Paper | Title | Page |
---|---|---|
MOPOPT063 | Reconstruction of Beam Parameters from Betatron Radiation Using Maximum Likelihood Estimation and Machine Learning | 407 |
|
||
Funding: US Department of Energy, Division of High Energy Physics, Contract No. DE-SC0009914 STFC Liver-pool Centre for Doctoral Training on Data Intensive Science, grant agreement ST/P006752/1 Betatron radiation that arises during plasma wakefield acceleration can be measured by a UCLA-built Compton spectrometer, which records the energy and angular position of incoming photons. Because information about the properties of the beam is encoded in the betatron radiation, measurements of the radiation can be used to reconstruct beam parameters. One method of extracting information about beam parameters from measurements of radiation is maximum likelihood estimation (MLE), a statistical technique which is used to determine unknown parameters from a distribution of observed data. In addition, machine learning methods, which are increasingly being implemented for different fields of beam diagnostics, can also be applied. We assess the ability of both MLE and other machine learning methods to accurately extract beam parameters from measurements. |
||
DOI • | reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-MOPOPT063 | |
About • | Received ※ 08 June 2022 — Revised ※ 15 June 2022 — Accepted ※ 24 June 2022 — Issue date ※ 26 June 2022 | |
Cite • | reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml) | |
WEPOST040 | Comparing Methods of Recovering Gamma Energy Distributions from PEDRO Spectrometer Responses | 1784 |
|
||
To calculate the energy levels of the photons emitted from high-energy particle interactions, the new pair spectrometer (PEDRO) channels the photons through several Beryllium nuclear fields to produce electron-positron pairs through the nuclear field interaction. This project compared several methods of reconstruction and determined which best predicts original energy distributions based on simulated spectra. These methods included using Maximum Likelihood Estimation, Machine Learning, and directly analyzing a response matrix that modeled PEDRO’s response to any photon energy distribution. We report that performing the direct analysis, also known as QR decomposition, on a PEDRO-generated spectrum provides by far the most accurate calculation of the spectrum’s original energy distribution. These methods were tested against results from experimental cases, including Nonlinear Compton Scattering and Filamentation. | ||
DOI • | reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2022-WEPOST040 | |
About • | Received ※ 15 June 2022 — Revised ※ 01 July 2022 — Accepted ※ 08 July 2022 — Issue date ※ 08 July 2022 | |
Cite • | reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml) | |