Low Level RF (LLRF) control systems of linear accelerators (LINACs) are typically implemented with heterodyne based architectures, which have complex analog RF mixers for up and down conversion. The Gen 3 RF System-on-Chip (RFSoC) device from AMD Xilinx integrates data converters with maximum RF frequency of 6 GHz. That enables direct RF sampling of C-band LLRF signal typically operated at 5.712 GHz without RF mixers, which can significantly simplify the system architecture. The data converters sample RF signals in higher order Nyquist zones and then up or down converted digitally by the integrated data path. The closed-loop feedback control firmware implemented in FPGA integrated in RFSoC can process the baseband signal from the ADC data path and calculate the updated phase and amplitude to be up-mixed by the DAC data path. We have developed an LLRF control RFSoC platform, which targets Cool Copper Collider (C3) and other C or S band LINAC research and development projects. In this paper, the architecture of the platform and the test results for some of the key performance parameters, such as phase and amplitude stability with our custom solid-state amplifier, will be described.

The Relativistic Heavy Ion Collider (RHIC) Run 23 program consisted of collisions of 100 GeV gold beams at two collision points for the first time since 2016; the sPHENIX collaboration used the beam to commission their new detector systems while STAR took physics data. Completion of sPHENIX construction pushed the start of the run to May, forcing the collider complex to operate over the summer months and incurring lower than normal availability due to heat and power dip related problems. Issues with dynamic pressure rise during acceleration through transition resulted in a slower ramp up of intensity compared to prior years. Finally, a failure of a warm-to-cold current lead interface in the valve box for the Main Magnet power supply forced the run to end. This paper will discuss the progress made by each experiment and the failure mode, repair and mitigation efforts in preparation for Run 24.

Reliable operation of a seeded Free Electron Laser requires the simultaneous control of several electron-beam, laser and accelerator parameters. With EEHG the complexity increases due to the second seed laser and the strong dependence of EEHG bunching to seeding parameters. With the recent upgrade of the FEL-1 line, FERMI is the first FEL facility to be operated in EEHG mode for users. This required a major work for developing software tools that could be used to easily set the FEL at the desired wavelength. We report here on the recent software developments at FERMI for the operations of the new FEL-1. An important prerequisite for EEHG is to determine both the electron beam energy spread and seed laser induced energy modulation. This is done by using HGHG time dependent bunching equations to match experimental parameters scans. With these data, optimal EEHG settings of the machine parameters are then calculated to reach the desired FEL wavelength. The requested parameters are then sent to interface tools that accurately control laser, undulator, chicane and electron beam.

Generating nanometer-scale density modulation has been pursued due to its potential for compact X-ray source applications. Realization of this nanometer modulation involves two key challenges: development of sub-micron-scale momentum modulation method and conversion method to density modulation without quality degradation. Addressing the first challenge, emittance exchange (EEX) beamline is a promising candidate. Its unique capability of transverse-to-longitudinal phase space exchange makes it compatible with various modulators imparting either transverse or longitudinal modulations. This versatility allows us to find optimal radiators, addressing the second challenge. Study on degradation sources and their effects on the beam are underway to realize nanometer-scale modulation using EEX beamline. We present most recent results from our simulation study.

Bunch trains have been considered as a promising means of generating intense, coherent radiation in compact accelerator facilities. However, conventional methods, which impart a sinusoidal modulation on the beam’s longitudinal phase space, are inefficient for generating a high bunching factor density modulation. Only a small fraction of a sinusoidal modulation, which has linearity, primarily forms density spikes while other particles under nonlinear correlation have limited contribution to these spikes. One way to improve such bunching efficiency is imparting a saw-tooth correlation, which has piecewise-linearities. This correlation maximizes the peaks of density spikes as more than 90% of particles will contribute to the spikes. While such correlation can be generated by a series of transverse wigglers, a single transverse wiggler with shaped poles to introduce higher harmonics can generate saw-tooth or saw-tooth-like correlations. We present a recent study on this new approach, employing a shaped-pole transverse wiggler.

Plasma temperature dynamics provide significant insight when evaluating ion source performance. Quantities such as beam emittance, or mean transverse energy, are strongly correlated with the source plasma temperature. At LANSCE there is currently no method implemented for measuring initial source emittance or implementing tunability of mean transverse energy through ion source control parameters. In this work we will discuss our demonstration of a new laser diagnostic tool for measuring H- beam emittance on the LANSCE H- ion source laser diagnostic stand. Our investigated method will be an extension of systems outlined for NIFS, and will be optimized for rapid response times, scanning the Doppler broadened Hydrogen-alpha emission line at a rate 10x faster than the plasma ignition time window (800 microseconds). We will show that our real-time, non-intrusive measurement approach will enable characterization and study of source control parameter effects on source plasma temperature for future emittance optimization.

In a generalized longitudinal strong focusing (GLSF) approach employed in steady-state microbunching (SSMB) storage rings, the objective is to achieve complete cancellation of modulations by ensuring that a particle's longitudinal position remains unchanged after passing through both modulators. This requires effective control over the deviation in longitudinal position, which arises from lattice nonlinearities. This paper derives analytical formulations for the mean and standard deviation of the particle position deviation, expressed in terms of the beam and lattice-dependent parameters. The aim is to provide valuable insights into the system's behavior and enable optimization of the GLSF unit's performance.

Magnetic hysteresis, eddy currents, and manufacturing imperfections pose significant challenges for beam operation in multi-cycling synchrotrons. Addressing the dynamic dependency of magnetic fields on cycling history is a current limitation for control room tools using existing models. This paper outlines recent advancements to solve this, presenting the outcome of operational tests utilizing data-driven approaches and an overview of the next steps. Notably, artificial neural networks, including long short-term memory networks, transformers and other time series analysis architectures, are employed to model static and dynamic effects in the main dipole magnets of the CERN SPS. These networks capture hysteresis and eddy current decays based on measured magnetic field and data from the real-time magnetic measurement system of the SPS main dipoles. Cycle-by-cycle feed-forward corrections are implemented through the CERN accelerator controls infrastructure, which propagate corrections of magnetic fields to corresponding adjustments in the current of the power converters feeding the magnets.

Transfer maps of different ion optical elements are usually obtained via ray tracing methods without taking into account the imperfections and misalignments of the optics. Normally beam profile monitors do not measure the full 6D phase-space, but only a portion of it. To verify the beam phase-space, we have constructed an Ion Optics Test Stand (IOTS) that is located at the Low Energy Branch (LEB) of the Jozef Stefan Institute in Ljubljana, Slovenia [1]. The IOTS consists of two Allison emittance scanners (AES) [2] with an electrode sandwiched between them, and is supplied by the LEB with a variety of ion beams with energies up to 20 keV. This allows us to automatically measure the 6D beam phase-space before and after the electrode and determine the electrodes transfer map. We will discuss the status of the IOTS, the emittance scanners, electrode transfer map measurements with them, and describe an example of AES--Einzel lens--AES test configuration. We will also show how the phase-space measurements performed with the IOTS can be used as a training ground of Machine Learning (ML) tools designed for ion optics optimization with respect to a preferred transport metric.

The 2D Orszag-Tang vortex magnetohydrodynamics (MHD) problem is studied through the use of physics-constrained convolutional neural networks (PCNNs). The density and the magnetic field are forecasted, and we also predict magnetic field given the velocity field of the fluid. We examined the incorporation of various physics constraints into the PCNNs: absence of magnetic monopoles, non-negativity of density and use of only relevant variables. Translation equivariance was present from the convolutional architecture. The use of a residual architecture and data augmentation was found to increase performance greatly. The most accurate models were incorporated into the simulation, with reasonably accurate results. For the prediction task, the PCNNs were evaluated against a physics-informed neural network (PINN), which had the ideal MHD induction equation as a soft constraint. The use of PCNNs for MHD has the potential to produce physically consistent real-time simulations to serve as virtual diagnostics in cases where inferences must be made with limited observables.

A particle accelerator is a time-varying complex system whose various components are regularly perturbed by external disturbances. The tuning of the accelerator can be a time-consuming process involving manual adjustment of multiple components, such as RF cavities, to minimize beam loss due to time-varying drifts. The high dimensionality of the system (~100 amplitude and phase RF settings in the LANSCE accelerator) makes it difficult to achieve optimal operation. The time-varying drifts and the dimensionality make system parameter estimation a challenging optimization problem. In this work, we propose a variational autoencoded latent regression (VAELR) model for robust estimation of system parameters using 2D unique projections of a charged particle beam's 6D phase space. In VAELR, VAE projects the phase space projections into a lower-dimensional latent space, and a dense neural network maps the latent space onto the space of system parameters. The trained network can predict system parameters for unseen phase space projections. Furthermore, VAELR can generate new projections by randomly sampling the latent space of VAE and also estimate the corresponding system parameters.

The main ring (MR) of the Japan proton accelerator research complex (J-PARC) delivers the high-intensity proton beams to the T2K long-baseline neutrino experiment. To observe charge-conjugation and parity-transformation violation in the lepton sector with high accuracy, the upgrade of the MR toward the beam power of 1.3 MW is mandatory. The magnet power supply system of MR was upgraded for this purpose during the long-term shutdown period in FY2021. However, the asymmetry of the closed-orbit distortion (COD) was observed after the upgrade. The cause of the asymmetry was attributed to the large ripples of the excitation currents for the bending magnets. The measures to reduce the ripples were applied to six identical power supplies for the bending magnets, and then the asymmetry was successfully corrected. This result suggests the tune region of the stable beam operation is expected to be improved since the effect of the non-structure resonance should be suppressed. This presentation reports the scheme of the ripple reduction for the excitation currents of the bending magnets and the measurement results of the COD in the MR.

The ATF2-3 beamline is the only facility in the world for testing the Final Focus Beamline of linear colliders and is essential for the ILC and the CLIC projects. A vertical electron beam size of 41 nm (within 10% of the target), a closed-loop intra-bunch feedback of latency 133 ns, and direct stabilization of the beam position at the Interaction Point to 41 nm (limited by IP BPM resolution) have all been achieved at ATF2. These results fulfilled the two main ATF2 design goals, but were obtained with reduced aberration optics and a bunch population of approximately 10% of the nominal value of 10^10 electrons. Recent studies indicate that the beam degradation with the beam intensity is due to the effects of wakefields. To overcome this intensity limitation, hardware upgrades including new vacuum chambers, magnets, IP-Beam Size Monitor laser, cavity BPMs, wakefield mitigation station, as well as a comprehensive R&D program to maximize the luminosity potential are being pursued in the framework of the ILC Technology Network. This new R&D program focuses on the study of wakefield mitigation techniques, correction of higher-order aberrations, tuning strategies, including AI techniques, as well as beam instrumentation issues, such as the BPMs, advanced Cherenkov Diffractive Radiation monitors, and fast feedback systems, among others. This paper summarizes the hardware upgrades, the R&D program and the results of the Fall 2023-Winter 2024 experimental campaign performed in ATF2-3.

