IBIC2022 - List of Authors (Hall, C.C.)

Paper	Title	Page
MOP10	Removing Noise in BPM Measurements with Variational Autoencoders	43
	J.P. Edelen, J.A. Einstein-Curtis, C.C. Hall, M.J. Henderson RadiaSoft LLC, Boulder, Colorado, USA A.L. Romanov Fermilab, Batavia, Illinois, USA
	Funding: This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of High Energy Physics under Award Number DE-SC0021699. Noise in beam measurements is an ever-present challenge in accelerator operations. In addition to the challenges presented by hardware and signal processing, new operational regimes, such as ultra-short bunches, create additional difficulties in routine beam measurements. Techniques in machine learning have been successfully applied in other domains to overcome challenges inherent in noisy data. Variational autoencoders (VAEs) are shown to be capable of removing significant leevels of noise. A VAE can be used as a pre-processing tool for noise removal before the de-noised data is analyzed via other methods, or the VAE can be directly used to make beam dynamics measurements. Here we present the use of VAEs as a tool for addressing noise in BPM measurements.
DOI •	reference for this paper ※ doi:10.18429/JACoW-IBIC2022-MOP10
About •	Received ※ 29 August 2022 — Revised ※ 10 September 2022 — Accepted ※ 11 September 2022 — Issue date ※ 24 November 2022
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

Paper

Title

Page

Removing Noise in BPM Measurements with Variational Autoencoders

43

J.P. Edelen, J.A. Einstein-Curtis, C.C. Hall, M.J. Henderson
RadiaSoft LLC, Boulder, Colorado, USA
A.L. Romanov
Fermilab, Batavia, Illinois, USA

Funding: This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of High Energy Physics under Award Number DE-SC0021699.
Noise in beam measurements is an ever-present challenge in accelerator operations. In addition to the challenges presented by hardware and signal processing, new operational regimes, such as ultra-short bunches, create additional difficulties in routine beam measurements. Techniques in machine learning have been successfully applied in other domains to overcome challenges inherent in noisy data. Variational autoencoders (VAEs) are shown to be capable of removing significant leevels of noise. A VAE can be used as a pre-processing tool for noise removal before the de-noised data is analyzed via other methods, or the VAE can be directly used to make beam dynamics measurements. Here we present the use of VAEs as a tool for addressing noise in BPM measurements.

DOI •

reference for this paper ※ doi:10.18429/JACoW-IBIC2022-MOP10

About •

Received ※ 29 August 2022 — Revised ※ 10 September 2022 — Accepted ※ 11 September 2022 — Issue date ※ 24 November 2022

Cite •

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