ReliableBeam optimization beam feedbackBeam feedback is essential for operating an accelerator with stable charged particleParticle beams. Furthermore, one must determine the setpoints of these charged particle beam feedbackBeam feedback loops for optimal parameters of the machine's final products (e.g., free electron laser (FEL) photon beam). This leads to the needs for the beam optimizationBeam optimization processes discussed in this chapter. More generally, optimizing the operation of an accelerator also includes setting up the accelerator subsystemsAccelerator subsystem for maximum performance like availability, stability, power efficiency, robustnessRobustness, and so on. In this chapter, we will introduce several widely used optimization algorithms focusing on online beam optimizationsBeam optimization. The test results of these algorithms for optimizing the bunch2 parameters of SwissFELSwissFEL will be demonstrated.

The ITER Real-Time Framework (RTF) is a software suite designed and developed to facilitate the implemen-tation of various real-time applications for ITER plant systems. The main driver to implement RTF was the Plasma Control System (PCS). RTF was designed as a base and development environment for PCS. However, due to its universal architecture, it can also be applied in other systems requiring real-time control or data processing. This paper presents the demonstration system to evaluate RTF in Thomson Scattering (TS) diagnostics. The system was developed to validate the integration of data acquisition hardware with RTF and check it in real-time data processing applications. The presented system covers the whole path of data acquisition, processing and archiving required for a typical Instrumentation and Control (I&C) system for ITER diagnostics. The main components of the RTF-based application developed for the system are device support for the pulsed giga-sample analog-to-digital converter, functional blocks implementing algorithms for analysing pulses from polychromator and calculating plasma electron temperature as well as interfaces for archiving raw data and publishing the measurement results. The overriding goal of the presented work was a detailed functional and performance evaluation of the implemented RTF-based system. The system was tested in laboratory using simulated data and in tokamak conditions at Korea Superconducting Tokamak Advanced Research (KSTAR). During the test campaign at KSTAR, the developed RTF-based application was integrated into the Thomson Scattering system and tested during regular operation of the tokamak with signals from a real polychromator. The results of measurements and performance evaluation are presented and discussed in the paper.

ParticleBeam feedback accelerators must provide stable beams for achieving their physics goals. Stabilizing the beam against external disturbances is a critical task of the beam control systemBeam control system. Typically, we introduce feedback controlFeedback control for mitigating the effects of slow disturbances like temperature or humidity changes and power supply driftsDrift. In this chapter, we present an overview of the beam feedbackBeam feedback control in accelerators, introduce a generic controller structure, and discuss several methods for feedback controller design. We highlight the control of multiple-input multiple-output static systems and introduce several controller design methods based on singular value decompositionSingular value decomposition, least-square with regularizationRegularization, and robust controlRobust control.

Machine learningBeam control is a class of data-driven methodologies to identify systemSystem models, make predictions, or determine control actionsAction. Machine learningMachine learning methods are attractive since less domain knowledge is required when applying them to accelerator design and operation. This chapter presents an overview of the machine learning algorithms for accelerator beam controlsBeam control. After a brief introduction, we discuss building surrogate modelsSurrogate model for accelerator subsystemsAccelerator subsystem and beam responses based on the input–outputOutput data. The neural networkNeural network and Gaussian process regression modelsGaussian process regression model are emphasized. These surrogate models are beneficial for adapting beam feedbackBeam feedback for different operating pointsOperating point, implementing feedforward controlFeedforward control, and accelerating beam optimizationBeam optimization processes. The concepts and algorithms of reinforcement learningReinforcement learning are then introduced and used to solve linear quadratic GaussianLinear quadratic Gaussian problems. In the end, we also summarize the machine learning applications in particleParticle accelerators beyond beam controlsBeam control.

This book systematically discusses the algorithms and principles for achieving stable and optimal beam parameters in particle accelerators

The subsystems of a particle accelerator must be turned on in a proper sequence to perform successful beam acceleration. Furthermore, the charged particles must interact with the RF fields (or the laser fields in case of laser-plasma acceleration is used) at the correct time for the desired accelerating phase. The timing and synchronization systems guarantee the required timing relations between the RF fields and the beam. The timing system defines the timing events and produces triggers if required for different subsystems to start or stop their operation. The synchronization system provides a common frequency and phase reference to all subsystems. We will introduce the basic concepts and architecture of the timing and synchronization systems in this chapter.