Fast single-turn injection kicker systems deflect incoming beam onto the orbit of the LHC. The higher intensities of High Luminosity (HL) LHC beams are predicted to cause the ferrite yokes of the LHC injection kicker magnets (MKI), in their current configuration, to heat up to their Curie temperature. Studies to reduce the beam induced heating have been carried out over the past years and resulted in a design featuring a water-cooled RF damper. A significant portion of the beam induced power has been relocated from the yoke to a ferrite in the RF damper. The ferrite damper is cooled via a copper sleeve, brazed to the ferrite, via a set of water pipes. The manufacturing of this RF damper system is challenging since different materials are brazed together to form a complex and fragile assembly, optimized for heat transfer, installed in an ultra-high vacuum environment. This paper outlines fabrication methods and their reproducibility, compares the results of measurements of the thermal interface between the ferrite and copper sleeve, and concludes on the challenges of assuring a production technique that results in a reliable and suitable thermal interface.

The LBNL Advanced Light Source, a pioneering third-generation soft x-ray synchrotron radiation source operating at 1.89 GeV with a 2~nm beam emittance, stands as one of the earliest facilities in its class, continually evolving to maintain its status at the forefront of soft x-ray sources. This paper reviews the most significant advancements in the accelerator's hardware and software infrastructures and presents the machine operational statistics over the last 15 years.

The Advanced Photon Source Upgrade (APS-U) project has developed and installed a multi-bend achromat (MBA) lattice operating at 6.0 GeV beam energy to replace the existing APS storage ring lattice that operated at 7.0 GeV. A major part of the project is to install 60 hybrid permanent magnet undulator (HPMU) insertion devices (IDs) that include 12 revolver undulators, each with two magnetic structures (for a total of 72 magnetic structures); and one electromagnetic undulator for intermediate energy x-rays (IEX). These IDs will outfit 35 sectors. We have developed new HPMU designs for five different period lengths used in 46 magnetic structures, and we will reuse 26 existing magnetic structures with four additional period lengths. Eight new superconducting undulators (SCUs) have been designed and built with two short period lengths and three different overall lengths [1-3]. The SCUs will be installed in both inline and canted configurations after beam commissioning is completed and the user runs start. Demanding field requirements for the undulators were expected to be challenging for the undulator tuning, especially given the tight schedule. All undulators underwent rigorous tuning and control system tests before they were installed in the new storage ring. We will provide a status and schedule update including presenting measurement results of the IDs.

Prior to assembly into the operational cryostat each superconducting undulator (SCU) at the Advanced Photon Source undergoes testing in a LHe bath cryostat where coil training and magnetic measurements are performed. If necessary, the baseline magnetic measurements are used for phase error tuning which is achieved by adjusting the magnetic gap of the SCU at prescribed locations. An optimization routine using a genetic algorithm is used to determine the magnitude of the gap change. Once complete, the SCUs are incorporated into the production cryostat and magnetic measurements of the final assembly are performed. Details of the process during phase error tuning and LHe bath testing of a 1.5 m-long SCU magnet are presented.

New Hybrid Permanent Magnet Undulators (HPMUs) have been designed and manufactured using servo motors for precise and reliable gap motion control for the Advanced Photon Source Upgrade (APS-U) project. Meanwhile, existing HPMUs equipped with legacy stepper motors are systematically replaced with servo motors. In parallel with mechanical modifications of the undulators, a comprehensive upgrade has been implemented for the control of the devices. This upgrade includes integration of standardized industrial components for replacement of motor controllers and motor drives using the Kollmorgen Programmable Controller Multi-axis Master (PCMM) controllers and the AKD2G series servo drives. Soft Input Output Controllers (IOCs) are developed and deployed to replace the legacy VME-based IOCs for both single-period undulators and Revolver undulators. In this paper, we will present the architecture of the new insertion device control system, including control mechanisms, interlock protocols, and tools for diagnostics and troubleshooting.

The present storage ring of the Siam Photon Source is equipped with the new 118 MHz capacitive type RF cavity, adapted from MAX-IV laboratory. This cavity has been installed in the ring since 2016. The cavity is operated with the digital low level RF controller and the solid-state RF amplifier. The system is running fine with less downtime and maintenance. After the full four insertion devices were added in the ring, there are instabilities detected in the beam signals. Investigation on the cavity were carried out with the simulation and measurement to characterize the higher order modes that may causes beam instabilities, especially the longitudinal modes. Simulation of the higher order modes will be presented. The modes properties from the measurements with various temperatures will also be presented. The cavity has two ports on its body reserved for the higher order modes damping mechanism. This study will be served as the baseline of the modes for the future designing of the damping mechanism.

The Advanced Photon Source at Argonne National Laboratory (ANL) is in the process of acquiring solid state amplifiers (SSA) from R&K Limited to replace four 1MW klystron rf systems that provide rf power to the storage ring cavities. This project is necessary primarily due to klystron obsolescence. Based on present needs for the APS Upgrade, twelve 160kW SS rf amplifier systems will be required to replace the legacy klystrons. Each of the 352-MHz SSA systems consist of a rack-mounted control unit, two 85kW amplifier cabinets, four 48-way coaxial combiners, and a single 4-way waveguide final combiner. The system is designed with particular attention to reliability and redundancy to help ensure high reliability metrics for the APS-U rf system.

The upcoming High-Luminosity Large Hadron Collider (HL-LHC) program requires a beam performance in the CERN Proton Synchrotron (PS) that is at the limits of the current RF systems. Following the discontinuation of the RF tube production of the driver amplifiers a new solid-state design has been developed using radiation-hard amplifier technology. In view that the current system architecture has reached its maximum achievable gain, the goal was to reduce the cavity impedance encountered by high-intensity circulating beams. This reduction is achieved by increasing the fast feedback gain around the 10 MHz cavities. A 400W modular driver amplifier based on GaN technology and its control system have been prototyped and are currently in the testing phase. The FETs have been qualified for radiation in J-PARC and they will undergo additional irradiation time in the PS tunnel at CERN to additionally qualify the amplifier in its entirety. The paper outlines the modeling phase, the challenges encountered during prototyping, and the achieved results.

Industrial accelerator applications require efficient, scalable, continuous wave (CW) microwave power systems. Magnetrons are inexpensive and efficient devices for converting electrical energy into microwave power; however, their power output is limited to approximately 100 kW. Cost effective power combining magnetron systems would serve the accelerator industry by providing practical and affordable RF power to accelerator applications. In a magic tee configuration, two oscillators can be power combined and locked to a common frequency. Researchers at General Atomics, in collaboration with Thomas Jefferson National Accelerator Facility, have constructed an experiment to demonstrate the power combining of magnetrons in a such a configuration. An analytic model is presented describing the power combining efficiency of a 4-port magic tee, accounting for two magnetron output signals, an injection signal, and a reactive load. The Adler-Chen model is solved numerically using robust computational geometry techniques*. These complete solutions provide insight to the phenomena of magnetron frequency locking and optimal combining efficiency, which are compared to experiment.

The magnetron as an efficient RF source for a compact industrial SRF accelerator has been proposed [1]. The performance of injection phase lock on two independent magnetron transmitters operated at 915 MHz, in CW mode with maximum power of 75 kW each has been demonstrated to satisfy for this application [2]. This industrial type magnetron has transformer and SCR rectifier on the DC anode power supply. Output power spectrum with phase locking can achieve noise reduction of -21 dBc at the 1st 60 Hz, -29 dBc at 1st 120 Hz with only -22.6 dBc injection power. Solenoid current increase of 16% can increase the magnetron relative natural frequency by 4e-4. Further solenoid current modulation with feedback control and the 2x75 kW power combining scheme with the WR975 magic-tee are to be further studied. We intend to use one 75 kW power station with InnoSys' switching DC power supplies to drive normal conducting and superconducting RF cavities for an industrial compact linac. We are also going to report on the 4x1.2 kW power combining experiment on the 2450 MHz magnetron system carried out at GA, including the control algorithm with modified magnetron heads with trim-coils and characterized at JLab.

A failure of the RHIC powering system occurred at the end of run 23 and led to the discovery of ruptured con-ductors on the 12x150A current leads used to feed cur-rent to the superconducting (SC) magnet circuits. These ruptured conductors are thought to have led to an electri-cal breakdown, first within the solder joint, and then across adjacent conductors of the same current lead assembly. A fatigue experiment has been set up to study the behavior of Sn96Ag4 solder joints under cycling load at cryogenic temperature. Mitigation measures to mini-mize further fatigue cycling have been implemented for the next RHIC run and will be discussed. This paper aims to describe our understanding of the solder joint cracking issue encountered and present the mitigation measures for future RHIC operation.

Paul Scherrer Institute (PSI) has led significant advancements in accelerator electronics development, leveraging Field Programmable Gate Arrays (FPGA) based Digital Signal Processing (DSP) across various critical systems, including Low Level RF (LLRF), Longitudinal Beam Loss Monitoring (LBLM), charge particle measurement via Integrating Current Transformers (ICT), Timing, Filling Pattern Monitor (FPM), Beam Position Monitor (BPM) and other essential beam instruments. Over the past decade, PSI’s approach to develop in-house control system platform (e.g. CPCI-S.0), has encouraged innovation. The strategic reorganization within PSI, fostering collaboration among FPGA firmware engineers, led to the inception of Open-Source FPGA DSP libraries hosted on GitHub. Serving as a comprehensive repository, these libraries empower developers by providing common FPGA IPs, fundamental DSP algorithms and Fixed-Point (FP) arithmetic units. Their presence advances prototype development by enabling rapid assembly of several measurement and or control concepts. In this contribution, we present the features and the transformative impact of the PSI Open-source FPGA libraries with a focus on LLRF. This initiative has not only empowered our team to provide valuable insights, but has also streamlined the integration of new recruits and students, enabling the seamless continuation of FPGA design frameworks.

The accelerator network of the Pohang Accelerator Laboratory (PAL) was initially designed and installed in 2015. It consists of three types of networks: a Public Network for external internet access, an Operation Network for accelerator operation and overall control, and a Control Network for device control and monitoring. From a hardware perspective, it comprises 2 firewalls, 1 intrusion prevention system, 4 backbones, 36 office network switches, and 77 switches for gallery and tunnel networks. Each network is physically or logically separated, and the backbone, serving as the main equipment, is configured in a redundant manner to prepare for failures

The High-Luminosity LHC project aims to enhance the integrated luminosity of the LHC machine by a factor of 10, by upgrading various components located in the LHC tunnel just before the collision points, with cutting-edge technologies. Among these innovations are the new superconducting magnets equipped with a combination of $Nb-Ti$ and $Nb_3Sn$. conductors. Over 100 magnets are being produced, each undergoing multiple production and test stages across different facilities worldwide, including laboratories outside CERN. Various technology systems are integrated into the magnets, involving collaboration with different groups for assembly work. Recognizing the complexity of this production process, a comprehensive production and test schedule at CERN was established. This paper elucidates the schedule tools implemented to oversee the entire resource loaded process. The compiled data serves to identify strategic or technical bottlenecks in the production flow. By adopting such an approach and simulating various production scenarios, the aim is to proactively address potential conflicts, to ensure the optimal allocation of resources and the readiness for installation during the Long Shutdown 3.

