenient, accurate beam tuning. We developed high-performance virtual accelerator software using "OpenXAL" and GPU-accelerated multi-particle beam transport simulations. The software allows real-time or custom parameter simulations and features control interfaces compatible with optimization algorithms. By designing tailored objective functions, desired beam size and distribution can be achieved in a few iterations.

Slides TUMBCMO13 [1.162 MB]

Poster TUMBCMO13 [1.011 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUMBCMO13

About •

Received ※ 04 October 2023 — Revised ※ 12 October 2023 — Accepted ※ 23 November 2023 — Issued ※ 23 November 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUMBCMO14

Initial Test of a Machine Learning Based SRF Cavity Active Resonance Control

cavity, controls, SRF, resonance

379

F.Y. Wang, J. Cruz
SLAC, Menlo Park, California, USA

We’ll introduce a high precision active motion controller based on machine learning (ML) technology and electric piezo actuator. The controller will be used for SRF cavity active resonance control, where a data-driven model for system motion dynamics will be developed first, and a model predictive controller (MPC) will be built accordingly. Simulation results as well as initial test results with real SRF cavities will be presented in the paper.

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUMBCMO14

About •

Received ※ 03 October 2023 — Revised ※ 14 November 2023 — Accepted ※ 27 November 2023 — Issued ※ 09 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPDP004

System Identification via ARX Model and Control Design for a Granite Bench at Sirius/LNLS

controls, experiment, feedback, acceleration

479

J.P.S. Furtado, I.E. Santos, T.R. Silva Soares
LNLS, Campinas, Brazil

Modern 4th generation synchrotron facilities demand mechanical systems and hardware capable of fine position control, improving the performance of experiments at the beamlines. In this context, granite benches are widely used to position systems such as optical elements and magnetos, due to its capacity of insulating interferences from the ground. This work aims to identify the transfer function that describes the motion of the granite bench at the EMA Beamline (Extreme conditions Methods of Analysis) and then design the control gains to reach an acceptable motion performance in the simulation environment before embedding the configuration into the real system, followed by the validation at the beamline. This improvement avoids undesired behaviour in the hardware or in the mechanism when designing the controller. The bench, weighting 1.2 tons, is responsible by carrying a coil, weighting 1.8 tons, which objective is to apply a 3 T magnetic field to the sample that receives the beam provided by the electrons accelerator. The system identification method applied in this paper is based on the auto-regressive model with exogenous inputs (ARX). The standard servo control loop of the Omron Delta Tau Power Brick controller and the identified plant were simulated in Simulink in order find the control parameters. This paper shows the results and comparison of the simulations and the final validation of the hardware performance over the real system.

Poster TUPDP004 [0.720 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUPDP004

About •

Received ※ 06 October 2023 — Accepted ※ 28 November 2023 — Issued ※ 17 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPDP098

Automatic Conditioning of High Voltage Pulsed Magnets

kicker, controls, PLC, vacuum

780

C.A. Lolliot, M.J. Barnes, N. Magnin, S. Pavis, P. Van Trappen
CERN, Meyrin, Switzerland
C. Monier
INSA Lyon, Villeurbanne, France

Fast pulsed kicker magnets are used across the various accelerators of CERN complex to inject and extract the beam. These kicker magnets, powered by high voltage pulsed generators and under vacuum, are prone to electrical breakdown during the pulse. To prepare the kicker magnet for reliable operation, or in case an electrical breakdown occurred, a conditioning is necessary: the magnet is pulsed gradually increasing the pulse voltage and length up to a value beyond operational conditions. This is a lengthy process that requires kicker experts on site to manually control the pulse voltage and length, and monitor the vacuum activity. For the start of LHC operation, a first automatic conditioning system was deployed on injection kicker magnet (MKI). Configurable voltage and pulse length ramps are automatically performed by the controller. In case abnormal vacuum activity occurs, the voltage is reduced and then the process continues. Based on this experience, a standardised algorithm has been developed, adding new features such as logarithmic ramp, or simulation of the whole conditioning cycle with test of reaction to vacuum activity. This new automatic conditioning system was deployed on several kicker systems across various CERN accelerators, allowing smoother conditioning, and great reduction on manpower. It also offers the possibility for further automate kicker system operation, starting automatically a magnet conditioning when needed without intervention of kickers experts, similarly as what was deployed for SPS Beam Dump System.

