Author: Cope, T.M.
Paper Title Page
WEPOW055 Bayesian Optimization of FEL Performance at LCLS 2972
 
  • M.W. McIntire, T.M. Cope, D.F. Ratner
    SLAC, Menlo Park, California, USA
  • S. Ermon
    Stanford University, Stanford, California, USA
  
  Funding: Research is supported by the U.S. Department of Energy under Contract No. DE-AC02-76SF00515.
The LCLS free-electron laser at SLAC is tuned via a huge number of parameters such as energy and magnet settings. Much of this tuning, including quadrupole magnet settings, is typically done by hand by the LCLS operators. In this paper we introduce an automated tuning system using Bayesian optimization, and describe its application to the optimization of noisy objectives such as FEL performance. We demonstrate with preliminary results from our implementation at LCLS that this system can improve both the speed of tuning procedures as well as the quality of the resulting solution. 		 
DOI • reference for this paper ※ DOI:10.18429/JACoW-IPAC2016-WEPOW055  
Export • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
WEPOY036 Progress in Automatic Software-based Optimization of Accelerator Performance 3064
 
  • S.I. Tomin, G. Geloni
    XFEL. EU, Hamburg, Germany
  • I.V. Agapov, I. Zagorodnov
    DESY, Hamburg, Germany
  • W.S. Colocho, T.M. Cope, A.B. Egger, D.F. Ratner
    SLAC, Menlo Park, California, USA
  • Y.A. Fomin, Y.V. Krylov, A.G. Valentinov
    NRC, Moscow, Russia
  
  Funding: partial support from Ioffe Roentgen Institute grant EDYN EMRAD
For modern linac- and storage-ring-based light sources certain amount of empirical tuning is used to reach ultimate performance. The possibility to perform such empirical tuning by automatic methods has now been demonstrated by several authors (e.g. I.Agapov et al. in proc IPAC 2015). In this paper we present the progress in development of our automatic optimisation software based on OCELOT and its applications to SASE FEL optimization at FLASH and LCLS, and its potential for storage ring optimization. 		 
DOI • reference for this paper ※ DOI:10.18429/JACoW-IPAC2016-WEPOY036  
Export • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
THPOY051 Upgrades to the SPEAR3 Single-Photon Bunch Measurement System 4223
 
  • T.M. Cope, S. Allison, W.J. Corbett, Y.H. Xu
    SLAC, Menlo Park, California, USA
  
  The SPEAR3 accelerator uses a Single Photon Time-Correlated Counting (TCSPC) system to accurately measure the time profile of electron bunches circulating in the storage ring. The detection hardware uses the PicoHarp 300 TCSPC processor module initially equipped with an available Hamamatsu H7360-01 photon counting head. The H7360-01 was later replaced with a PicoQuant Hybrid-06 PMA to decrease single-photon arrival time jitter. At the same time we adopted an EPICS-based TCSPC software package developed at DIAMOND for robust data acquisition and display. In this paper we report on recent beam profile measurements and upgrades to the data acquisition software system including installation of a local EPICS IOC for real-time access to the bunch profile from SLAC's centralized Accelerator Control Room (ACR). High-level operator interface and monitoring applications developed in Python are discussed.  
DOI • reference for this paper ※ DOI:10.18429/JACoW-IPAC2016-THPOY051  
Export • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)