The operation of the COler SYnchrotron COSY and its further development ended in October 2023. We briefly review the operation of the accelerator facility and continuous development of its sub-systems. Additionally, this work is put in context of the transformation process that COSY operation and the Institute of Nuclear Physics (IKP) of the Research Center Jülich went through starting 2015. Furthermore, the decommissioning strategy along with the possible further use of COSY components are discussed.

Recently, the heavy ion synchrotron SIS18 at GSI was for the first time operated with a dual-isotope beam, made up of 12C3+ and 4He+. Such a beam can be used to improve carbon radiotherapy by providing online information on dose deposition, where the helium ions serve as a probe beam traversing the patient while depositing a negligible dose. For this, the accelerator has to deliver a slowly extracted beam with a fixed fraction of helium over the spill. The difference in mass-to-charge ratio of 4He compared to 12C is small enough to permit simultaneous acceleration and to make the two isotopes practically indistinguishable for the accelerator instrumentation. Yet, it may cause a temporal shift between the two components in the spill owing to the sensitivity of slow extraction to tiny tune variations. We investigated different extraction methods, and examined the time-wise stability of the dual-isotope beam with a beam monitoring setup installed in the GSI biophysics experiment room. A constant helium fraction was obtained using transverse knock-out extraction with adjusted chromaticity.

The Advanced Photon Source (APS) facility has just completed an upgrade to become one of the world’s brightest storage-ring light sources. For the first time, machine learning (ML) methods have been developed and used as part of the baseline commissioning plan. One such method is Bayesian optimization (BO) – a versatile tool for efficient high-dimensional single and multi-objective tuning, as well as surrogate model construction and other purposes. In this paper we will present our development work on adapting BO to practical control room problems such as tuning linac and booster transmission efficiency, injection stabilization, enlarging storage ring dynamic and momentum apertures, and various other tasks. We will also show first experimental results of these efforts, including achieving initial beam capture in the APS-U storage ring. Given the success of BO methods at APS, we are working on tighter ML method integration into the standard control room procedures through a dedicated graphical interface.

The CERN Super Proton Synchrotron (SPS) offers slow-extracted, high-intensity proton beams at 400 GeV/c for 3 fixed targets in the CERN North Experimental Area (NA) with a spill length of about 5 seconds. Since first commissioning in the late seventies, the NA has seen a steady increase in users, many of which requiring improved spill quality control. Slow extraction is sensitive to small perturbations with the effect of reduced spill quality. While some of these effects have been addressed in recent years, continuous compensation of intensity fluctuations at 50 Hz harmonics originating from power converter ripple has been particularly difficult to achieve. In 2023, the deployment of two techniques - "Empty-Bucket Channeling" and active control with Adaptive Bayesian Optimization – resulted in a significant suppression of these intensity modulations. This paper focuses on using Adaptive Bayesian Optimization for 50 Hz harmonic control. The chosen algorithm is described, together with details of integration in the CERN control system. The 2023 results are presented and complemented with an overview of the next steps.

Efficient control of frequency detuning for the radio-frequency quadrupole (RFQ) at the Facility for Rare Isotope Beams (FRIB) is still challenging. The transport delay and the complicated heat transfer process in the cooling water control system convolute the control problem. In this work, a long-short term memory (LSTM)-based Koopman model is proposed to deal with this time-delayed control problem. By learning the time-delayed correlations hidden in the historical data, this model can predict the behavior of RFQ frequency detuning with given control actions. With this model, a model predictive control (MPC) strategy is developed to pursue better control performance.

In accelerator labs like GSI/FAIR, automating complex systems is key for maximizing physics experiment time. This study explores the application of a data-driven model predictive control (MPC) to refine the multi-turn injection (MTI) process into the SIS18 synchrotron, departing from conventional numerical optimization methods. MPC is distinguished by its reduced number of optimization steps and superior ability to control performance criteria, effectively addressing issues like delayed outcomes and safety concerns, including septum protection. The study focuses on a highly sample-efficient MPC approach based on Gaussian processes, which lies at the intersection of model-based reinforcement learning and control theory. This approach merges the strengths of both fields, offering a unified and optimized solution and yielding a safe and fast state-based optimization approach beyond classical reinforcement learning and Bayesian optimization. Our study lays the groundwork for enabling safe online training for the SS18 MTI issue, showing great potential for applying data-driven control in similar scenarios.

This paper addresses the automation of particle accelerator control through reinforcement learning (RL). It highlights the potential to increase reliable performance, especially in light of new diagnostic tools and the increasingly complex variable schedules of specific accelerators. We focus on the physics simulation of the AWAKE electron line, an ideal platform for performing in-depth studies that allow a clear distinction between the problem and the performance of different algorithmic approaches for accurate analysis. The main challenges are the lack of realistic simulations and partially observable environments. We show how effective results can be achieved through meta-reinforcement learning, where an agent is trained to quickly adapt to specific real-world scenarios based on prior training in a simulated environment with variable unknowns. When suitable simulations are lacking or too costly, a model-based method using Gaussian processes is used for direct training in a few shots only. The work opens new avenues for implementing control automation in particle accelerators, significantly increasing their efficiency and adaptability.

Reinforcement Learning (RL) has demonstrated its effectiveness in solving control problems in particle accelerators. A challenging application is the control of the microbunching instability (MBI) in synchrotron light sources. Here the interaction of an electron bunch with its emitted coherent synchrotron radiation leads to complex non-linear dynamics and pronounced fluctuations. Addressing the control of intricate dynamics necessitates meeting stringent microsecond-level real-time constraints. To achieve this, RL algorithms must be deployed on a high-performance electronics platform. The KINGFISHER system, utilizing the AMD-Xilinx Versal family of heterogeneous computing devices, has been specifically designed at KIT to tackle these demanding conditions. The system implements an experience accumulator architecture to perform online learning purely through interaction with the accelerator while still satisfying strong real-time constraints. The preliminary results of this innovative control paradigm at the Karlsruhe Research Accelerator (KARA) will be presented. Notably, this represents the first experimental attempt to control the MBI with RL using online training only.

Reinforcement Learning (RL) is a unique learning paradigm that is particularly well-suited to tackle complex control tasks, can deal with delayed consequences, and can learn from experience without an explicit model of the dynamics of the problem. These properties make RL methods extremely promising for applications in particle accelerators, where the dynamically evolving conditions of both the particle beam and the accelerator systems must be constantly considered. While the time to work on RL is now particularly favorable thanks to the availability of high-level programming libraries and resources, its implementation in particle accelerators is not trivial and requires further consideration. In this context, the Reinforcement Learning for Autonomous Accelerators (RL4AA) international collaboration was established to consolidate existing knowledge, share experiences and ideas, and collaborate on accelerator-specific solutions that leverage recent advances in RL. Here we report on two collaboration workshops, RL4AA'23 and RL4AA'24, which took place in February 2023 at the Karlsruhe Institute of Technology and in February 2024 at the Paris-Lodron Universität Salzburg.

The Advanced Light Source (ALS) storage ring employs various feedback and feedforward systems to stabilize the circulating electron beam thus ensuring delivery of steady synchrotron radiation to the users. In particular, active correction is essential to compensate for the significant perturbations to the transverse beam size induced by user-controlled tuning of the insertion devices, which occurs continuously during normal operation. Past work at the ALS already offered a proof-of-principle demonstration that Machine Learning (ML) methods could be used successfully for this purpose. Recent work has led to the development of a more robust ML-algorithm capable of continuous retraining and its routine deployment into day-to-day machine operation. In this contribution we focus on technical aspects of gathering the training data and model analysis based on archived data from 2 years of user operation as well as on the model implementation including the interface of an EPICS Input/Output Controller (IOC) into a Phoebus Panel, enabling operator-level supervision of the Beam Size Control (BSC) tool during regular user operation.

In this work we explore the estimation and control of a particle accelerator simulation of the 800 MeV linac at Los Alamos National Lab. We use a convolution neural network model with a low dimensional latent space to predict the phase space projections of the beam and beam loss, which are mapped from accelerator settings. In deploying the model, we assume phase space predictions cannot be measured but beam loss can, and we apply a feedback using the error in beam loss prediction to tune the latent space. With beam loss and phase space predictions well correlated, we apply constrained optimization techniques, simultaneous with phase space prediction, to control the beam phase space while keeping beam loss from reaching unsafe levels.

We present a unifying approach to generative deep learning-based 6D phase space diagnostics which combines neural networks, physics models, and adaptive feedback. Our approach includes a physics-constrained neural network (PCNN) for calculating the electromagnetic fields of intense relativistic charged particle beams via 3D convolutional neural networks. Unlike the popular physics-informed neural networks (PINNs) approach, in which soft physics constraints are added as part of the network training cost function, our PCNNs respect hard physics constraints, such as ▽·B=0, by construction. Our 3D convolutional PCNNs map entire large (256×256×256 pixel) 3D volumes of time-varying current and charge densities to their associated electromagnetic fields. We demonstrate the method on space charge dominated, relativistic (5 MeV), short (hundreds of fs), high charge (2 nC) electron beams, such as those in the injector sections of modern free electron laser and plasma wakefield accelerators. We show that the method is accurate, respects physics constraints, and that the trained 3D convolutional PCNNs perform electromagnetic calculations orders of magnitude faster than traditional solvers which require a O(N^2) process for calculating the space charge fields of intense charged particle beams. We show how we combine this with an online physics model, adaptive feedback, and automatic differentiation for real-time predictions of the 2D projections of the 6D phase space of charged particle beams.

The Data Platform is a fully independent system for management and retrieval of heterogeneous, time-series data required for machine learning and general data science applications deployed at large particle accelerator facilities. It is an independent subsystem within the larger Machine Learning Data Platform (MLDP) which provides full-stack support for such facilities and applications [1]. The Data Platform maintains the heterogeneous data archive along with all associated metadata and post-acquisition user annotations. It also facilitates all interactions between data scientists and the data archive, thus it directly supports all back-end data science use cases. Accelerator facilities include thousands of data sources sampled at high frequencies, so ingestion performance is a key requirement and the current challenge. We describe the operation, architecture, performance, and development status of the Data Platform.

Osprey DCS is developing the Machine Learning Data Platform (MLDP) supporting data science applications specific to large particle accelerator facilities and other large experimental physics facilities. It represents a “data-science ready” host platform providing integrated support for advanced data science applications used for diagnosis, modeling, control, and optimization of these facilities. There are three primary functions of the platform: 1) high-speed data acquisition, 2) archiving and management of time-correlated, heterogeneous data, and 3) comprehensive access and interaction with archived data. The objective is to provide full-stack support, from low-level hardware acquisition to broad data accessibility within a portable, standardized platform offering a data-centric interface for accelerator physicists and data scientists. Osprey DCS has developed a working prototype MLDP* and is now pursuing full-scale development. We present an overview of the MLDP including use cases, architecture, and deployment, along with the current development status. The MLDP is deployable at any facility, however, the low-level acquisition component is EPICS based.