Poster TUPDP098 [0.328 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUPDP098

About •

Received ※ 06 October 2023 — Revised ※ 21 October 2023 — Accepted ※ 05 December 2023 — Issued ※ 17 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPDP109

Tickit: An Event-Based Multi-Device Simulation Framework

framework, controls, hardware, EPICS

823

A. Emery, T.M. Cobb, C.A. Forrester, G. O’Donnell
DLS, Oxfordshire, United Kingdom

Tickit is an event-based multi-device simulation framework providing configuration and orchestration of complex simulations. It was developed at Diamond Light Source in order to overcome limitations presented to us by some of our existing hardware simulations. With the Tickit framework, simulations can be addressed with a compositional approach. It allows devices to be simulated individually while still maintaining the interconnected behaviour exhibited by their hardware counterparts. This is achieved by modelling the interactions between devices, such as electronic signals. Devices can be collated into larger simulated systems providing a layer of simulated hardware against which to test the full stack of Data Acquisition and Controls tools. We aim to use this framework to extend the scope and improve the interoperability of our simulations; enabling us to further improve the testing of current systems and providing a preferential platform to assist in development of the new Acquisition and Controls tools.

Poster TUPDP109 [0.703 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUPDP109

About •

Received ※ 29 September 2023 — Revised ※ 21 October 2023 — Accepted ※ 04 December 2023 — Issued ※ 18 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPDP115

Machine Learning for Compact Industrial Accelerators

cavity, controls, industrial-accelerators, network

846

J.P. Edelen, J.A. Einstein-Curtis, M.J. Henderson, M.C. Kilpatrick
RadiaSoft LLC, Boulder, Colorado, USA
J.A. Diaz Cruz, A.L. Edelen
SLAC, Menlo Park, California, USA

Funding: This material is based upon work supported by the DOE Accelerator R&D and Production under Award Number DE-SC0023641.
The industrial and medical accelerator industry is an ever-growing field with advancements in accelerator technology enabling its adoption for new applications. As the complexity of industrial accelerators grows so does the need for more sophisticated control systems to regulate their operation. Moreover, the environment for industrial and medical accelerators is often harsh and noisy as opposed to the more controlled environment of a laboratory-based machine. This environment makes control more challenging. Additionally, instrumentation for industrial accelerators is limited making it difficult at times to identify and diagnose problems when they occur. RadiaSoft has partnered with SLAC to develop new machine learning methods for control and anomaly detection for industrial accelerators. Our approach is to develop our methods using simulation models followed by testing on experimental systems. Here we present initial results using simulations of a room temperature s-band system.

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUPDP115

About •

Received ※ 06 October 2023 — Accepted ※ 05 December 2023 — Issued ※ 18 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUSDSC02

Integrating Online Analysis with Experiments to Improve X-Ray Light Source Operations

experiment, interface, real-time, framework

921

N.M. Cook, E.G. Carlin, J.A. Einstein-Curtis, R. Nagler, R. O’Rourke
RadiaSoft LLC, Boulder, Colorado, USA
A.M. Barbour, M. Rakitin, L. Wiegart, H. Wijesinghe
BNL, Upton, New York, USA

Funding: This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research under Award Number DE-SC00215553.
The design, execution, and analysis of light source experiments requires the use of sophisticated simulation, controls and data management tools. Existing workflows require significant specialization to accommodate specific beamline operations and data pre-processing steps necessary for more intensive analysis. Recent efforts to address these needs at the National Synchrotron Light Source II (NSLS-II) have resulted in the creation of the Bluesky data collection framework, an open-source library for coordinating experimental control and data collection. Bluesky provides high level abstraction of experimental procedures and instrument readouts to encapsulate generic workflows. We present a prototype data analysis platform for integrating data collection with real time analysis at the beamline. Our application leverages Bluesky in combination with a flexible run engine to execute user configurable Python-based analyses with customizable queueing and resource management. We discuss initial demonstrations to support X-ray photon correlation spectroscopy experiments and future efforts to expand the platform’s features.

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUSDSC02

About •

Received ※ 06 October 2023 — Revised ※ 22 October 2023 — Accepted ※ 11 December 2023 — Issued ※ 14 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THMBCMO17

FAIR Data of Physical and Digital Beamlines

experiment, software, controls, GUI

1231

G. Günther, O. Mannix, O. Ruslan, S. Vadilonga
HZB, Berlin, Germany

Simulations play a crucial role in instrument design, as a digital precursor of a real-world object they contain a comprehensive description of the setup. Unfortunately, this digital representation is often neglected once the real instrument is fully commissioned. To preserve the symbiosis of simulated and real-world instrument beyond commissioning we connect the two worlds through the instrument control software. The instrument control simultaneously starts measurements and simulations, receives feedback from both, and directs (meta)data to a NeXus file - a standard format in photon and neutron science. The instrument section of the produced NeXus file is enriched with detailed simulation parameters where the current state of the instrument is reflected by including real motor positions such as incorporating the actual aperture of a slit system. As a result, the enriched instrument description increases the reusability of experimental data in sense of the FAIR principles. The data is ready to be exploited by machine-learning techniques, such as for predictive maintenance applications as it is possible to perform simulations of a measurement directly from the NeXus file. The realization at the Aquarius beamline * at Bessy II in connection with the Ray-UI simulation software ** and RayPyNG API *** serves as a prototype for a more general application.
* https://www.helmholtz-berlin.de/forschung/oe/wi/optik-strahlrohre/projekte/aquarius_en.html
** https://doi.org/10.1063/1.5084665
*** https://pypi.org/project/raypyng

