Machine learning (ML) has the potential for significant impact on the modeling, operation, and control of particle accelerators due to its ability to model nonlinear behavior, interpolate on complicated surfaces, and adapt to system changes over time. Anomaly detection in particular has been highlighted as an area where ML can significantly impact the operation of accelerators. These algorithms work by identifying subtle behaviors of key variables prior to negative events. Efforts to apply ML to anomaly detection have largely focused on subsystems such as RF cavities, superconducting magnets, and losses in rings. However, dedicated efforts to understand how to apply ML for anomaly detection in linear accelerators have been limited. In this paper the use of autoencoders is explored to identify anomalous behavior in measured data from the Fermilab low-energy linear accelerator.

https://www.mdpi.com/2078-2489/12/6/238/pdf

Autoencoder Based Analysis of RF Parameters in the Fermilab Low Energy Linac

The Fermilab Proton Plan, tasked to increase the intensity and reliability of the Proton Source, has identified the Low Level Radio Frequency (LLRF) system as the critical component to be upgraded in the Linac. The current 201.25 MHz Drift Tube Linac LLRF system was designed and built over 35 years ago and does not meet the higher beam quality requirements under the new Proton Plan. A new VXI based LLRF system has been designed to improve cavity vector regulation and reduce beam losses. The upgrade includes an adaptive feedforward system for beam loading compensation, a new phase feedback system, and a digital phase comparator for cavity tuning. The new LLRF system is phase locked to the 805 MHz reference line, currently used as frequency standard in the higher energy accelerating section of the Linac. This paper will address the current status of the project, present the advancements in both amplitude and phase stability over the old LLRF system, and discuss commissioning plans.

L.J.Allen

Papers

New LLRF System for Fermilab 201.25 MHz Linac