Advances in high-performance computing have enabled detailed physics simulations, including those with nonlinear collective effects such as space charge, to be deployed online in a control room setting to aid operator intuition and be used directly in automatic tuning. Simultaneously, machine learning (ML) has enabled deployment of detailed models online with sub-second execution time, opened up new avenues for adapting simulation models to more closely match real accelerator behavior, and enabled novel ways to combine detailed physics simulations and ML-based tuning. This contribution will provide an overview of how these tools are being developed and successfully applied at SLAC, with an emphasis on experimental demonstrations. This includes improvements in adaptive calibration methods, novel approaches to simulation (e.g. differentiable physics combined with ML), and the use of system models in ML-based tuning (e.g. Bayesian optimization with system model priors, iterative simulation and ML tuning to aid LCLS-II injector commissioning). Discussion of the software infrastructure required to achieve this and deploy these solutions into regular operation will also be discussed.

For a large ring, the response matrix contains tens of thousands of data points, allowing for a comprehensive representation of the ring’s linear optics. However, when it comes to fourth-generation diffraction-limited rings that utilize strong sextupoles and octupoles, the response matrix derived from closed orbit tracking is significantly influenced by nonlinearity. This nonlinearity poses challenges for the Linear Optics from Closed Orbit (LOCO) method in terms of matching lattice parameters and correcting lattice errors. In this study, we propose the utilization of an evolutionary algorithm that integrates multiple techniques to enhance exploration capabilities. By this, we aim to find the global optimal solution, which can effectively address the issues associated with the response matrix correction and achieve a larger dynamic aperture compared to the linear LOCO approach.

In our pursuit to tailor a precise electron bunch with a photoinjector, fine-tuning laser parameters, especially those influencing the photocathode pulse, is pivotal. Our ongoing research integrates machine learning, training neural networks with experimental data from ATF. The first approach involves generating a downstream photocurrent image to replicate the emission profile, serving as a fitness function for neural network training. The second approach employs an emittance scan during each iteration of the neural network-controlled laser profile, using magnetic optics and beam profile monitors, with calculated beam emittance as an additional fitness function. Our research aims to demonstrate the potential superiority of the neural network in achieving precise laser shaping for electron beam optimization. Leveraging real data, our goal is to reduce electron beam emittance through optimized laser profiles, underscoring the impactful applications of machine learning in advancing photoinjector technology.

The PERLE (Powerful Energy Recovery Linac for Experiments) project requires an injector capable of delivering a beam current of 20 mA at a total beam energy of 7 MeV with 500 pC bunches. These requirements present challenges for achieving the high quality beam required for the main ERL loop. At low energy and high bunch charge, the electron bunches will predominantly experience emittance growth due to the space charge effects. The compensation of this emittance growth will be performed with the traditional method of two solenoids a single bunching cavity and a linac to reach the intended injection energy. Additionally, the control of longitudinal and transverse bunch size must be performed to meet the requirements at the end of the injector. For stable operation of PERLE a rms bunch length of < 3 mm is required, with transverse emittances < 6 mm·mrad and acceptable transverse size. Presented here is the re-optimization of the injector settings used during commissioning for two alternative DC photoguns. It is found that the former cathode does not perform to the standard of previous optimizations. However, a newly procured cathode when optimized can meet the requirements for PERLE.

Photocathodes play key roles in supplying electron beams for diverse research facilities. Among them, semiconductor photocathodes stand out for their high quantum efficiency (QE). Typically, a high QE, long operation time, low thermal emittance and fast response time are desired for the accelerator community. However, the performance of semiconductor photocathodes is extremely sensitive to growth conditions. In this presentation, I will delve into recent advancements in semiconductor photocathodes fabrication at Applied Cathode Enhancement and Robustness Technologies (ACERT) of Los Alamos National Laboratory (LANL). These updates allow us to fine tune growth parameters and fabricate photocathodes with high QE and low emittance at high gradient to meet the requirements of photocathodes for Cathodes and Radiofrequency Interactions in Extremes (CARIE) project at LANL. Specifically, I will highlight our progress in developing a control system that enables to accurately control growth parameters. Furthermore, I will show our preliminary results focusing on the fabrication of CsSb and CsTe photocathodes using both sequential and co-deposition methods.

The Accelerator Operations and Technology division at Los Alamos Neutron Science Center (LANSCE) is working on designing and implementing an Instrumentation and Controls System (ICS) for the Cathodes and Radio-frequency Interactions in Extremes (CARIE) project. The system will utilize open-source Experimental Physics and Industrial Control System (EPICS) developed for scientific facilities for control, monitoring, and data acquisition. The hardware form factors will include National Instrument’s (NI) cRIO automation controller for industrial-like slow inputs/outputs and NI’s PXIe for high-speed data acquisition for diagnostic signals featuring masked and event-based time window capture. In this paper, we will discuss the reasons that led to the design, the hardware and software design specifics, the challenges that we faced during implementation, including the EPICS device support for NI PXIe, as well as the advantages and drawbacks of our system given the experimental nature of the CARIE project.

The Inverse Compton Scattering Interaction Point (ICS-IP) vacuum chamber provides a UHV environment where the electron and IR laser beams are overlapped in space and time to generate hard X-rays between 4 and 20 keV. The chamber has over two dozen motorized stages that position YAG screens with ~10 nm precision utilizing the EPICS framework for instrumentation interface. Using agile programming methods, MATLAB GUIs were created to control all the motors inside the chamber. Each YAG screen has a linear array of holes ranging between 10 microns and 2 mm that are imaged by cameras mounted on top of the chamber. Programmable focus lenses and IR mirrors are positioned to focus the IR laser at the interaction point. An X-ray optic is mounted onto a six degree of freedom nano-positioner enabling capture and collimation of X-rays coming from the IP. The X-ray optic can also be extracted from the beam path to transport the freely diverging X-rays to the experiment hutch for imaging experiments. We present the systems integration of the chamber, diagnostics elements, and control software and comment on its performance during instrument commissioning.

One major factor contributing to electron beam stability in a storage ring is the mechanical vibrations of magnets. At NSLS-II, we employ electromagnetic actuators to induce controlled vibrations in the support girders of the magnets. Beam position monitors distributed around the ring measure the spatial and frequency distribution of beam oscillations. The collected data is used to create and validate a computer model through a simulated commissioning tool, simulating beam motion caused by magnet vibrations. This computational model is useful for establishing mechanical stability specifications for the low-emittance upgrade of NSLS-II.

The Radio Frequency System-on-Chip (RFSoC) FPGA-based high-performance bunch-by-bunch beam position monitor (BxB BPM) was developed and commissioned at NSLS-II. The new BxB BPM features a 14-bit 5 Gsps ADC, directly sampling 2 ns four-button signals, and digital signal processing with a synchronized 500 MHz RF reference clock. The BxB BPM provides 32 K points of ADC raw data, 5 K turns for up to 1320 bunch amplitude and position data, 2.6 million turn-by-turn (TxT) data points, 10 K turns of circular buffer, and 10 Hz streaming data. The potential applications include, but are not limited to measuring injection transient, efficiency, ion instability detection, and single/multi-bunch motion analysis. A ~15 μm single-bunch resolution was confirmed with the beam test. This paper will present the beam test results, hardware FPGA firmware architecture, and control system interface for operation.

Efficient operation of the Large Hadron Collider (LHC) relies on accurate longitudinal beam measurements to diagnose beam instabilities and verify the correctness of bunch-shaping techniques. To achieve this goal, a diagnostic system was developed to perform high-resolution measurements of longitudinal bunch profiles. High-performance oscilloscopes, synchronized to precise accelerator events, are employed to carry out the measurements, acquiring data from wideband wall-current monitors installed in the machine. This paper provides details on the implementation of the system, highlighting its current and future applications that will play a key role in increasing beam intensity in the LHC.

A low-latency, real-time diagnostic system for the analysis of longitudinal Schottky signals in CERN’s antiproton chain has been developed. The system, installed in CERN’s Antiproton Decelerator (AD), processes the combined output of two low-noise, wideband AC beam transformers. It uses a GPU and the NVIDIA CUDA Toolkit, exploiting the directly sampled data and hardware features provided by the low-level radio-frequency (LLRF) VMEBus Switched Serial (VXS) system and its companion ObsBox server, to implement the FFT-based multi-harmonic spectral analysis needed to set up and monitor the stochastic and electron cooling processes. Longitudinal beam properties, such as mean momentum and momentum spread, are also derived to evaluate and log the machine performance. This paper describes the implementation of the system and its integration within the CERN control system, achieved using the Front-End Software Architecture (FESA) framework and a graphics co-processor directly installed in the Front-End computer (FEC), running a real-time operating system environment. Preliminary results of its usage in the Extra Low ENergy Antiproton (ELENA) ring and next steps to process bunched beam spectra are also presented.

To realize very precise measurement of the muon spin precession frequency in the level of sub-ppm, a muon beam is injected into a precisely adjusted storage magnet of sub-ppm uniformity via “Three-dimensional spiral beam injection scheme [1]” at J-PARC muon g-2/EDM experiment. This injection scheme requires a strongly X-Y coupled beam which is applied by eight rotating quadrupoles on the 10m of beam transport line [2]. Currently we have two scenarios of set of rotation angles (1) 45 or 60 degrees fixed, (2) any angles. In this presentation, strategy to precise control of the X-Y coupling at the beam transport line is discussed: how to control/monitor X-Y coupled phase space with eight rotatable quadrupole magnets including its alignment requirements for the case of (1) and (2). Results of alignment of the newly developed mount system for the rotating quad is also introduced. A pair of dedicated magnets called active shield multipole magnet (ASXM) will be set at the entrance and the exit of the beam channel of the storage magnet yoke. These devices will guarantee how well the beam phase space is matched with requirements at the reference point inside the storage magnet [3].

Ultrafast high repetition-rate laser systems are essential to modern scientific and industrial applications. Variations in critical figures of merit, such as focal position, can significantly impact efficacy for applications involving laser plasma interactions, such as electron beam acceleration and radiation generation. 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. We present the deployment and testing of this scheme at the BELLA Center at Lawrence Berkeley National Laboratory. Online optical adjustments are made to a telescopic lens to provide the desired correction on millisecond timescales. A framework for generating a low-level hardware description of ML-based correction algorithms on FPGA hardware is coupled directly to the beamline using the AMD Xilinx Vitis AI toolchain in conjunction with deployment scripts.

The r-process fission cycle terminates the synthesis of heavy elements in binary neutron-star mergers. Fission processes of transuranium nuclides will be studied in electrofission reactions at the thrice-recirculating electron accelerator S-DALINAC*. Due to the minuscule fissile target, the experimental setup requires an active beam-stabilization system with high accuracy and a beam position resolution in the sub-millimeter range. Requirements and concepts for this system regarding beam diagnostics elements, feedback control and readout electronics will be presented. The usage of a cavity beam position monitor and optical transition radiation screens to monitor the required beam parameters will be discussed in detail. Additionally, various measurements including a study of beam stability performed in the injector section of the S-DALINAC to assess requirements and limits for the beam-stabilization system will be presented. Finally, the application of advanced machine learning methods, such as neural networks and agent-based reinforcement learning, will be discussed.