Slides THMBCMO17 [0.632 MB]

Poster THMBCMO17 [0.828 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-THMBCMO17

About •

Received ※ 06 October 2023 — Accepted ※ 11 December 2023 — Issued ※ 14 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THMBCMO18

Advancements in Beamline Digital Twin at BESSYII

operation, software, experiment, MMI

1236

S. Vadilonga, G. Günther, S. Kazarski, R. Ovsyannikov, S.S. Sachse, W. Smith
HZB, Berlin, Germany

This presentation reports on the status of beamline digital twins at BESSY II. To provide a comprehensive beamline simulation experience we have leveraged BESSY II’s x-ray tracing program, RAY-UI[*], widely used for beamline design and commissioning and best adapted to the requirements of our soft X-ray source BESSY II. We created a Python API, RayPyNG, capable to convert our library of beamline configuration files produced by RAY-UI into Python objects[**]. This allows to embed beamline simulation into Bluesky[***], our experimental controls software ecosystem. All optical elements are mapped directly into the Bluesky device abstraction (Ophyd). Thus beamline operators can run simulations and operate real systems by a common interface, allowing to directly compare theory predictions with real-time results[****]. We will discuss the relevance of this digital twin for process tuning in terms of enhanced beamline performance and streamlined operations. We will shortly discuss alternatives to RAY-UI like other software packages and ML/AI surrogate models.
[*]https://doi.org/10.1063/1.5084665
[**]https://raypyng.readthedocs.io/
[***]https://doi.org/10.1080/08940886.2019.1608121
[****]https://raypyng-bluesky.readthedocs.io/

Slides THMBCMO18 [0.333 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-THMBCMO18

About •

Received ※ 06 October 2023 — Accepted ※ 11 December 2023 — Issued ※ 16 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THPDP060

Beam Instrumentation Simulation in Python

electron, electronics, instrumentation, framework

1454

M. Gonzalez-Berges, D. Alves, A. Boccardi, V. Chariton, I. Degl’Innocenti, S. Jackson, J. Martínez Samblas
CERN, Meyrin, Switzerland

The design of acquisition electronics for particle accelerator systems relies on simulations in various domains. System level simulation frameworks can integrate the results of specific tools with analytical models and stochastic analysis. This allows the designer to estimate the performance of different architectures, compare the results, and ultimately optimize the design. These simulation frameworks are often made of custom scripts for specific designs, which are hard to share or reuse. Adopting a standard interface for modular components can address these issues. Also, providing a graphical interface where these components can be easily configured, connected and the results visualised, eases the creation of simulations. This paper identifies which characteristics ISPy (Instrumentation Simulation in Python) should fulfill as a simulation framework. It subsequently proposes a standard format for signal-processing simulation modules. Existing environments which allow script integration and an intuitive graphical interface have then been evaluated and the KNIME Analytics Platform was the proposed solution. Additionally, the need to handle parameter sweeps for any parameter of the simulation, and the need for a bespoke visualisation tool will be discussed. Python has been chosen for all of these developments due to its flexibility and its wide adoption in the scientific community. The ensuing performance of the tool will also be discussed.

Poster THPDP060 [2.931 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-THPDP060

About •

Received ※ 07 October 2023 — Accepted ※ 08 December 2023 — Issued ※ 12 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THPDP062

Controls Optimization for Energy Efficient Cooling and Ventilation at CERN

controls, operation, ECR, software

1465

D. Monteiro, R. Barillère, N. Bunijevac, I. Rühl
CERN, Meyrin, Switzerland

Cooling and air conditioning systems play a vital role for the operation of the accelerators and experimental complex of the European Organization for Nuclear Research (CERN). Without them, critical accelerator machinery would not operate reliably as many machines require a fine controlled thermodynamic environment. These operation conditions come with a significant energy consumption: about 12% (75 GWh) of electricity consumed by the Large Hadron Collider (LHC) during a regular run period is devoted to cooling and air conditioning. To align with global CERN objectives of minimizing its impact on the environment, the Cooling and Ventilation (CV) group, within the Engineering Department (EN), has been developing several initiatives focused on energy savings. A particular effort is led by the automation and controls section which has been looking at how controls and automation strategies can be optimized without requiring costly hardware changes. This paper addresses several projects of this nature, by presenting their methodology and results achieved to date. Some of them are particularly promising as real measurements revealed that electricity consumption was more than halved after implementation. Due to the pertinence of this effort in the current context of energy crisis, the paper also draws a careful reflection on how it is planned to be further pursued to provide more energy-efficient cooling and ventilation services at CERN.

