NAPAC2019 - Table of Session: THXBA (Thursday Parallel Session 1)

Paper	Title	Page
THXBA1	Machine Learning Demonstrations on Accelerators
	A.L. Edelen SLAC, Menlo Park, California, USA
	Machine learning has been used in various ways to improve acclerator operation including the development of surrogate models to improve real-time modeling, advanced optimization of accelerator operating configurations such as quadrupole or undulator strengths, development of virtual diagnostics to ’measure’ accelerator and beam parameters, and prognostics to improve operating time.
	Slides THXBA1 [31.075 MB]
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THXBA2	Analysis of Beam Position Monitor Requirements with Bayesian Gaussian Regression	912
WEPLM09	use link to see paper's listing under its alternate paper code
	Y. Li, R.S. Rainer BNL, Upton, New York, USA W.X. Cheng ANL, Lemont, Illinois, USA Y. Hao FRIB, East Lansing, Michigan, USA
	Funding: This research is supported by U.S. Department of Energy under Contract No. DE-SC0012704, and the NSF under Cooperative Agreement PHY-1102511. With a Bayesian Gaussian regression approach, a systematic method for analyzing a storage ring’s beam position monitor (BPM) system requirements has been developed. The ultimate performance of a ring-based accelerator, based on brightness or luminosity, is determined not only by global parameters, but also by local beam properties at some particular points of interest (POI). BPMs used for monitoring the beam properties, however, can not be located at these points. Therefore, the underlying and fundamental purpose of a BPM system is to predict whether the beam properties at POIs reach their desired values. The prediction process is a regression problem with BPM readings as the training data, but containing random noise. A Bayesian Gaussian regression approach can determine the probability distribution of the predictive errors, which can be used to conversely analyze the BPM system requirements. This approach is demonstrated by using turn-by-turn data to reconstruct a linear optics model, and predict the brightness degradation for a ring-based light source. The quality of BPMs was found to be more important than their quantity in mitigating predictive errors.
	Slides THXBA2 [3.205 MB]
	Poster THXBA2 [7.083 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-NAPAC2019-THXBA2
About •	paper received ※ 16 August 2019 paper accepted ※ 04 September 2019 issue date ※ 08 October 2019
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THXBA3	Adaptive Machine Learning and Feedback Control for Automatic Particle Accelerator Tuning	916
	A. Scheinker LANL, Los Alamos, New Mexico, USA
	Free electron lasers (FEL) and plasma wakefield accelerators (PWA) are creating more and more complicated electron bunch configurations, including multi-color modes for FELs such as LCLS and LCLS-II and custom tailored bunch current profiles for PWAs such as FACET-II. These accelerators are also producing shorter and higher intensity bunches than before and require an ability to quickly switch between many different users with various specific phase space requirements. For some very exotic setups it can take hours of tuning to provide the beams that users require. In this work, we present results adaptive machine learning and model independent feedback techniques and their application in both the LCLS and European XFEL to 1) control electron bunch phase space to create desired current profiles and energy spreads by tuning FEL components automatically, 2) maximize the average pulse output energy of FELs by automatically tuning over 100 components simultaneously, 3) preliminary results on utilizing these techniques for non-invasive electron bunch longitudinal phase space diagnostics at PWAs.
	Slides THXBA3 [8.110 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-NAPAC2019-THXBA3
About •	paper received ※ 27 August 2019 paper accepted ※ 15 September 2019 issue date ※ 08 October 2019
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THXBA4	Update on BPM Signal Processing Circuitry Development at AWA	919
	W. Liu, M.E. Conde, D.S. Doran, G. Ha, J.G. Power, J.H. Shao, C. Whiteford, E.E. Wisniewski ANL, Lemont, Illinois, USA C. Jing Euclid Beamlabs LLC, Bolingbrook, USA
	Funding: The US Department of Energy, Office of Science Beam position monitor (BPM) is widely used in accelerator facilities worldwide. It is a device which is capable of providing, non-destructively, accurate beam centroid and charge information of a passing charged beam. A typical BPM system contains customized hardware and specialized processing electronics. The cost is often too high for small facilities to afford them. As a small facility, Argonne Wakefield Accelerator (AWA) decided to develop a solution with high cost-efficiency to fit in its budget. Some details about the development are presented in this paper.
	Slides THXBA4 [8.544 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-NAPAC2019-THXBA4
About •	paper received ※ 29 August 2019 paper accepted ※ 31 August 2019 issue date ※ 08 October 2019
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THXBA5	Electro-Optic Sampling Beam Position Monitor
	K.D. Hunt-Stone, R.A. Ariniello, J.R. Cary, C.E. Doss CIPS, Boulder, Colorado, USA J.R. Cary Tech-X, Boulder, Colorado, USA M.D. Litos Colorado University at Boulder, Boulder, Colorado, USA
	Funding: This work was supported in part by the U.S. D.O.E. grant number DE-SC0017906. Electron beam-driven plasma wakefield accelerator (PWFA) experiments at SLAC’s FACET-II research facility will require diagnostics that can measure the transverse position of both the drive beam and the witness beam in a single shot. This is a challenge for ordinary beam position monitors due to the close temporal spacing between the two bunches, usually on the order of 300 fs. Here we will discuss the concept for an electro-optic sampling beam position monitor (EOS-BPM) that can measure the transverse position of the individual bunches with roughly 10 µm spatial resolution, and 50 fs temporal resolution. The EOS-BPM has the advantage of being a non-destructive, single shot measurement. It uses two EO crystals on either side of the beamline. The half-cycle THz fields of the electron beams induce a birefringence in the crystals which are probed by a chirped laser pulse. The longitudinal current profile is spectrally encoded into the probe laser, while the transverse position for each bunch is encoded in the relative strength of the signal in either crystal. We present simulations demonstrating the effectiveness of an EOS-BPM in the context of PWFA experiments planned for FACET-II.
	Slides THXBA5 [23.339 MB]
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