As part of the International Muon Collider study, a beam vacuum window is being developed at CERN. It is required for the final cooling, where the charged particles travel from the vacuum chamber to the absorber; here, the beam loses momentum to cross a second window entering in a RF cavity that increases the longitudinal momentum. The best absorber for the final cooling is hydrogen. As the absorber should be installed inside a high field focusing solenoid, the hydrogen density should be as high as possible, ideally liquid or high pressure gas, to have a reasonable solenoid length. To evaluate the performance of the window, it is necessary to study the tightness at cryogenic temperatures, resistance to burst, high temperature and beam-induced damage. The main objective of the proposed work is to design and validate a versatile bulge test setup for the mechanical characterization of thin windows at different pressures and temperatures to cover all operating conditions, from 77 K to 293.15 K and ideally above. Due to the low thicknesses, a non-contact measuring technique based on a confocal chromatic sensor is proposed.

During the last years of operation of the COSY facility, significant improvements were made in beam control and diagnostics. Many systems have been upgraded from a Tcl/Tk based control system to EPICS. One of the advantages of EPICS is the coherent communication via Process Variables (PVs). This allowed us not only to control the synchrotron and its injection beam line (IBL) through GUIs but also allowed us to control the beam with algorithms. In our case, these algorithms covered a range of applications from variation of the currents of the electromagnets up to more advanced techniques of AI/ML such as Bayesian Optimization or beam control with Reinforcement Learning. Due to the unified nature of the PVs, the algorithms can be fed with a plethora of input parameters such as beam positions, beam current, or even live images of a camera. Depending on the algorithm, it is also possible to switch the target quantity (e.g. from measured current at the beam cups to the intensity of the injected beam at COSY). The algorithms can also trigger model calculations and access their results, if desired. We present an overview of different applications and our efforts to prepare COSY for them.

A surface treatment device has been established at the Wuxi Platform, enabling chemical polishing treatment on coupon samples. Currently, several samples treated with buffered chemical polishing (BCP) have been utilized in the investigation of nitrogen doping and medium-temperature baking mechanisms. This paper presents the development process of this device along with the experimental outcomes. In the future, we plan to enhance the device to facilitate electropolishing (EP) treatment on coupon samples.

The magnetic measurements laboratory of the Frascati National Laboratories of INFN is one of the pole of the Innovative Research Infrastructure for applied Superconductivity (IRIS). This infrastructure aims at upgrading laboratories to carry out basic research on magnetism and superconducting materials, test of superconducting magnets, wires, tapes, cables. The LNF pole will be devoted to testing SC coils and magnets at room temperature. These measurements are recommended during the manufacturing phase, since they allow the validation of the assembly and the detection of defects at early stages of production, before the cryogenic tests are carried out. Part of the equipment is already available, including a stretched wire bench, a rotating coil system, a NMR probe, gaussmeters, instruments for high precision electrical measurements. The IRIS upgrade will include a 3D Hall probe mole system, a pulsed wire bench, a 5-axes coordinatometer, high-stability power supplies of various sizes, a calibration system. The flexibility of the instruments will allow to cover a large range of magnetic measurements, from point maps to integrated fields, from multipolar analysis to fiducialization.

Thomas Jefferson National Accelerator Facility (Jefferson Lab) currently studies the feasibility of upgrading its energy to 22GeV. It considers addition of six more linac passes. The highest energy passes will share two new arcs designed using the Fixed-Field Alternating Gradient (FFA) technology. The FFA arcs are built using permanent combined-function magnets. They will be connected to the linacs through transition sections that will match the optics of all six passes to the linacs. With the high number of constraints and the limited space available, we are investigating a parametric resonance technique to match the optics quasi-independently at each energy. A resonance is excited at each individual energy to selectively control its optics. The resonant dipole and quadrupole kick harmonics are imposed for all energies simultaneously using Panofsky corrector magnets placed throughout the FFA arcs. This paper presents the current progress on that transition section design.

Our research focuses on the design of a beamline. Due to the numerous beamline components involved, without strict optimization of each component's parameters, the transmitted temporal profile of beam may distort, failing to meet the expected requirements. Additionally, different initial temporal profile of the beam will undergo longitudinal shaping during transmission through the beamline. Therefore, we aim to determine the combination of initial beam temporal profile at the cathode and the parameters of the beamline components based on the specific beam distribution at the exit. We propose the application of an improved multi-objective genetic algorithm to solve this problem. Through multiple optimization iterations for a given temporal profile, our algorithm consistently identifies multiple suitable combinations of initial beam temporal profile and beamline component parameters to produce the desired specific temporal profile of the beam.

The automation upgrade of the photoinjector for the Compact X-Ray Light Source (CXLS) at Arizona State University is described. As the accelerator vault of the CXLS is only 10 meters long, the photoinjector drive laser is located in an enclosure inside the vault. Since ionizing radiation is present in this room during operations, it necessitates remote control of all devices used to optimize the laser spot. This includes multiple shutters, Galil motors, picomotors, a mirror flipper, LEDs, and remote lens controllers. To actuate these devices, a GUI was created with the use of MATLAB AppDesigner which communicates with the hardware through EPICS (Experimental Physics and Industrial Control System). Challenges with this GUI are described, along with the team’s efforts to finalize the control software. After these upgrades, the photoinjector laser characteristics can be adjusted remotely during operation and changes to the drive laser’s position, shape, and intensity can be made without interrupting beam time.

MEDICYC (MEDical CYClotron) is an isochronous cyclotron dedicated to radiotherapy which was built and commissioned in Nice, France, in 1990 by a local team aided by experts from CERN. The cyclotron accelerates H- to a maximum energy of 65 MeV and uses stripping to extract a proton beam. Its primary purpose is treating ocular melanoma by protontherapy but a significant research activity is also present on beam-lines dedicated for this purpose. An extensive refurbishment program of the cyclotron has been started to cope with the end-of-life and/or the obsolescence of several sub-systems. In this context, a new high-level cyclotron control system has been developed and implemented in 2021-2023. The primary responsibility of the system is the high-level coordination of the H- source, the RF system and the beam-line and cyclotron magnets to produce and deliver a beam with a given set of characteristics. A secondary responsibility is the collection, visualization and analysis of sub-system and beam data for monitoring and pre-emptive fault detection. In this contribution, the control system software architecture is presented and the infrastructure on which the systems are deployed is laid out.

At the tandem ion accelerator laboratory of the Jožef Stefan Institute (JSI) in Ljubljana, Slovenia we are developing a control system for the Low Energy Branch (LEB) ion beamline. This activity is ongoing simultaneously with the hardware construction of the ion beamline branch dedicated to the research with low-energy ion beams with energies up to 30 keV. The LEB instrumentation is categorized into: a) Ion sources, b) Ion beam transport optics, and c) Accessories, including specialized detector systems and devices, used to prepare and maintain optimal experimental conditions. Therefore, key functionalities of the control system include the control of devices like vacuum pumps, power supplies, etc., data acquisition from sensors and detector systems, and ensuring reliable autonomous operation for high-precision physics experiments [1]. The control system will be implemented within the Experimental Physics and Industrial Control System (EPICS) environment [2], providing us with the tools required to develop a comprehensive and scalable control system. In this work, we present a block scheme, a device list, the prototype control system architecture of a minimal control system prototype currently operational in our laboratory.

This paper presents a cost-effective asset management system (AMS) designed to optimize the workflow of the accelerator control system for the Advanced Light Source Upgrade (ALS-U) project at LBNL. The AMS stores all essential information about equipment, including location, owner, hardware details, and firmware versions. Its user-friendly interface provides consistent access throughout the equipment lifecycle, from quality assurance to installation, through label printing, QR codes, and the Web application. By streamlining workflows and improving data consistency, the AMS contributes significantly to the efficiency and success of the ALS-U project.

Transitioning from the aging ALS Geo MACRO motor controller, this paper details the meticulous selection process for a new, cost-effective standard to fulfill the diverse motion requirements of the upcoming ALS-U project. Targeting primarily simple stepper motors with varying current demands, the chosen solution seamlessly integrates into the existing ALS-U EPICS environment while preserving the established ALS motion architecture and EPICS IOC support. Notably, the solution maintains independence from Delta-Tau technology while comprehensively accommodating the project's required range of servo/stepper motor types and offering dedicated support for critical subsystems like Beam Scraper and Cold Finger motion. This document delves into the exhaustive selection process, from comprehensively summarizing the current architecture and ALS-U requirements to meticulously analyzing the results of a year-long evaluation of diverse vendor offerings.

This paper presents the design and status of Accumulator Ring (AR) RF Equipment Protection System (EPS) of Advanced Light Source Upgrade project at LBNL. The key components of AR RF EPS include a Master Interlock PLC subsystem handling supervisory control and slow interlocks in ms scale, an FPGA-based LLRF Controller managing fast interlocks in µs scale, a 60 kW high-power amplifier with standalone PLC-based slow (ms scale) and FPGA-based fast (µs scale) protection systems, and an RF Drive Control Chassis acting as primary RF mitigation device. The design of AR RF EPS is presented along with internal RF and external AR subsystems interfaces.

The LANSCE accelerator is an 800 MeV linear accelerator delivering beams for more than fifty years. As it has aged, maintenance and upgrades to its control system software components have become challenging and often deferred due to operational and schedule constraints. As a result, we have a wide variety of new and old software, difficult to re-use, with a large staff burden. Data is stored in redundant sources, inconsistent formats, and outdated technology. Multiple tools exist for the same tasks. Some production software is updated without proper processes. We describe our approach to modernizing LANSCE control system software with proper development processes. We consider reduction of diversity, redundancies, data sources. Migration to modern technologies is also discussed. We explore the possibility of language standardization, and describe our database implementation and other future plans. Lifecycle management is also considered. This years-long effort will utilize a risk-based strategy to address the most urgent issues while also ensuring steady progress, ultimately resulting in a coherent and maintainable suite of control system software.

The elliptically polarized undulator with a period length of 56 mm, called EPU56, is part of the Taiwan Photon Source (TPS) phase-III beamline project. Its control system is built within the EPICS framework using motion controllers and EtherCAT. The control systems of EPU56 include a safety interlock system, which automatically stops movement based on limit switches, torque limit switches,emergency stop button, and readings from the enclosed linear optical encoder. In addition, the control system offers settings for adjusting the correction magnets' power supply and employs optical absolute encoder motors to control the movement of the Gap and Phase. In order to maintain stability during movement, PID control is applied to the motion process by the motion controller. To further enhance precision, the system also employs an integrator limit within the motion controller for additional adjustments. This paper describes the development of the control system and the enhancements made to the insertion device movement process.

The TPS (Taiwan Photon Source) is a third generation 3 GeV synchrotron light source. Some beamlines use synchrotron pulses in conjunction with laser pulses for pump-probe experiments, which is a time-resolved experiment method for capturing the temporal evolution of the pumped process. Periodic X-ray pulses are provided by the synchrotron light source as detecting light (Probe), and laser pulses can be used as a pump to excite a target, which changes a certain property when excited. Pump-probe experiments re-quire a synchronized laser system to alter the delay time between X-ray pulses and laser pulses. It has been built a laser synchronizer and timing support system. One direct digital synthesizer (DDS) with fine delay adjustment can change the laser pump pulse relative to the X-ray pulse. The clock fanout buffer with output dividers provides the synchronized clocks required by the laser oscillator and laser source. An SBC (single-board computer) is employed as a control interface The software architecture is created using the EPICS framework, which is compatible with the TPS control system, and a GUI with the ability to adjust the time delay is created. The efforts will be described in this report.