As more accelerators adopt superconducting cavities or require high RF field stability, the LLRF system becomes critical for meeting the accelerator performance objectives. Different control strategies have been developed to solve the RF control problems raised by different machines. In this chapter, we give an overview of the RF control strategies widely used in LLRF systems. We will briefly compare the feedback and feedforward control, the amplitude/phase and in-phase/quadrature (I/Q) control, the generator-driven resonator (GDR), self-excited loop (SEL) and phase-locked loop (PLL) control, the analog and digital control, and the single-cavity and vector-sum control. The pros and cons of the control strategies will also be discussed.

The linearity of the high-power RF (HPRF) and low-level RF (LLRF) components is critical for the RF control performance. A nonlinear RF detector causes errors in the measurements of the amplitude and phase. Moreover, the RF driving chain nonlinearity produces higher-order harmonics in the output and results in gain and phase shift varying with the input power. In this chapter, the nonlinear effects in an accelerator RF system will be discussed. We also introduce several widely used approaches to linearize the RF amplifiers and to deal with the nonlinearity in feedback controllers.

A (mathematical) model of the RF system is essential for the design and analysis of LLRF systems. The model here describes the (dynamical) relationship between the outputs and inputs of the RF system in the form of transfer functions. The transfer functions provide insights into the behavior and limitations of the RF system. For example, we may predict the RF system outputs in the presence of RF drive, beam loading and external distrubances (e.g., electromagnetic noise, thermal drifts or mechanical vibrations) using the transfer functions. Another useful information that the model can provide is the stability limits for the loop phase, loop gain and loop delay, which helps to determine these operational parameters of the RF controller. We will derive the RF system model through its physical principles, which can, therefore, also be used to optimize the operational parameters and to identify the characteristics of the RF system.

Providing a bridge between RF control engineering and beam physics, this book systematically discusses the core algorithms and technologies for ...

RF detectors transform the RF frequency signals to baseband signals in the format of complex envelopes (i.e., phasors) or amplitude and phase. RF actuators perform the inverse transformations. This chapter starts with an introduction to several widely used analog or digital RF detection schemes. Then we focus on the digital solutions and provide a few popular RF detection algorithms. The non-I/Q demodulation algorithm will be discussed in detail. Furthermore, some advanced topics of RF detection involving reference tracking will be discussed. Finally, we present some basic RF actuation schemes, such as the direct up-conversion, the single sideband (SSB) up-conversion, and the intermediate frequency (IF) up-conversion. These up-conversion schemes are also frequently found in commercial RF transmitters.

One of the LLRF system's core functions is controlling the RF field for particle beam acceleration. An RF controller implements feedback or feedforward algorithms, regulating the RF field to follow set points or stabilizing it against disturbances. This chapter will discuss the principles of GDR, SEL, and PLL control strategies. Several widely used feedback and feedforward algorithms will be introduced. These algorithms can be applied to practical LLRF systems directly. We will also discuss the cavity resonance control, which keeps the cavity on-resonance and is essential to enhance the performance of RF field controls.

Several design variants of chopper-based digital signal integrators have been tested to evaluate the optimal solution to achieve the ITER magnetics diagnostic requirements. A maximum flux-equivalent drift of 500 mu V.s/hour is one of the key ITER magnetics diagnostic constraints for the integrators. The flux drift must be below the specified limit whilst the device satisfies other stringent specifications such as, 500 V galvanic isolation, 14-bit ENOB and environment magnetic field tolerance up to 10 mT. This paper presents the results of some of the tests performed on the integrator prototypes developed. These include tests to verify the integrator drift during long experiments when subjected to different conditions, e.g., imposition of a common mode voltage and input signals with a frequency spectrum that challenges the design limits.

