Detecting anomalous SRF cavity behavior with unsupervised learning

We present an unsupervised learning framework for detecting anomalous superconducting radio-frequency (SRF) cavity behavior at the Continuous Electron Beam Accelerator Facility (CEBAF), emphasizing its initial performance and effectiveness. Key to the system’s success was the development of data acq...

Full description

Saved in:
Bibliographic Details
Main Authors: Hal Ferguson, Jiang Li, Adam Carpenter, Chris Tennant, Dillon Thomas, Dennis Turner
Format: Article
Language:English
Published: American Physical Society 2025-03-01
Series:Physical Review Accelerators and Beams
Online Access:http://doi.org/10.1103/PhysRevAccelBeams.28.034602
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We present an unsupervised learning framework for detecting anomalous superconducting radio-frequency (SRF) cavity behavior at the Continuous Electron Beam Accelerator Facility (CEBAF), emphasizing its initial performance and effectiveness. Key to the system’s success was the development of data acquisition systems (DAQs) that capture fast-sampled, information-rich signals, essential for detecting transient effects. The approach involves creating daily cavity-specific models using principal component analysis to handle variations in rf signal behavior and mitigate performance degradation from data drift. This unsupervised method eliminates the need for expensive labeling by continuously updating models with recent data. Deployed and operational for 3 months before a scheduled shutdown, the system successfully identified several issues with DAQ signals, confirming its effectiveness. Despite access to only a fraction of CEBAF’s SRF cavity signals, the framework efficiently detected several instances requiring intervention, demonstrating a significant improvement over traditional, labor-intensive methods of manual plot inspection.
ISSN:2469-9888