NAPAC2019 - List of Authors (Scheinker, A.)

Paper	Title	Page
WEXBB1	Adaptive Machine Learning and Automatic Tuning of Intense Electron Bunches in Particle Accelerators	609
	A. Scheinker LANL, Los Alamos, New Mexico, USA
	Machine learning and in particular neural networks, have been around for a very long time. In recent years, thanks to growth in computing power, neural networks have reshaped many fields of research, including self driving cars, computers playing complex video games, image identification, and even particle accelerators. In this tutorial, I will first present an introduction to machine learning for beginners and will also touch on a few aspects of adaptive control theory. I will then introduce some problems in particle accelerators and present how they have been approached utilizing machine learning techniques as well as adaptive machine learning approaches, for automatically tuning extremely short and high intensity electron bunches in free electron lasers.
	Slides WEXBB1 [58.913 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-NAPAC2019-WEXBB1
About •	paper received ※ 28 August 2019 paper accepted ※ 06 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)

Paper

Title

Page

Adaptive Machine Learning and Automatic Tuning of Intense Electron Bunches in Particle Accelerators

609

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

Machine learning and in particular neural networks, have been around for a very long time. In recent years, thanks to growth in computing power, neural networks have reshaped many fields of research, including self driving cars, computers playing complex video games, image identification, and even particle accelerators. In this tutorial, I will first present an introduction to machine learning for beginners and will also touch on a few aspects of adaptive control theory. I will then introduce some problems in particle accelerators and present how they have been approached utilizing machine learning techniques as well as adaptive machine learning approaches, for automatically tuning extremely short and high intensity electron bunches in free electron lasers.

Slides WEXBB1 [58.913 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-NAPAC2019-WEXBB1

About •

paper received ※ 28 August 2019 paper accepted ※ 06 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)