Home > Research > Publications & Outputs > Explainable machine learning for breakdown pred...

Research

Associated organisational unit

Links

Text available via DOI:

Keywords

View graph of relations

Explainable machine learning for breakdown prediction in high gradient rf cavities

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Published

Standard

Explainable machine learning for breakdown prediction in high gradient rf cavities. / Obermair, Christoph; Cartier-Michaud, Thomas; Apollonio, Andrea et al.
In: Physical Review Accelerators and Beams, Vol. 25, No. 10, 104601, 04.10.2022.

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Harvard

Obermair, C, Cartier-Michaud, T, Apollonio, A, Millar, W, Felsberger, L, Fischl, L, Bovbjerg, HS, Wollmann, D, Wuensch, W, Catalan-Lasheras, N, Boronat, M, Pernkopf, F & Burt, G 2022, 'Explainable machine learning for breakdown prediction in high gradient rf cavities', Physical Review Accelerators and Beams, vol. 25, no. 10, 104601. https://doi.org/10.1103/physrevaccelbeams.25.104601

APA

Obermair, C., Cartier-Michaud, T., Apollonio, A., Millar, W., Felsberger, L., Fischl, L., Bovbjerg, H. S., Wollmann, D., Wuensch, W., Catalan-Lasheras, N., Boronat, M., Pernkopf, F., & Burt, G. (2022). Explainable machine learning for breakdown prediction in high gradient rf cavities. Physical Review Accelerators and Beams, 25(10), Article 104601. https://doi.org/10.1103/physrevaccelbeams.25.104601

Vancouver

Obermair C, Cartier-Michaud T, Apollonio A, Millar W, Felsberger L, Fischl L et al. Explainable machine learning for breakdown prediction in high gradient rf cavities. Physical Review Accelerators and Beams. 2022 Oct 4;25(10):104601. doi: 10.1103/physrevaccelbeams.25.104601

Author

Obermair, Christoph ; Cartier-Michaud, Thomas ; Apollonio, Andrea et al. / Explainable machine learning for breakdown prediction in high gradient rf cavities. In: Physical Review Accelerators and Beams. 2022 ; Vol. 25, No. 10.

Bibtex

@article{c800a8207ef744959c179adc7c198c90,
title = "Explainable machine learning for breakdown prediction in high gradient rf cavities",
abstract = "The occurrence of vacuum arcs or radio frequency (rf) breakdowns is one of the most prevalent factors limiting the high-gradient performance of normal conducting rf cavities in particle accelerators. In this paper, we search for the existence of previously unrecognized features related to the incidence of rf breakdowns by applying a machine learning strategy to high-gradient cavity data from CERN{\textquoteright}s test stand for the Compact Linear Collider (CLIC). By interpreting the parameters of the learned models with explainable artificial intelligence (AI), we reverse-engineer physical properties for deriving fast, reliable, and simple rule–based models. Based on 6 months of historical data and dedicated experiments, our models show fractions of data with a high influence on the occurrence of breakdowns. Specifically, it is shown that the field emitted current following an initial breakdown is closely related to the probability of another breakdown occurring shortly thereafter. Results also indicate that the cavity pressure should be monitored with increased temporal resolution in future experiments, to further explore the vacuum activity associated with breakdowns.",
keywords = "Surfaces and Interfaces, Physics and Astronomy (miscellaneous), Nuclear and High Energy Physics",
author = "Christoph Obermair and Thomas Cartier-Michaud and Andrea Apollonio and William Millar and Lukas Felsberger and Lorenz Fischl and Bovbjerg, {Holger Severin} and Daniel Wollmann and Walter Wuensch and Nuria Catalan-Lasheras and Mar{\c c}{\`a} Boronat and Franz Pernkopf and Graeme Burt",
year = "2022",
month = oct,
day = "4",
doi = "10.1103/physrevaccelbeams.25.104601",
language = "English",
volume = "25",
journal = "Physical Review Accelerators and Beams",
issn = "2469-9888",
publisher = "American Physical Society",
number = "10",

}

RIS

TY - JOUR

T1 - Explainable machine learning for breakdown prediction in high gradient rf cavities

AU - Obermair, Christoph

AU - Cartier-Michaud, Thomas

AU - Apollonio, Andrea

AU - Millar, William

AU - Felsberger, Lukas

AU - Fischl, Lorenz

AU - Bovbjerg, Holger Severin

AU - Wollmann, Daniel

AU - Wuensch, Walter

AU - Catalan-Lasheras, Nuria

AU - Boronat, Marçà

AU - Pernkopf, Franz

AU - Burt, Graeme

PY - 2022/10/4

Y1 - 2022/10/4

N2 - The occurrence of vacuum arcs or radio frequency (rf) breakdowns is one of the most prevalent factors limiting the high-gradient performance of normal conducting rf cavities in particle accelerators. In this paper, we search for the existence of previously unrecognized features related to the incidence of rf breakdowns by applying a machine learning strategy to high-gradient cavity data from CERN’s test stand for the Compact Linear Collider (CLIC). By interpreting the parameters of the learned models with explainable artificial intelligence (AI), we reverse-engineer physical properties for deriving fast, reliable, and simple rule–based models. Based on 6 months of historical data and dedicated experiments, our models show fractions of data with a high influence on the occurrence of breakdowns. Specifically, it is shown that the field emitted current following an initial breakdown is closely related to the probability of another breakdown occurring shortly thereafter. Results also indicate that the cavity pressure should be monitored with increased temporal resolution in future experiments, to further explore the vacuum activity associated with breakdowns.

AB - The occurrence of vacuum arcs or radio frequency (rf) breakdowns is one of the most prevalent factors limiting the high-gradient performance of normal conducting rf cavities in particle accelerators. In this paper, we search for the existence of previously unrecognized features related to the incidence of rf breakdowns by applying a machine learning strategy to high-gradient cavity data from CERN’s test stand for the Compact Linear Collider (CLIC). By interpreting the parameters of the learned models with explainable artificial intelligence (AI), we reverse-engineer physical properties for deriving fast, reliable, and simple rule–based models. Based on 6 months of historical data and dedicated experiments, our models show fractions of data with a high influence on the occurrence of breakdowns. Specifically, it is shown that the field emitted current following an initial breakdown is closely related to the probability of another breakdown occurring shortly thereafter. Results also indicate that the cavity pressure should be monitored with increased temporal resolution in future experiments, to further explore the vacuum activity associated with breakdowns.

KW - Surfaces and Interfaces

KW - Physics and Astronomy (miscellaneous)

KW - Nuclear and High Energy Physics

U2 - 10.1103/physrevaccelbeams.25.104601

DO - 10.1103/physrevaccelbeams.25.104601

M3 - Journal article

VL - 25

JO - Physical Review Accelerators and Beams

JF - Physical Review Accelerators and Beams

SN - 2469-9888

IS - 10

M1 - 104601

ER -