Data acquisition (DAQ) is an ubiquitous feature in modern particle accelerator measurement and control systems. At the Paul Scherrer Institut (PSI), a next generation of electronic devices is being designed to meet the demands of upcoming renewal of facilities. The new developments utilize the CompactPCI Serial (CPCI-S.0) platform, and will cover a diverse set of applications, including Low Level Radio Frequency (LLRF), Longitudinal Beam Loss Monitoring (LBLM), and Filling Pattern Monitoring (FPM) systems. Careful design considerations and selection of an optimal architecture are crucial to fulfill a variety of DAQ requirements such as maximum frequency of acquisition, size of the data and different modes of triggering. In this contribution, we focus on the real-time DAQ implementations utilizing a multiprocessor system on chip (MPSoC) technology. We review the IP components developed in-house at PSI that provide the DAQ functionality. We demonstrate, that by reusing the IP components development, prototyping and testing of applications requiring the DAQ are accelerated.

Transverse bunch-by-bunch (BxB) feedback system has been designed, fabricated, installed, and tested with beam at the Advanced Photon Source Upgrade (APS-U) storage ring. The transverse feedback system (TFB) is designed to suppress coupled bunch instabilities and single bunch instabilities. It adapted a stripline kicker design which has the same profile as the APS-U injection/extraction kickers. The system uses digital controllers which provide powerful diagnostics, in addition to its major functionality for feedback control. This paper presents the status of the TFB system including early beam commissioning results.

Advanced Photon Source Upgrade (APS-U) Fast Orbit Feedback (FOFB) system uses 160 fast and 160 slow corrector magnets to stabilize orbit measured at 560 Beam Position Monitors (BPM). We plan to operate both fast and slow correctors in a unified feedback algorithm at 22 kHz correction rate. Mid-ranging control is a proven approach for feedback systems with two manipulated inputs each exerting distinct dynamic effects to regulate a single output. This method resets the fast input to its chosen DC setpoint and proves beneficial when cost of fast input is more than the slower one. Unified operation of fast and slow correctors is a fitting application to mid-ranging concept which is well founded for two input one output systems. In this work, based on the cross-directional nature of the FOFB system we developed a multi-variable approach to mid-ranging control. It can be applied to FOFB with multiple fast and slow correctors, and multiple BPMs. Performance of proposed scheme is tested in simulations with APS-U FOFB prototype model in MATLAB. The feedback loop with fast and slow correctors is stable with mid-ranging algorithm, and the fast corrector drives effectively tracked setpoints.

During the Long Shutdown 2 (LS2 2019-2021) of the LHC, the orbit feedback correction software (OFB) of the LHC was redesigned to satisfy new requirements for LHC Run 3 (2022-2025) and to clean up legacy functionalities. The OFB is an essential component of LHC high intensity operation since the orbit must be stabilized to a fraction of the beam size during the entire LHC machine cycle. Redesigning such an essential and complex system during shutdowns requires thorough testing of the system functionality. The existing OFB testing system has been reviewed and improved based on the experience of LHC Run 2. An automatic, continuous integration tool has been put in place to validate future software developments before putting them in production. The solution for the OFB testing will be presented in this contribution.

Charge stripping is inherent for high power ion accelerators such as the FRIB LINAC. However, at high power, strippers require motion to prolong the operational life of the stripping media, or by flowing a liquid Lithium film. The charge stripping process introduces energy losses that vary with the actual Lithium film thickness, which can result in observable beam losses along the tuned beamline at high on-target beam power, above ~100 kW, if not adequately mitigated. BPM phase feedback is used in real-time to compensate for these effects, controlling upstream RF cavities in order to maintain a constant beam energy and phase post-stripper, which significantly reduces beam energy fluctuations.

In the SIS18 heavy-ion synchrotron at GSI, RF beam phase feedback systems are developed and tested with beam for the damping of coherent longitudinal bunch oscillations. In particular, a beam phase control system is currently commissioned for the damping of longitudinal dipole oscillations. The feedback system has to cope with both, a large RF frequency span (400 kHz to 5.4 MHz) and synchrotron frequencies of up to 6 kHz. It has to be compatible with several beam manipulation schemes such as dual-harmonic, bunch merging, and bunch compression. The system relies on recent upgrades of the SIS18 LLRF topology including a newly developed multi-purpose DSP system that is used for the RF cavity synchronization as well as for RF beam feedback. This paper describes the LLRF concept of the RF beam phase feedback at SIS18 and presents results from machine experiments with beam where an adaptive feedback filter for damping longitudinal dipole oscillations during the whole SIS18 machine cycle was realized and successfully applied. Finally, an outlook will be given towards the full integration into the central control system and towards the SIS100 bunch-by-bunch longitudinal feedback system.

The Full Remote Alignment System (FRAS) is an alignment system remotely controlled and monitored that comprises almost one thousand permanent sensors distributed along the 200 meters of equipment that will be installed in the frame of the High Luminosity LHC (HL-LHC) project on either side of the ATLAS and CMS detectors. The sensors, along with their electronics and a system of motorized actuators, will be used to adjust the relative positions of the components remotely, in real time, with no human intervention needed in the irradiated environment of the tunnel. In this contribution we describe the design and the implementation of the position estimation algorithm which is a core-component of the FRAS. This algorithm will process the data provided by all the sensors to determine exact positions and orientations of the associated components in real-time. The position estimation module is designed as a reusable C++ library and builds on the existing CERN LGC, a modular least-square software. It will be fully integrated into the FRAS software stack and is entirely file-less during operation. In this paper we will demonstrate its performance in a realistic case study and showcase its ability to provide position updates on a much higher frequency than the required 1 Hz.

The novel Coupling-Loss-Induced-Quench (CLIQ) concept will be part of the quench protection system of the High Luminosity Large Hadron Collider (HL-LHC) Inner Triplet superconducting magnets at CERN. Several units of two distinct CLIQ prototype variants were produced to validate the CLIQ novel protection concept and define the system parameters for the required performance. Subsequently, these units were further enhanced by introducing additional redundancy, advanced monitoring systems, and improved safety features. These improvements culminated in the development of the third and final version. This paper provides insights into the evolution from prototypes to the final version to be installed in the machine, shedding light on the outcomes of comprehensive safety and electromagnetic compatibility (EMC) tests, coupled with extensive operational assessments.

The Quench Heater Discharge Power Supplies (HDS) are magnet protection devices installed in the Large Hadron Collider (LHC) that, upon detection of a magnet quench, release energy into the copper-plated stainless-steel strip heaters, inducing a resistive transition all along the superconducting coils. Such a distributed internal heating ensures an even energy dissipation across the entire volume, preventing local overheating and magnet damage. Over 6000 HDS units have been operational in the LHC tunnel since 2007. The new HDS design for protection of the High Luminosity LHC (HL-LHC) Inner Triplet magnets, to be installed in the Long Shutdown starting in 2026, calls for a more advanced design with up-to-date components resulting in a higher reliability of the HDS units. Several HDS prototypes were produced at CERN, eventually culminating in the development of the HL-LHC HDS version to be installed in the accelerator. This paper describes the design of the upgraded HDS units and the comprehensive safety and electromagnetic compatibility (EMC) tests, coupled with extensive operational tests, including irradiation tests, that have been conducted.

The Safe Machine Parameter (SMP) system is an electronic hardware-based system which has been an integral part of the LHC’s machine protection strategy since it started operation. Its primary objective is to provide several parameters and interlock signals to critical machine protection users across the LHC and SPS accelerators, whilst prioritizing high reliability and availability. After almost two decades of operation, there is a need to upgrade the SMP hardware electronics. In the High Luminosity LHC era the requirements of connected systems have changed, leading to new system functions and operational requirements which must be integrated into the new design. This paper details the electronic design considerations of developing the second-generation SMP. The general distribution of parameters relies on the CERN WhiteRabbit timing network renovation, for which dedicated high-precision clock components were selected and tested on a prototype board. Details of the hardware design and validation are discussed, along with the comprehensive upgrades aimed at delivering an SMP system with expanded monitoring and diagnostic features.

In the context of LATINO (Laboratory in Advanced Technologies for INnOvation) and Rome Technopole Projects founded by Regione Lazio and NextGenerationEu, and directly involved in the EuPRAXIA@SPARC_Lab flagship project, a testing facility for X-band (TEX) has been established at the Frascati National Laboratories of INFN. TEX is dedicated to the examination of radiofrequency X/C-band, aiming to develop and test the technologies and systems of a particle accelerator operating under such conditions. Given the complex nature of such a system and the advancement of technology to the forefront of the state of the art, it is imperative to have an advanced Machine Protection System (MPS) characterized by high reliability, availability, and safety, in accordance with IEC-61508 standards. Currently in development is a prototype MPS designed to autonomously initiate procedures to control operations and avert anomalies. An EPICS supervisor oversees the management of all devices and monitoring connected subsystems. Additionally, a real-time interlock system, based on distributed FPGA, is employed to swiftly respond to vacuum and RF systems during the next RF pulse.

SIRIUS is a 4th generation synchrotron light source built and operated by the Brazilian Synchrotron Light Laboratory (LNLS). Recently, investigations of noise sources and the storage ring RF plant identification enabled a fine-tuning of the Digital Low-Level Radio Frequency (DLLRF) parameters. This paper presents the main improvements implemented, which include the mitigation of 60Hz noise from the LLRF Front End and the optimization of the control system parameters. Optimizations in the machine were based on an adjusted model of the SIRIUS storage ring RF plant. Tests with the model's parameters showed that the system's stability was strongly dependent on phase shifts introduced by nonlinearities from the high power RF sources. The new parameters significantly improved the control performance, increasing the bandwidth of the system and reducing longitudinal oscillations. BPM (Beam Position Monitor) and BbB (Bunch-by-Bunch) systems were employed to quantify longitudinal beam stability improvements.

In 2023, the KEK-PF 2.5 GeV ring LLRF system was replaced from a conventional analog to an FPGA-based digital system. The hardware and software of our digital LLRF system were developed by customizing the LLRF technologies established at the SPring-8 and J-PARC. In our system, we adopted the non-IQ direct sampling method for RF detection. We set the sampling frequency at 8/13 (307.75 MHz) of the RF frequency, where the denominator (13) is the divisor of the harmonic number (312) of the storage ring. This allows us to detect the transient variation of the cavity voltage that is synchronized with the beam revolution. To compensate this voltage variation, we plan to implement a feedforward technique. These functions will be useful to improve the bunch lengthening performance in a double RF system for KEK future synchrotron light source. The new digital LLRF system has been already installed and used for the user operation. At the nominal beam current of 450 mA, the variation of the cavity voltage amplitude and phase were within ±0.06% and ±0.06°, respectively. In this presentation, we introduce the details of our new system and report on the commissioning results.

