Precise calibration of the cavity phase signals is necessary for the operation of any particle accelerator. For many systems this requires human in the loop adjustments based on measurements of the beam parameters downstream. Some recent work has developed a scheme for the calibration of the cavity phase using beam measurements and beam-loading however this scheme is still a multi-step process that requires heavy automation or human in the loop. In this paper we analyze a new scheme that uses only RF signals reacting to beam-loading to calculate the phase of the beam relative to the cavity. This technique could be used in slow control loops to provide real-time adjustment of the cavity phase calibration without human intervention thereby increasing the stability and reliability of the accelerator.

In this paper we present the design, simulation, and proof of principle results of an optimization based adaptive feed-forward algorithm for beam-loading compensation in a high impedance room temperature cavity. We begin with an overview of prior developments in beam loading compensation. Then we discuss different techniques for adaptive beam loading compensation and why the use of Newton's Method is of interest for this application. This is followed by simulation and initial experimental results of this method.

This paper describes a recent effort to develop and benchmark a simulation tool for the analysis of radio frequency (RF) transients and their compensation in an H-linear accelerator. Existing tools in this area either focus on electron linear accelerators (LINACs) or lack fundamental details about the low level radio frequency system that are necessary to provide realistic performance estimates. In this paper, we begin with a discussion of our computational models followed by benchmarking with existing beam-dynamics codes and measured data. We then analyze the effect of RF transients and their compensation in the Proton Improvement Plan-II LINAC, followed by an analysis of calibration errors and how Newton's method-based feedback scheme can be used to regulate the beam energy to within the specified limits.

The PIP-II project is a program to upgrade the Fermilab accelerator complex. The PIP-II linac includes a 2.1 MeV Medium Energy Beam Transport (MEBT) section that incorporates a unique chopping system to perform arbitrary, bunch-by-bunch removal of 162.5 MHz structured beam. The MEBT chopping system will consist of two identical kickers working together and a beam absorber. One design of two having been proposed has been a 200 Ohm characteristic impedance traveling wave dual-helix kicker driven with custom designed high-speed switches. This paper reports on the first performance results of one prototype kicker built, installed and tested with beam at the PIP-II Injector Test (PIP2IT) facility. The helix deflector design details are discussed. The electrical performance of the high-speed switch driver operating at 500 V bias is presented. Tests performed were chopping beam at 81.25 MHz for microseconds as well as with a truly arbitrary pattern for 550 μs bursts having a 45 MHz average switching rate and repeating at 20 Hz.

The PIP-II accelerator is a proposed upgrade to the Fermilab accelerator complex that will replace the existing, 400 MeV room temperature LINAC with an 800 MeV superconducting LINAC. Part of this upgrade includes a new injection scheme into the booster that levies tight requirements on the LLRF control system for the cavities. In this paper we discuss the challenges of the PIP-II accelerator and the present status of the LLRF system for this project.

In this paper we develop an a priori method for simulating dynamic resonant frequency and temperature responses in a radio frequency quadrupole (RFQ) and its associated water-based cooling system respectively. Our model provides a computationally efficient means to evaluate the transient response of the RFQ over a large range of system parameters. The model was constructed prior to the delivery of the PIP-II Injector Test RFQ and was used to aid in the design of the water-based cooling system, data acquisition system, and resonance control system. Now that the model has been validated with experimental data, it can confidently be used to aid in the design of future RFQ resonance controllers and their associated water-based cooling systems. Without any empirical fitting, it has demonstrated the ability to predict absolute temperature and frequency changes to 11% accuracy on average, and relative changes to 7% accuracy.

The PIP-II injector test radio frequency quadrupole (RFQ) arrived at Fermilab in the fall of 2015. The RFQ is a 162.5MHz Haccelerator with a nominal drive power of 100kW, which produces a bunched Hbeam at 2.1MeV. In this paper we discuss commissioning, operational performance, and improvements to the low level RF (LLRF) control system for the RFQ. We begin by describing the general system configuration and initial simulation results. We will then highlight temperature related issues in the high power RF system, which necessitate active control over the phase balance of the two amplifiers. Finally we demonstrate performance of the RF feedback and feed-forward compensation needed to meet specification during a 20-microsecond beam pulse.

Particle accelerators are host to myriad nonlinear and complex physical phenomena. They often involve a multitude of interacting systems, are subject to tight performance demands, and should be able to run for extended periods of time with minimal interruptions. Often times, traditional control techniques cannot fully meet these requirements. One promising avenue is to introduce machine learning and sophisticated control techniques inspired by artificial intelligence, particularly in light of recent theoretical and practical advances in these fields. Within machine learning and artificial intelligence, neural networks are particularly well-suited to modeling, control, and diagnostic analysis of complex, nonlinear, and time-varying systems, as well as systems with large parameter spaces. Consequently, the use of neural network-based modeling and control techniques could be of significant benefit to particle accelerators. For the same reasons, particle accelerators are also ideal test-beds for these techniques. Many early attempts to apply neural networks to particle accelerators yielded mixed results due to the relative immaturity of the technology for such tasks. The purpose of this paper is to re-introduce neural networks to the particle accelerator community and report on some work in neural network control that is being conducted as part of a dedicated collaboration between Fermilab and Colorado State University (CSU). We describe some of the challenges of particle accelerator control, highlight recent advances in neural network techniques, discuss some promising avenues for incorporating neural networks into particle accelerator control systems, and describe a neural network-based control system that is being developed for resonance control of an RF electron gun at the Fermilab Accelerator Science and Technology (FAST) facility, including initial experimental results from a benchmark controller.

As part of the PIP-II Injector Experiment (PXIE) accelerator, a four-vane radio frequency quadrupole (RFQ) accelerates a 30-keV, 1-mA to 10-mA H- ion beam to 2.1 MeV. It is designed to operate at a frequency of 162.5 MHz with arbitrary duty factor, including continuous wave (CW) mode. The resonant frequency is controlled solely by a water-cooling system. We present an initial neural network model of the RFQ frequency response to changes in the cooling system and RF power conditions during pulsed operation. A neural network model will be used in a model predictive control scheme to regulate the resonant frequency of the RFQ.

For the PIP-II Injector Test (PI-Test) at Fermilab, a four-vane radio frequency quadrupole (RFQ) is designed to accelerate a 30-keV, 1-mA to 10-mA, H- beam to 2.1 MeV under both pulsed and continuous wave (CW) RF operation. The available headroom of the RF amplifiers limits the maximum allowable detuning to 3 kHz, and the detuning is controlled entirely via thermal regulation. Fine control over the detuning, minimal manual intervention, and fast trip recovery is desired. In addition, having active control over both the walls and vanes provides a wider tuning range. For this, we intend to use model predictive control (MPC). To facilitate these objectives, we developed a dedicated control framework that handles higher-level system decisions as well as executes control calculations. It is written in Python in a modular fashion for easy adjustments, readability, and portability. Here we describe the framework and present the first control results for the PI-Test RFQ under pulsed and CW operation.

Colorado State University (CSU) and Fermi National Accelerator Laboratory (Fermilab) have been developing a control system to regulate the resonant frequency of an RF electron gun. As part of this effort, we present initial test results for a benchmark temperature controller that combines a machine learning-based model and a predictive control algorithm. This is part of an on-going effort to develop adaptive, machine learning-based tools specifically to address control challenges found in particle accelerator systems.