Paper

Title

Page

THXBA1

Machine Learning Demonstrations on Accelerators

A.L. Edelen
SLAC, Menlo Park, California, USA

Machine learning has been used in various ways to improve acclerator operation including the development of surrogate models to improve real-time modeling, advanced optimization of accelerator operating configurations such as quadrupole or undulator strengths, development of virtual diagnostics to ’measure’ accelerator and beam parameters, and prognostics to improve operating time.

Slides THXBA1 [31.075 MB]

Export •

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

THXBA2

Analysis of Beam Position Monitor Requirements with Bayesian Gaussian Regression

912

WEPLM09

use link to see paper's listing under its alternate paper code

Y. Li, R.S. Rainer
BNL, Upton, New York, USA
W.X. Cheng
ANL, Lemont, Illinois, USA
Y. Hao
FRIB, East Lansing, Michigan, USA

Funding: This research is supported by U.S. Department of Energy under Contract No. DE-SC0012704, and the NSF under Cooperative Agreement PHY-1102511.
With a Bayesian Gaussian regression approach, a systematic method for analyzing a storage ring’s beam position monitor (BPM) system requirements has been developed. The ultimate performance of a ring-based accelerator, based on brightness or luminosity, is determined not only by global parameters, but also by local beam properties at some particular points of interest (POI). BPMs used for monitoring the beam properties, however, can not be located at these points. Therefore, the underlying and fundamental purpose of a BPM system is to predict whether the beam properties at POIs reach their desired values. The prediction process is a regression problem with BPM readings as the training data, but containing random noise. A Bayesian Gaussian regression approach can determine the probability distribution of the predictive errors, which can be used to conversely analyze the BPM system requirements. This approach is demonstrated by using turn-by-turn data to reconstruct a linear optics model, and predict the brightness degradation for a ring-based light source. The quality of BPMs was found to be more important than their quantity in mitigating predictive errors.