An accurate physics simulation model is key to accelerator operation because all beam control and optimization algorithms require good understanding of the accelerator and its elements. For the AGS Booster, discrepancies between the real physical system and online simulation model have been a long-standing issue. Due to the lack of a reliable model, the current practice of beam control relies mainly on empirical tuning by experienced operators, which may be inefficient or sub-optimal. In this work, we investigate two main factors that can cause discrepancies between simulation and reality in the AGS Booster: magnet misalignments and magnet transfer functions. We developed a orbit response measurement script that collects real machine data in the Booster for model calibration. By matching simulated data with real data, we can develop a more accurate simulation model for future polarization optimizations, and build the foundation for a fully functional digital-twin.

Beam emittance plays the crucial role in a Beam transportation system. At a fixed-target beamline off the AGS Booster Synchrotron, beam emittance is determined through measuring the beam width via a segmented multi-wire ion chamber (SWIC) and varying quadrupole strength. The width of the beam signal (as Full Width Half Max) on the SWIC passes through a minimum value and the resulting dataset of FWHM per magnet current is used to fit a function. Using this technique, new controls software has been developed to set up measurements, acquire data, and perform analysis through a python-based scripts to calculate the emittance along the NASA Space Radiation Laboratory (NSRL) beamline. Initial results of the program are presented to for various points along the beamline in a variety of conditions.

Many accelerator control rooms rely on envelope models to simulate beam dynamics because they are fast and accurate at tracking the beam core. Particle-in-Cell (PIC) models, however, can track particles inside and outside the core and, with the improvements of computers, are now fast enough to be used in control rooms. The Spallation Neutron Source (SNS) at Oak Ridge National Laboratory is currently developing a tool to use a Particle-in-Cell model for control room applications. This report covers the progress so far and the future goals of using PyORBIT, a Particle-in-Cell simulation model, in the SNS control room.

The recent development of advanced black box optimization algorithms has promised order of magnitude improvements in optimization speed when solving accelerator physics problems. These algorithms have been implemented in the python package Xopt, which has been used to solve online and offline accelerator optimization problems at a wide number of facilities, including at SLAC, Argonne, BNL, DESY, ESRF, and others. In this work, we describe updates to the Xopt framework that expand its capabilities and improves optimization performance in solving online optimization problems. We also discuss how Xopt has been incorporated into the Badger graphical user interface that allows easy access to these advanced control algorithms in the accelerator control room. Finally, we describe how to integrate machine learning based surrogate models provided by the LUME-model package into online optimization via Xopt.

The conditioning of room temperature cavities is an exhausting process. To prevent damage to the cavity and auxiliary equipment, this potentially long process needs constant supervision or extensive safety precautions. Additionally, the unpredictability of every new conditioning makes the development of effective classical algorithms difficult. To reduce the workload for everyone involved and to increase the efficiency of the conditioning process, it was decided to develop a machine learning algorithm with the goal of fully automated conditioning in mind. To reach this goal, it is planned to train the model on the data of already conducted conditionings of room temperature cavities, a virtual cavity and several more conditionings to be conducted soon. In this paper, the status of development, problems and challenges as well as the planned future progression shall be summarized.

The Institute for Applied Physics (IAP) installed a permanent test setup for up to \SI{60}{\kilo\watt} of RF power in cw mode for conditioning. The goal is to establish a time efficient test procedure for the future MYRRHA CH-cavities. Three test stands were designed to accommodate up to three cavities simultaneously. All stations are curently tested via the \SI{175}{\mega\hertz} MAX RFQ prototype at IAP with the new test setup. Vacuum, Low-level-, and high-power measurements have been successfully performed.

Additive manufacturing ("AM") has become a powerful tool for rapid prototyping and manufacturing of complex geometries. A 433 MHz IH-DTL cavity has been constructed to act as a proof of concept for direct additive manufacturing of linac components. In this case, the internal drift tube structure has been produced from 1.4404 stainless steel, as well as pure copper using AM. We present the most recent results of vacuum, low level RF, as well as RF conditioning of the cavity.

Radio Frequency Knock Out (RF-KO) resonant slow extraction is commissioned at the Cooler Synchrotron (COSY) Jülich for the first time to extract the stored beam and deliver spills with constant particle rates to the experiments. Therefore, transverse RF excitation generated with a software-defined radio is applied to control the extraction rate. A built-in feedback system adjusts the excitation amplitude to maintain the desired extraction rate. To suppress fluctuations of the particle rate on timescales of milliseconds and below, an optimization algorithm is used to tune the RF excitation signals. The method was used extensively during the final run of COSY in 2023, reliably delivering stable beams to various users.

Electronic logbooks contain valuable information about activities and events concerning their associated particle accelerator facilities. However, the highly technical nature of logbook entries can hinder their usability and automation. As natural language processing (NLP) continues advancing, it offers opportunities to address various challenges that logbooks present. This work explores jointly testing a tailored Retrieval Augmented Generation (RAG) model for enhancing the usability of particle accelerator logbooks at institutes like DESY, BESSY, Fermilab, BNL, SLAC, LBNL, and CERN. The RAG model uses a corpus built on logbook contributions and aims to unlock insights from these logbooks by leveraging retrieval over facility datasets, including discussion about potential multimodal sources. Our goals are to increase the FAIR-ness (findability, accessibility, interoperability, and reusability) of logbooks by exploiting their information content to streamline everyday use, enable macro-analysis for root cause analysis, and facilitate problem-solving automation.

The following paper embarks on an in-depth exploration of extraction kickers employed at some of the most renowned particle physics and neutron science facilities worldwide. Specifically, we delve into the extraction kickers utilized at the Spallation Neutron Source, Fermi National Accelerator Laboratory, Los Alamos Neutron Science Center, and delve into the novel inductive adder structures. These facilities represent the forefront of scientific research, housing state-of-the-art technologies and extraction kicker systems that play a fundamental role in advancing our understanding of particle physics, neutron science, and related domains. Throughout the paper, we will investigate the design principles, operational intricacies, and technological innovations associated with these extraction kickers. By analyzing existing research and scholarly works, we aim to provide a comprehensive overview of the unique challenges and advancements encountered at each facility.

As one of the state-of-the-art radiotherapy approaches, proton therapy possesses conformal dose profiles yet expensive cost. Designing a facility with a small footprint and a high treatment efficiency is the main goal for researchers to fulfill the potential of proton therapy and make it more affordable both for vendors and patients. In this contribution, the design of a light-weight proton therapy gantry based on the alternating-gradient canted-cosine-theta (AG-CCT) super-conducting (SC) magnet is presented. The AG-CCT magnets adopt large bores and combined function design. With fine field harmonic control and fringe field shape optimization of the magnets, the multi-particle tracking results prove that the gantry achieves a momentum acceptance of ±8%. So that the full energy range from 70 to 230 MeV can be covered with merely 3 field switch points. Combined with a fast degrader component, whose switch time is below 50 ms, the energy modulation speed can be greatly fastened. To fully utilize the advantages of the large momentum acceptance gantry, the energy spread of the proton beam is expanded and a reduced treatment plan is proposed. Compared with the standard treatment plan, the energy layers number of a prostate case is reduced by 61.3% with comparable plan quality. In summary, the proposed gantry has significant superiority both in manufacture and clinical aspects.

To rule out Solid-State Power Amplifier (SSPA) modules with defects due to handmade and reduce time cost of maintenance for deployed modules, it is essential to establish a comprehensive testing platform that includes a complete quality control system. In this study, we developed a platform with function of manipulating driving power and shutting down when failures are detected.

The installation of the STAR High-Energy Linac, the energy upgrade of the Southern European Thomson Back-Scattering Source for Applied Research (STAR) project at the University of Calabria, was conducted by INFN by the end of 2023. This paper presents the testing procedures aimed at confirming the consistency, completeness, and quality of the STAR accelerator upgrade installation (electron beam energy boost from 65 MeV up to 150 MeV). We illustrate the installation and testing of the electrical, hydraulic and related automation and auxiliary systems. We will discuss the high-power commissioning of the two C-band RF power stations and testing of the low-level C-band RF system and control system configuration based on EPICS. Finally, we will describe the layout and testing of the vacuum system, the characterization and commissioning of the magnets with related power supplies and the assessment of the installed diagnostics devices. The linac commissioning as well as electron beam measurements are planned for Summer 2024, due to pending radioprotection authorizations.

The APS Upgrade (APS-U) multi-bend acromat storage ring requires 1000 high-stability unipolar magnet power supplies. A precision current measurement and calibration system has been developed to independently measure the power supply output current to ensure the accuracy and repeatability of the supplies. The measurement system uses custom commercial DCCT current transducers along with APS-U-designed electronics. The calibration system is designed to perform on-demand calibration of all 1000 DC measurement channels simultaneously using a single current reference source instrument. The calibration system includes a precision current multiplier and impedance buffer based on a novel use of DCCT technology that provides a local precision calibration current for up to 6 DCCTs in series through multi-turn low impedance calibration windings. All system components have been received and passed acceptance testing; the full system is currently being installed in the new storage ring and full-scale evaluation will begin in early 2024. This paper describes the system design and presents preliminary test results.

The High Energy Photon Source (HEPS) is the fourth-generation synchrotron photon source. Compared with the third-generation synchrotron photon source, the brightness is 100-1000 times higher, and the electron emittance of the storage ring is low to the diffraction limit of light. Through physical calculations, it is required that the stability of the storage ring quadrupole magnet power supply be better than 10ppm, and the accuracy of output current be better than 80ppm. This high demand for technical parameter poses a challenge to the development of high precision and stability power supplies. The main circuit topology of the power supply adopts a phase shifted full bridge soft switching scheme, which avoids interference caused by switching noise and improves power stability and efficiency. The high-precision digital power supply controller based on FPGA improves the sampling speed and control accuracy of the power supply, and the constant temperature control circuit ensures that the output current of the power supply meets the requirements of HEPS for power supply performance. In the batch testing section, a testing facility was built to test the stability, accuracy, repeatability, voltage ripple, and other parameter of high precision and stability power supplies. After a year and a half of testing, the performance tests of 1066 power supplies, including linear accelerators power supplies, booster power supplies, storage rings power supplies, dipole and quadrupole combined power supplies, dipole and quadrupole power supplies, were completed. The results all met and exceeded the design specifications. The HEPS high precision and stability power supply meets the design requirements in terms of current stability, accuracy, repeatability, voltage ripple, and other aspects. The batch test results show that the power supply performance using the full bridge phase shifting soft switching technology combined with high-precision digital controller scheme is excellent, and the power supply consistency is good, providing a guarantee for the successful operation of HEPS in the future.

The high-energy photon source (HEPS) under construction in Beijing is an excellent photon source with an emissivity better than 60pm rad. HEPS adopts on axis injection. The fast pulse power supply for booster injection adopts a topology structure of LC resonant discharge based on heavy hydrogen thyristor. The energy storage scheme of pulse capacitors adopts a design scheme of DC charging. The local control station of the fast pulse power supply for the enhancer is mainly responsible for the timing control, charging control, interlock control, protection of the kicker, and remote control. Fast pulse power supplies have high reliability, which poses challenges to the development of local control stations for fast pulse power supplies. The local control station adopts a high-performance programmable logic controller (PLC) as the control core, and applies standard modbus and ethernet for communication protocol to control equipment. A local control station prototype has been built. Through system joint testing, the designed local control station can achieve power control and protection, remote control of the local station, and interlocking protection of the magnet power supply.

