Coincident learning for beam-based rf station fault identification using phase information at the SLAC linac coherent light source

Liang, Jia; Colocho, William; Decker, Franz-Josef; Humble, Ryan; Morris, Ben; O’Shea, Finn H.; Steele, David A.; Zhang, Zhe; Darve, Eric; Ratner, Daniel

doi:10.1103/zmmr-ry9h

Coincident learning for beam-based rf station fault identification using phase information at the SLAC linac coherent light source

Journal Article · Wed Dec 03 00:00:00 EST 2025 · Physical Review Accelerators and Beams

DOI:https://doi.org/10.1103/zmmr-ry9h· OSTI ID:3005876

^[1]; ^[2]; ^[2]; ^[3]; ^[2]; ^[2]; ^[2]; ^[2]; ^[1]; ^[2]

Stanford Univ., CA (United States)
SLAC National Accelerator Laboratory (SLAC), Menlo Park, CA (United States)
Cerebras Systems, Sunnyvale, CA (United States)

Anomalies in radio-frequency (rf) stations can result in unplanned downtime and performance degradation in linear accelerators such as SLAC’s Linac Coherent Light Source (LCLS). Detecting these anomalies is challenging due to the complexity of accelerator systems, high data volume, and scarcity of labeled fault data. Prior work identified faults using beam-based detection, combining rf amplitude and beam position monitor data. Due to the simplicity of the rf amplitude data, classical methods are sufficient to identify faults, but the recall is constrained by the low-frequency and asynchronous characteristics of the data. In this work, we leverage high-frequency, time-synchronous rf phase data to enhance anomaly detection in the LCLS accelerator. Due to the complexity of phase data, classical methods fail, and we instead train deep neural networks within the Coincident Anomaly Detection (CoAD) framework. We find that applying CoAD to phase data detects nearly 3 times as many anomalies as when applied to amplitude data, while achieving broader coverage across rf stations. Furthermore, the rich structure of phase data enables us to cluster anomalies into distinct physical categories. Through the integration of auxiliary system status bits, we link clusters to specific fault signatures, providing additional granularity for uncovering the root cause of faults. We also investigate interpretability via Shapley values, confirming that the learned models focus on the most informative regions of the data and providing insight for cases where the model makes mistakes. This work demonstrates that phase-based anomaly detection for rf stations improves both diagnostic coverage and root cause analysis in accelerator systems and that deep neural networks are essential for effective analysis.

View Accepted Manuscript (DOE)

Research Organization:: SLAC National Accelerator Laboratory (SLAC), Menlo Park, CA (United States)

Sponsoring Organization:: USDOE Office of Science (SC), Basic Energy Sciences (BES)

Grant/Contract Number:: AC02-76SF00515

OSTI ID:: 3005876

Journal Information:: Physical Review Accelerators and Beams, Journal Name: Physical Review Accelerators and Beams Journal Issue: 12 Vol. 28; ISSN 2469-9888

Publisher:: American Physical Society (APS)Copyright Statement

Country of Publication:: United States

Language:: English

References (16)