The local radiation shielding design for the detector of ITER VUV edge imaging spectrometer is evaluated based on the MCNP calculation using a local port cell model of ITER upper port #18. A back-illuminated CCD (Charge-Coupled Device), the envisaged VUV (Vacuum Ultraviolet) detector for ITER VUV edge imaging spectrometer will be installed at ITER upper port #18 port cell region, in which a harsh radiation environment is expected with neutron flux higher than 106 neutrons cm(-2) s(-1) mainly thermalized from d-t neutrons in plasma as well as high gamma ray dose of several tens kGy mainly from 16N isotopes in water coolant. For the evaluation of the radiation exposure to the detector, the local port cell model is developed to reduce both the calculation time and statistical error. The boundary neutron source based on MCNP result using C-lite model as well as gamma source based on ITER radiation map has been utilized for the analysis of local port cell model. Since the radiation exposure to the back-illuminated CCD should be mitigated as much as possible to minimize the radiation damage to the detector as well as single event upset, local shielding design options for the VUV detector with various shapes, thicknesses, and material compositions are evaluated. The result shows that the neutron flux and gamma dose at the location of VUV detector can be mitigated below 100 n cm(-2) s(-1) and 10 Gy, respectively, which are the alert thresholds for non-critical electronics in ITER. (C) 2017 Elsevier B.V. All rights reserved.

The main goal of this work is to demonstrate that a digital integrator based on the signal chopping concept is capable of attaining the ITER requirements. In particular, the ITER magnetics diagnostic requires a maximum flux drift of 500μVs/hour, among other specifications, for the signal integrators. As of today, known commercial integration modules do not fully comply simultaneously with all ITER magnetics requirements. A first phase of prototyping, presented in this work, comprises the development and testing of four design variants. Combinations of a SAR ADC (AD7960) and a Delta-Sigma ADC (ADS1675) with different analog front ends were used for the corresponding integrator prototypes. The designs have a common interface to an FPGA based system that receives the data acquired during the tests and streams it through a GbE link to a PC, where real-time digital integration of the signals is performed using the MARTe control framework. The GbE network also acts as the interfacing medium for the data archiving, through the connection of the integrator prototypes under test to an MDSplus based environment. This paper presents the integrator prototype designs developed and tests done so far.

The preliminary design of XRCS Survey spectrometer for ITER has been developed addressing many challenges such as designing a 8.0 m long, vacuum extending sight-tube that interfaces mystal spectrometer. placed in the port-cell, with equatorial portsplug (EPP-11) while allowing 50 nun machine movements, and opthnizing neutron shield design so that systems can fit into the available space arid still the shutdown dose rates (SDDR) remains within the safe limits. The design detailing has been done for the sightstube and its components addressing the ITER specific requirements. Engineering and nentronic analysis are perfonfied for estimating the thermal displaceinent, stresses in the fron Eend components, neutron flux on the sight-tube components, SDDRs in the interspace region etc.

To aid in assessing the functional performance of ITER, Fission Chambers (FC) based on the neutron diagnostic use case deliver timestamped measurements of neutron source strength and fusion power. To demonstrate the Plant System Instrumentation & Control (I&C) required for such a system, ITER Organization (IO) has developed a neutron diagnostics use case that fully complies with guidelines presented in the Plant Control Design Handbook (PCDH). The implementation presented in this paper has been developed on the PXI Express (PXIe) platform using products from the ITER catalog of standard I&C hardware for fast controllers. Using FlexRIO technology, detector signals are acquired at 125 MS/s, while filtering, decimation, and three methods of neutron counting are performed in real-time via the onboard Field Programmable Gate Array (FPGA). Measurement results are reported every 1 ms through Experimental Physics and Industrial Control System (EPICS) Channel Access (CA), with real-time timestamps derived from the ITER Timing Communication Network (TCN) based on IEEE 1588-2008. Furthermore, in accordance with ITER specifications for CODAC Core System (CCS) application development, the software responsible for the management, configuration, and monitoring of system devices has been developed in compliance with a new EPICS module called Nominal Device Support (NDS) and RIO/FlexRIO design methodology.