Nowadays, the real-time control is more and more popular in the particle accelerator field because it is a powerful tool for stable operation and beam loss suppression in the particle accelerator. However, in the Japan Proton Accelerator Research Complex (J-PARC) Main Ring (MR), real-time control has not been widely used in magnet power supplies yet. Magnet power supplies are very easily affected by disturbances from external factors, such as environmental temperature, device aging, power grid voltage and current fluctuations, and so on. Therefore, it is worth developing a real-time digital controller with general functions for the magnet power supplies to observe and suppress these disturbances. In this paper, we propose the design of a general-purpose intelligence controller for the magnet power supply realized by a System-on-Chip (SoC) Field Programmable Gate Array (FPGA). This digital controller can also be used as a high-resolution data acquisition system, a pattern generator, and a high-precision current control system for magnet power supplies.

This paper presents a study on the upgrade and modernization of the magnet power supplies of the KARA (Karlsruhe Research Accelerator) storage ring. The existing power supplies, which have been in operation for more than two decades, were facing obsolescence and operational limitations. To ensure the continued availability and reliability of the facility for the next decade and beyond, a comprehensive refurbishment was required. The project involved the replacement and upgrade of the power supplies for the dipole and sextupole magnets at KARA, as well as for the dipole and quadrupole magnets in the booster. A key aspect of this modernization effort beside an improvement in efficiency and stability is the migration from a custom control system to EPICS running embedded on the power supplies. This paper provides an in-depth analysis of the motivations, goals, and technical aspects of the power supply modernization project as well as first measurements with the new power supplies and the project status.

This paper investigates the design and implementation of a TPS correction magnet power supply using a combination of a linear power operational amplifier (PA05) and a pre-regulator voltage controller. The PA05 linear power operational amplifier features bipolar output, high internal power dissipation, and wide bandwidth. Utilizing a DCCT sensor as a current feedback element integrated with the pre-regulator voltage controller, a closed-loop current modulation circuit is formed, providing the variable voltage required by the linear power operational amplifier. We have successfully developed a prototype of a linear power operational amplifier power supply with a pre-regulator voltage controller for TPS correction magnets through these measures. Design validation is achieved through control loops, resulting in fast and stable output current performance. The output current ripple is maintained below 100 μA, and the rise time during the step response is 75 μs. During the frequency response test using a 0.1 V interference signal, the gain margin remained within -3 dB at an 11.2 kHz bandwidth, and the phase margin was within -45 ° over a range of 5.1 kHz. The long-term stability of the output current is maintained at ten ppm. Finally, a hardware prototype circuit is assembled in the power laboratory with input voltage ranging from ±24 V, an output current of ±20 A, and a maximum rated power of 240 W.

In this paper, the focus is on the development of a bipolar high-current correction magnet power supply for the future TPS-II permanent magnet corrector coil. The maximum output current of the prototype is speci-fied as 20 A, operating at a voltage of 48 V. This con-figuration enhances the amplitude of the trim magnetic core correction magnetic field, thereby providing greater flexibility in manufacturing the permanent magnet corrector coil. The Danisense DP50-IP-B DCCT is the current feedback component to design a power supply with high current and stability. MOSFETs are configured in a full bridge setup serving as power switches. The driving frequency is set at 40 kHz. Analogue modulation control circuitry and pro-tection circuits ensure precise current control loop modulation. Finally, a hardware prototype circuit is constructed in the power supply laboratory with an input voltage of 48 V, an output current of 20 A, a maximum power of 960 W, and the current ripple com-ponent maintained within 400 μA. This validates the control loop design of the prototype, demonstrating the capability to achieve rapid and stable output cur-rent performance. The small-signal bandwidth tested using a 1V input reference signal shows a -3 dB band-width of 8.51 kHz. Long-term current stability is with-in ±10 ppm, and the interface is compatible with exist-ing TPS correction magnet power supply interfaces, allowing for direct operation within the current sys-tem.

The Linac Coherent Light Source II (LCLS-II) at the SLAC National Accelerator Laboratory represents a groundbreaking advancement in the realm of Free Electron X-Ray Laser (XFEL) science. This 1.3 GHz continuous-wave superconducting RF LINAC is designed to generate 4 GeV electron bunches up to one MHz, propelling the capabilities of XFEL sources. Achieving a significant milestone, the LCLS-II successfully reached its 2K operating temperature with the first electrons in October 2022, culminating in the generation of the first x-rays in September 2023. This paper offers an overview of the diverse array of DC magnet power supplies (PSs) employed in LCLS-II, which can be categorized into two sections: warm and superconducting. The warm section comprises of two crucial types of PSs-intermediate and trim. Notably, these PSs are subjected to tight stability requirements as low as 20 ppm. The warm section has close to 600 PSs. In the superconducting section, an extra level of complexity is added by including a quench protection circuit to protect the magnets in case of a sudden loss of superconductivity. PSs in this section also have a stability requirement of 0.02 %. The superconducting section has 105 PSs. This paper also discusses the system design and performance of these PSs.

This paper introduces the strategy for operating the cryogenic system of the ESS superconducting LINAC, emphasizing the integration of individual cryomodules and valve boxes within an unified system. The study focuses on the practical implementation of this strategy at Test Stand 2 (TS2) as a pilot project, validating the proposed control system in a real-world setting. The paper evaluates the primary goals which include performing functional tests, successful implementation, identifying control system shortcomings, and collecting valuable operator feedback for continuous improvement.

The end station for TLS beamline 07A (BL07A) primarily serves industrial applications, catering to various sample inspection requirements within the industry in multiple modes. In the past, manual switching of modes for different experiments was time-consuming, requiring the re-installment a lot components and taking long time to readjust the gas proportion for different samples. To enhance the efficiency of the BL07A experimental station, an automated design has been implemented, utilizing multiple self-made motorized platforms. Another crucial aspect for improved experimental efficiency is the use of a combination of vacuum pumps, flow meters, and electromagnetic valves for gas replacement system, significantly reducing the time needed for this process. The automated system is currently operational, reducing the operation time for experimental equipment switching from several hours to two minutes. The execution time for the gas replacement process has also been drastically reduced from 100 minutes to 5 minutes.

The 100 mm periodic permanent magnet Wiggler (W100) was installed in the 31st straight section of the TPS storage ring in September 2020, during a prolonged shutdown of the TPS. It provides photon energy ranging from 5 to 50 keV for user experimental applications. The mechanical structure of this facility involves assembling and connecting it to the upper and lower magnetic arrays, each approximately 500 mm in length. Precise control of the gap between the magnetic arrays and accurate adjustments are required. This report primarily describes the assembly process of various components of W100 and the inspection items along with the results.

The contemporary advancement in particle accelerator technology necessitates precise control over beam manipulation for various experimental and industrial applications. One pivotal aspect of this control resides in the generation and modulation of high-voltage pulse to manipulate the trajectory and behavior of particle beams within the accelerator systems. This extensive study delves into the design, development, and characterization of an adjustable-length pulse generator specifically tailored for a beam extraction kicker system, which is employed to navigate the beam out of the photon storage ring. The primary aim of this research is to engineer a versatile and reliable pulse-length modulation mechanism for a high-voltage pulse generation, which is capable of producing adjustable pulses with ultra-fine precision to meet the demanding requirements of beam manipulation within the accelerator setup. The system's design encompasses a meticulous integration of electronic components, waveform shaping modules, and control mechanisms to achieve the desired output.

Brookhaven National Laboratory has recently been selected as the site for the Electron-Ion Collider (EIC). The EIC will consist of two intersecting accelerators, one producing an intense beam of electrons, the other a high-energy beam of protons or heavier atomic nuclei, which are steered into head-on collisions. One of the sections of the EIC beamline will require a hadron injection fast kicker system. RadiaBeam is developing GaN-based pulser with ±50 kV voltage amplitude, <4 ns rise and fall times, 40 ns pulse width. In this paper, we discuss the development progress.

The 3D design of large accelerators like the Large Hadron Collider (LHC) requires coordination among equipment, services, and infrastructures. As numerous systems are designed, procured, and installed, 3D integration studies are important steps at any stages of a project, starting from the conceptual phase with space reservations, envelopes and interfaces, followed by the technical design phase managing the detailed and simplified 3D models, and finishing by the installation phase with follow-up of discrepancies. While the first phases serve to validate the accelerator configuration and design, the installation phase is followed by a reverse engineering process to verify the ‘as-built’ configuration, representing the final actual setup of the accelerator. At CERN, the 3D integration office for the accelerators assumes responsibility for collecting, aggregating, centralizing, and checking the 3D models provided by CERN design offices such as equipment owners, electrical, civil engineering, metallic structure, transport, handling, cooling, and ventilation services. This office manages 3D space, avoiding mechanical interferences before and during the installation phase. This paper describes the CAD, PDM and PLM methodologies used for 3D integration of the accelerators at CERN, highlighting their critical aspects and specificities.

The goal of the High Luminosity-Large Hadron Collider (HL-LHC) Inner Triplet (IT) String test, is to validate the assembly and connection procedures and tools required for its construction, to assess the collective behavior of the superconducting magnet chain in conditions as close as possible to those of their operation in the HL-LHC and to provide a training opportunity for the equipment teams for their work in the LHC tunnel. The IT String includes the systems required for operation at nominal conditions, such as the cryogenics, powering and quench protection systems. This contribution describes the individual system and short circuit tests performed at the IT String as part of the hardware commissioning and preparation for the full exploitation of the facility. After describing the IT String infrastructure, the individual system tests performed on the cryogenic and the associated vacuum systems are detailed. Moreover, the individual system and short circuit tests executed on the warm powering systems part of the magnet circuit including power converters, energy extraction systems and the DC connections are described. The powering interlock controller used for the global interlocking of the magnet circuits is also validated during this phase. The tests described involve the same steps as those planned for the LHC collider. Therefore, they validate the systems to be installed and ensure the time-efficient execution of activities for the HL-LHC project.

It has been seven years since the Taiwan Photon Source (TPS) started to serve users in 2016. Sixteen beamlines had been installed in the first and second phases of TPS beamline project. The third phase project was also launched in 2021. Considering the experimental hall is more compact and power saving issue, our research aimed to analyze a modified air conditioning system with better cooling efficiency through Computational Fluid Dynamic (CFD) simulation. One twelfth of the TPS experimental hall and two beamlines are modeled.

Tokyo Denkai has been producing niobium for superconducting cavities since 1985. We have also produced niobium for L-band cavities since the beginning of their development, and have a large number of production records. In particular, more than 20,000 pieces have been delivered to TESLA based on the XFEL-007 specifications for the European XFEL, LCLS-II, LCLS-II HE, and SHINE projects. In this report, we present a statistical evaluation of measured data on the actual mechanical properties of niobium sheets in a mass production of niobium sheets based on nearly identical specifications. Specifically, histograms of hardness, RRR, and tensile testing (rolling and transverse direction) of niobium sheets were drawn to evaluate the data variability. The data for all items were normally distributed, indicating that quality was controlled. In addition, the relationship between rolling direction and all tensile test items (yield stress, maximum stress, and elongation) were examined. Positive correlations were observed for yield stress and maximum stress. I report on the quality data and statistical results of the same product over a period of more than 10 years.