Poster THPDP062 [7.056 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-THPDP062

About •

Received ※ 05 October 2023 — Accepted ※ 08 December 2023 — Issued ※ 10 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

FR2AO02

A Digital Twin for Neutron Instruments

neutron, software, experiment, detector

1626

S. Nourbakhsh, Y. Le Goc, P. Mutti
ILL, Grenoble, France

Data from virtual experiments are becoming an extremely valuable asset for research infrastructures in a multitude of aspects and different actors: for instrument scientists to develop and optimise current and future instruments; for training external users in the usage of the instrument control system; for scientists in studying, quantifying and reducing instrumental effects on acquired data. Furthermore large sets of simulated data are also a necessary ingredient for the development of surrogate models for faster and more accurate simulation, reduction and analysis of the data. The development of a digital twin of an instrument can answer such different needs with a single unified approach wrapping in a user-friendly envelop the knowledge about the instrument physical description, the specific of the simulation packages and their interaction, and the high performing computing setup. In this article we will present the general architecture of the digital twin prototype developed at the ILL in the framework of the PANOSC European project in close collaboration with other research facilities (ESS and EuXFel). The communication patterns (based on ZQM) and interaction between the control system (NOMAD), simulation software (McStas), instrument description and configuration, process management (CAMEO) will be detailed. The adoption of FAIR principles for data formats and policies in combination with open-source software make it a sustainable project both for development and maintenance in the mid and long-term.

Slides FR2AO02 [1.245 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-FR2AO02

About •

Received ※ 31 October 2023 — Revised ※ 02 November 2023 — Accepted ※ 05 December 2023 — Issued ※ 07 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

FR2AO04

A Physics-Based Simulator to Facilitate Reinforcement Learning in the RHIC Accelerator Complex

cavity, controls, booster, diagnostics

1630

L.K. Nguyen, K.A. Brown, M.R. Costanzo, Y. Gao, M. Harvey, J.P. Jamilkowski, J. Morris, V. Schoefer
BNL, Upton, New York, USA

Funding: Work supported by Brookhaven Science Associates, LLC under Contract No. DE-SC0012704 with the U.S. Department of Energy.
The successful use of machine learning (ML) in particle accelerators has greatly expanded in recent years; however, the realities of operations often mean very limited machine availability for ML development, impeding its progress in many cases. This paper presents a framework for exploiting physics-based simulations, coupled with real machine data structure, to facilitate the investigation and implementation of reinforcement learning (RL) algorithms, using the longitudinal bunch-merge process in the Booster and Alternating Gradient Synchrotron (AGS) at Brookhaven National Laboratory (BNL) as examples. Here, an initial fake wall current monitor (WCM) signal is fed through a noisy physics-based model simulating the behavior of bunches in the accelerator under given RF parameters and external perturbations between WCM samples; the resulting output becomes the input for the RL algorithm and subsequent pass through the simulated ring, whose RF parameters have been modified by the RL algorithm. This process continues until an optimal policy for the RF bunch merge gymnastics has been learned for injecting bunches with the required intensity and emittance into the Relativistic Heavy Ion Collider (RHIC), according to the physics model. Robustness of the RL algorithm can be evaluated by introducing other drifts and noisy scenarios before the algorithm is deployed and final optimization occurs in the field.

Slides FR2AO04 [2.694 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-FR2AO04

About •

Received ※ 04 October 2023 — Accepted ※ 05 December 2023 — Issued ※ 16 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

FR2AO05

Python Library for Simulated Commissioning of Storage-Ring Accelerators

MMI, lattice, storage-ring, closed-orbit

1637

L. Malina, I.V. Agapov, J. Keil, E.S.H. Musa, B. Veglia
DESY, Hamburg, Germany
N. Carmignani, L.R. Carver, L. Hoummi, S.M. Liuzzo, T.P. Perron, S.M. White
ESRF, Grenoble, France
T. Hellert
LBNL, Berkeley, California, USA

Simulations of the commissioning procedure became vital to the storage-ring lattice design process. The achievable tolerances on lattice imperfections, such as equipment misalignments or magnet gradient errors, would, without correction, prohibit reaching the design parameters. We present a Python library which includes an extensive set of error sources in the accelerator lattice and provides a variety of correction algorithms to commission a storage ring. The underlying beam dynamics simulations are performed with pyAT. This project builds upon previous works and expands them in the direction of realistic control room experience and software maintainability. The performance is demonstrated using example commissioning studies, and further development plans are discussed.

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-FR2AO05

About •

Received ※ 06 October 2023 — Revised ※ 27 October 2023 — Accepted ※ 05 December 2023 — Issued ※ 19 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)