Tensor canonical correlation analysis Min, Eun Jeong; Chi, Eric C.; Zhou, Hua Stat, Vol. 8, Issue 1 https://doi.org/10.1002/sta4.253	journal	January 2019
Few-shot fault diagnosis of particle accelerator power system using a bidirectional discriminative prototype network Yang, Zhe; Ye, Rongbing; Jiang, Lingli Applied Soft Computing, Vol. 170 https://doi.org/10.1016/j.asoc.2025.112726	journal	February 2025
Time-Series Deep Learning Anomaly Detection for Particle Accelerators Marcato, Davide; Bortolato, Damiano; Martinelli, Valentina IFAC-PapersOnLine, Vol. 56, Issue 2 https://doi.org/10.1016/j.ifacol.2023.10.1855	journal	January 2023
A machine learning approach for particle accelerator errant beam prediction using spatial phase deviation Yucesan, Yigit A.; Blokland, Willem; Ramuhalli, Pradeep Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, Vol. 1063 https://doi.org/10.1016/j.nima.2024.169232	journal	June 2024
Particle accelerator power system early fault diagnosis based on deep learning and multi-sensor feature fusion Jiqing, Zhou; Deming, Li; Haijun, Su Engineering Research Express, Vol. 6, Issue 2 https://doi.org/10.1088/2631-8695/ad55a6	journal	June 2024
A smart alarm for particle accelerator beamline operations Tennant, Chris; Freeman, Brian; Kazimi, Reza Machine Learning: Science and Technology, Vol. 4, Issue 1 https://doi.org/10.1088/2632-2153/acb98d	journal	February 2023
Robust errant beam prognostics with conditional modeling for particle accelerators Rajput, Kishansingh; Schram, Malachi; Blokland, Willem Machine Learning: Science and Technology, Vol. 5, Issue 1 https://doi.org/10.1088/2632-2153/ad2e18 015044	journal	March 2024
Accelerating cavity fault prediction using deep learning at Jefferson Laboratory Rahman, Md M.; Carpenter, A.; Iftekharuddin, K. Machine Learning: Science and Technology, Vol. 5, Issue 3 https://doi.org/10.1088/2632-2153/ad7ad6	journal	September 2024
Outlook towards deployable continual learning for particle accelerators Rajput, Kishansingh; Lin, Sen; Edelen, Auralee Machine Learning: Science and Technology, Vol. 6, Issue 3 https://doi.org/10.1088/2632-2153/adeb45	journal	July 2025
Uncertainty aware anomaly detection to predict errant beam pulses in the Oak Ridge Spallation Neutron Source accelerator Blokland, Willem; Rajput, Kishansingh; Schram, Malachi Physical Review Accelerators and Beams, Vol. 25, Issue 12 https://doi.org/10.1103/PhysRevAccelBeams.25.122802	journal	December 2022
Beam-based rf station fault identification at the SLAC Linac Coherent Light Source Humble, Ryan; O’Shea, Finn H.; Colocho, William Physical Review Accelerators and Beams, Vol. 25, Issue 12 https://doi.org/10.1103/PhysRevAccelBeams.25.122804	journal	December 2022
Machine-learning-based pressure-anomaly detection system for SuperKEKB accelerator Suetsugu, Yusuke Physical Review Accelerators and Beams, Vol. 27, Issue 6 https://doi.org/10.1103/PhysRevAccelBeams.27.063201	journal	June 2024
hdbscan: Hierarchical density based clustering McInnes, Leland; Healy, John; Astels, Steve The Journal of Open Source Software, Vol. 2, Issue 11 https://doi.org/10.21105/joss.00205	journal	March 2017
Pyrcca: Regularized Kernel Canonical Correlation Analysis in Python and Its Applications to Neuroimaging Bilenko, Natalia Y.; Gallant, Jack L. Frontiers in Neuroinformatics, Vol. 10 https://doi.org/10.3389/fninf.2016.00049	journal	November 2016
Early Fault Detection in Particle Accelerator Power Electronics Using Ensemble Learning Radaideh, Majdi I.; Pappas, Chris; Wezensky, Mark International Journal of Prognostics and Health Management, Vol. 14, Issue 1 https://doi.org/10.36001/ijphm.2023.v14i1.3419	journal	June 2023
LCLS RF Station Phase Anomaly Candidate Dataset Liang, Jia; Colocho, William; Decker, Franz-Josef SLAC National Accelerator Laboratory (SLAC), Menlo Park, CA (United States) https://doi.org/10.71929/2584368	dataset	January 2025

Similar Records

LCLS RF Station Phase Anomaly Candidate Dataset

Dataset · Wed May 21 00:00:00 EDT 2025 · OSTI ID:2584368

Investigating Anomalies in Compute Clusters: An Unsupervised Learning Approach

Conference · Wed Nov 01 00:00:00 EDT 2023 · OSTI ID:2439937

Related Subjects

Accelerator Physics (physics.acc-ph)
Beam diagnostics
Beam loss
FOS: Physical sciences
Machine learning
Radio frequency techniques

Coincident learning for beam-based rf station fault identification using phase information at the SLAC linac coherent light source

Citation Formats

References (16)

Similar Records

Related Subjects