Author: Fazio, M.A.
Paper Title Page
WE2AA04 Data Analysis and Control of an MeV Ultrafast Electron Diffraction System using Machine Learning 650
  • T.B. Bolin, S. Biedron, M.A. Fazio, M. Martínez-Ramón, S.I. Sosa Guitron
    UNM-ECE, Albuquerque, USA
  • M. Babzien, M.G. Fedurin, J.J. Li, M.A. Palmer
    BNL, Upton, New York, USA
  • S. Biedron
    Element Aero, Chicago, USA
  • S. Biedron
    UNM-ME, Albuquerque, New Mexico, USA
  MeV ultrafast electron diffraction (MUED) is a pump-probe material characterization technique to study ultrafast lattice dynamics with high temporal and spatial resolution. It is a relatively young technology that has the potential to shed light onto some of the most puzzling problems in physics. This complex instrument can be advanced into a turn-key high-throughput tool with the aid of machine learning (ML) mechanisms together with high-performance computing. The MUED instrument located in the Accelerator Test Facility of Brookhaven National Laboratory was employed in this work to test different ML approaches for both data analysis and control. We characterized three materials using MUED: graphite, black phosphorous and gold thin films. Diffraction patterns were acquired in single shot mode and different ML methodologies were applied to reduce image noise. Convolutional neural network autoenconder and variational autoenconder models were utilized to extract the noise features and increase the signal-to-noise ratio. The energy jitter of the electron beam was analyzed after noise reduction of the single shot diffraction patterns.  
video icon
  please see instructions how to view/control embeded videos  
slides icon Slides WE2AA04 [12.865 MB]  
DOI • reference for this paper ※  
About • Received ※ 30 August 2022 — Revised ※ 02 September 2022 — Accepted ※ 15 September 2022 — Issue date ※ 20 September 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)