IPAC2021 - List of Authors (Pernkopf, F.)

Paper	Title	Page
MOPAB344	Machine Learning Models for Breakdown Prediction in RF Cavities for Accelerators	1068
	C. Obermair, A. Apollonio, T. Cartier-Michaud, N. Catalán Lasheras, L. Felsberger, W.L. Millar, W. Wuensch CERN, Geneva, Switzerland C. Obermair, F. Pernkopf TUG, Graz, Austria
	Radio Frequency (RF) breakdowns are one of the most prevalent limits in RF cavities for particle accelerators. During a breakdown, field enhancement associated with small deformations on the cavity surface results in electrical arcs. Such arcs degrade a passing beam and if they occur frequently, they can cause irreparable damage to the RF cavity surface. In this paper, we propose a machine learning approach to predict the occurrence of breakdowns in CERN’s Compact LInear Collider (CLIC) accelerating structures. We discuss state-of-the-art algorithms for data exploration with unsupervised machine learning, breakdown prediction with supervised machine learning, and result validation with Explainable-Artificial Intelligence (Explainable AI). By interpreting the model parameters of various approaches, we go further in addressing opportunities to elucidate the physics of a breakdown and improve accelerator reliability and operation.
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2021-MOPAB344
About •	paper received ※ 20 May 2021 paper accepted ※ 16 July 2021 issue date ※ 11 August 2021
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

MOPAB345	Machine Learning with a Hybrid Model for Monitoring of the Protection Systems of the LHC	1072
	C. Obermair, A. Apollonio, Z. Charifoulline, M. Maciejewski, A.P. Verweij CERN, Geneva, Switzerland C. Obermair, F. Pernkopf TUG, Graz, Austria
	The LHC is the world’s largest particle accelerator and uses a complex set of sophisticated and highly reliable machine protection systems to ensure a safe operation with high availability for particle physics production. The data gathered during several years of successful operation allow the use of data-driven methods to assist experts in finding anomalies in the behavior of those protection systems. In this paper, we derive a model that can extend the existing signal monitoring applications for the LHC protection systems with machine learning. Our hybrid model combines an existing threshold-based system with a SVM by using signals, manually validated by experts. Even with a limited amount of data, the SVM learns to integrate the expert knowledge and contributes to a better classification of safety-critical signals. Using this approach, we analyze historical signals of quench heaters, which are an important part of the quench protection system for superconducting magnets. Particularly, it is possible to incorporate expert decisions into the classification process and to improve the failure detection rate of the existing quench heater discharge analysis tool.
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2021-MOPAB345
About •	paper received ※ 20 May 2021 paper accepted ※ 19 July 2021 issue date ※ 01 September 2021
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

Paper

Title

Page

Machine Learning Models for Breakdown Prediction in RF Cavities for Accelerators

1068

C. Obermair, A. Apollonio, T. Cartier-Michaud, N. Catalán Lasheras, L. Felsberger, W.L. Millar, W. Wuensch
CERN, Geneva, Switzerland
C. Obermair, F. Pernkopf
TUG, Graz, Austria

Radio Frequency (RF) breakdowns are one of the most prevalent limits in RF cavities for particle accelerators. During a breakdown, field enhancement associated with small deformations on the cavity surface results in electrical arcs. Such arcs degrade a passing beam and if they occur frequently, they can cause irreparable damage to the RF cavity surface. In this paper, we propose a machine learning approach to predict the occurrence of breakdowns in CERN’s Compact LInear Collider (CLIC) accelerating structures. We discuss state-of-the-art algorithms for data exploration with unsupervised machine learning, breakdown prediction with supervised machine learning, and result validation with Explainable-Artificial Intelligence (Explainable AI). By interpreting the model parameters of various approaches, we go further in addressing opportunities to elucidate the physics of a breakdown and improve accelerator reliability and operation.

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2021-MOPAB344

About •

paper received ※ 20 May 2021 paper accepted ※ 16 July 2021 issue date ※ 11 August 2021

Export •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

MOPAB345

Machine Learning with a Hybrid Model for Monitoring of the Protection Systems of the LHC

1072

C. Obermair, A. Apollonio, Z. Charifoulline, M. Maciejewski, A.P. Verweij
CERN, Geneva, Switzerland
C. Obermair, F. Pernkopf
TUG, Graz, Austria

The LHC is the world’s largest particle accelerator and uses a complex set of sophisticated and highly reliable machine protection systems to ensure a safe operation with high availability for particle physics production. The data gathered during several years of successful operation allow the use of data-driven methods to assist experts in finding anomalies in the behavior of those protection systems. In this paper, we derive a model that can extend the existing signal monitoring applications for the LHC protection systems with machine learning. Our hybrid model combines an existing threshold-based system with a SVM by using signals, manually validated by experts. Even with a limited amount of data, the SVM learns to integrate the expert knowledge and contributes to a better classification of safety-critical signals. Using this approach, we analyze historical signals of quench heaters, which are an important part of the quench protection system for superconducting magnets. Particularly, it is possible to incorporate expert decisions into the classification process and to improve the failure detection rate of the existing quench heater discharge analysis tool.

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2021-MOPAB345

About •

paper received ※ 20 May 2021 paper accepted ※ 19 July 2021 issue date ※ 01 September 2021

Export •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)