Slides THXBA2 [3.205 MB]

Poster THXBA2 [7.083 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-NAPAC2019-THXBA2

About •

paper received ※ 16 August 2019 paper accepted ※ 04 September 2019 issue date ※ 08 October 2019

Export •

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

THXBA3

Adaptive Machine Learning and Feedback Control for Automatic Particle Accelerator Tuning

916

A. Scheinker
LANL, Los Alamos, New Mexico, USA

Free electron lasers (FEL) and plasma wakefield accelerators (PWA) are creating more and more complicated electron bunch configurations, including multi-color modes for FELs such as LCLS and LCLS-II and custom tailored bunch current profiles for PWAs such as FACET-II. These accelerators are also producing shorter and higher intensity bunches than before and require an ability to quickly switch between many different users with various specific phase space requirements. For some very exotic setups it can take hours of tuning to provide the beams that users require. In this work, we present results adaptive machine learning and model independent feedback techniques and their application in both the LCLS and European XFEL to 1) control electron bunch phase space to create desired current profiles and energy spreads by tuning FEL components automatically, 2) maximize the average pulse output energy of FELs by automatically tuning over 100 components simultaneously, 3) preliminary results on utilizing these techniques for non-invasive electron bunch longitudinal phase space diagnostics at PWAs.

Slides THXBA3 [8.110 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-NAPAC2019-THXBA3

About •

paper received ※ 27 August 2019 paper accepted ※ 15 September 2019 issue date ※ 08 October 2019

Export •

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

THXBA4

Update on BPM Signal Processing Circuitry Development at AWA

919

W. Liu, M.E. Conde, D.S. Doran, G. Ha, J.G. Power, J.H. Shao, C. Whiteford, E.E. Wisniewski
ANL, Lemont, Illinois, USA
C. Jing
Euclid Beamlabs LLC, Bolingbrook, USA

Funding: The US Department of Energy, Office of Science
Beam position monitor (BPM) is widely used in accelerator facilities worldwide. It is a device which is capable of providing, non-destructively, accurate beam centroid and charge information of a passing charged beam. A typical BPM system contains customized hardware and specialized processing electronics. The cost is often too high for small facilities to afford them. As a small facility, Argonne Wakefield Accelerator (AWA) decided to develop a solution with high cost-efficiency to fit in its budget. Some details about the development are presented in this paper.

Slides THXBA4 [8.544 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-NAPAC2019-THXBA4

About •

paper received ※ 29 August 2019 paper accepted ※ 31 August 2019 issue date ※ 08 October 2019

Export •

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

THXBA5

Electro-Optic Sampling Beam Position Monitor

K.D. Hunt-Stone, R.A. Ariniello, J.R. Cary, C.E. Doss
CIPS, Boulder, Colorado, USA
J.R. Cary
Tech-X, Boulder, Colorado, USA
M.D. Litos
Colorado University at Boulder, Boulder, Colorado, USA

Funding: This work was supported in part by the U.S. D.O.E. grant number DE-SC0017906.
Electron beam-driven plasma wakefield accelerator (PWFA) experiments at SLAC’s FACET-II research facility will require diagnostics that can measure the transverse position of both the drive beam and the witness beam in a single shot. This is a challenge for ordinary beam position monitors due to the close temporal spacing between the two bunches, usually on the order of 300 fs. Here we will discuss the concept for an electro-optic sampling beam position monitor (EOS-BPM) that can measure the transverse position of the individual bunches with roughly 10 µm spatial resolution, and 50 fs temporal resolution. The EOS-BPM has the advantage of being a non-destructive, single shot measurement. It uses two EO crystals on either side of the beamline. The half-cycle THz fields of the electron beams induce a birefringence in the crystals which are probed by a chirped laser pulse. The longitudinal current profile is spectrally encoded into the probe laser, while the transverse position for each bunch is encoded in the relative strength of the signal in either crystal. We present simulations demonstrating the effectiveness of an EOS-BPM in the context of PWFA experiments planned for FACET-II.

Slides THXBA5 [23.339 MB]

Export •

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