FEL2019 - List of Authors (Hiller, N.)

Paper	Title	Page
WEP038	Commissioning and Stability Studies of the SwissFEL Bunch-Separation System	404
	M. Paraliev, S. Dordevic, R. Ganter, C.H. Gough, N. Hiller, R.A. Krempaská, D. Voulot PSI, Villigen PSI, Switzerland
	SwissFEL is a linear electron accelerator based, X-ray Free Electron Laser at the Paul Scherrer Institute, Switzerland. It is a user oriented facility capable of producing short, high brightness X-ray pulses covering the spectral range from 1 to 50 Å. SwissFEL is designed to run in two electron bunch mode in order to serve simultaneously two experimental beamline stations (hard and soft X-ray one) at its full repetition rate. Two closely spaced (28 ns) electron bunches are accelerated in one RF macro pulse up to 3 GeV. A high stability resonant kicker system and a Lambertson septum magnet are used to separate the bunches and to send them to their respective beamlines. With the advancement of the construction of the second beamline (Athos) the bunch-separation system was successfully commissioned. In order to confirm that the beam separation process is fully transparent a stability study of the electron beam and the free electron laser in the main beamline (Aramis) was done.
	Poster WEP038 [0.945 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-FEL2019-WEP038
About •	paper received ※ 19 August 2019 paper accepted ※ 25 August 2019 issue date ※ 05 November 2019
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

WED01	Experience with Short-Period, Small Gap Undulators at the SwissFEL Aramis Beamline	564
	T. Schmidt, M. Aiba, A.D. Alarcon, C. Arrell, S. Bettoni, M. Calvi, A. Cassar, E. Ferrari, R. Follath, R. Ganter, N. Hiller, P.N. Juranič, C. Kittel, F. Löhl, E. Prat, S. Reiche, T. Schietinger, D. Voulot, U.H. Wagner PSI, Villigen PSI, Switzerland N.J. Sammut University of Malta, Faculty of Engineering, Msida, Malta
	The SwissFEL Aramis beamline provides hard X-ray FEL radiation down to 1 Angström with 5.8 GeV and short period, 15 mm, in-vacuum undulators (U15). To reach the maximum designed K-value of 1.8 the U15s have to be operated with vacuum gaps down to 3.0 mm. The thirteen-undulator modules are 4 m long and each of them is equipped with a pair of permanent magnet quadrupoles at the two ends, aligned magnetically to the undulator axis. Optical systems and dedicated photon diagnostics are used to check the alignment and improve the K-value calibration. In this talk the main steps of the undulator commissioning will be recalled and a systematic comparison between the magnetic results and the electron and photon based measurements will be reported to highlight achievements and open issues.
	Slides WED01 [13.825 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-FEL2019-WED01
About •	paper received ※ 28 August 2019 paper accepted ※ 06 November 2019 issue date ※ 05 November 2019
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

THP061	Bayesian Optimisation for Fast and Safe Parameter Tuning of SwissFEL	707
	J. Kirschner, A. Krause, M. Mutný, M. Nonnenmacher ETH, Zurich, Switzerland A. Adelmann, N. Hiller, R. Ischebeck PSI, Villigen PSI, Switzerland
	Parameter tuning is a notoriously time-consuming task in accelerator facilities. As tool for global optimization with noisy evaluations, Bayesian optimization was recently shown to outperform alternative methods. By learning a model of the underlying function using all available data, the next evaluation can be chosen carefully to find the optimum with as few steps as possible and without violating any safety constraints. However, the per-step computation time increases significantly with the number of parameters and the generality of the approach can lead to slow convergence on functions that are easier to optimize. To overcome these limitations, we divide the global problem into sequential subproblems that can be solved efficiently using safe Bayesian optimization. This allows us to trade off local and global convergence and to adapt to additional structure in the objective function. Further, we provide slice-plots of the function as user feedback during the optimization. We showcase how we use our algorithm to tune up the FEL output of SwissFEL with up to 40 parameters simultaneously, and reach convergence within reasonable tuning times in the order of 30 minutes (< 2000 steps).
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-FEL2019-THP061
About •	paper received ※ 13 August 2019 paper accepted ※ 27 August 2019 issue date ※ 05 November 2019
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

Paper

Title

Page

Commissioning and Stability Studies of the SwissFEL Bunch-Separation System

404

M. Paraliev, S. Dordevic, R. Ganter, C.H. Gough, N. Hiller, R.A. Krempaská, D. Voulot
PSI, Villigen PSI, Switzerland

SwissFEL is a linear electron accelerator based, X-ray Free Electron Laser at the Paul Scherrer Institute, Switzerland. It is a user oriented facility capable of producing short, high brightness X-ray pulses covering the spectral range from 1 to 50 Å. SwissFEL is designed to run in two electron bunch mode in order to serve simultaneously two experimental beamline stations (hard and soft X-ray one) at its full repetition rate. Two closely spaced (28 ns) electron bunches are accelerated in one RF macro pulse up to 3 GeV. A high stability resonant kicker system and a Lambertson septum magnet are used to separate the bunches and to send them to their respective beamlines. With the advancement of the construction of the second beamline (Athos) the bunch-separation system was successfully commissioned. In order to confirm that the beam separation process is fully transparent a stability study of the electron beam and the free electron laser in the main beamline (Aramis) was done.

Poster WEP038 [0.945 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-FEL2019-WEP038

About •

paper received ※ 19 August 2019 paper accepted ※ 25 August 2019 issue date ※ 05 November 2019

Export •

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

WED01

Experience with Short-Period, Small Gap Undulators at the SwissFEL Aramis Beamline

564

T. Schmidt, M. Aiba, A.D. Alarcon, C. Arrell, S. Bettoni, M. Calvi, A. Cassar, E. Ferrari, R. Follath, R. Ganter, N. Hiller, P.N. Juranič, C. Kittel, F. Löhl, E. Prat, S. Reiche, T. Schietinger, D. Voulot, U.H. Wagner
PSI, Villigen PSI, Switzerland
N.J. Sammut
University of Malta, Faculty of Engineering, Msida, Malta

The SwissFEL Aramis beamline provides hard X-ray FEL radiation down to 1 Angström with 5.8 GeV and short period, 15 mm, in-vacuum undulators (U15). To reach the maximum designed K-value of 1.8 the U15s have to be operated with vacuum gaps down to 3.0 mm. The thirteen-undulator modules are 4 m long and each of them is equipped with a pair of permanent magnet quadrupoles at the two ends, aligned magnetically to the undulator axis. Optical systems and dedicated photon diagnostics are used to check the alignment and improve the K-value calibration. In this talk the main steps of the undulator commissioning will be recalled and a systematic comparison between the magnetic results and the electron and photon based measurements will be reported to highlight achievements and open issues.

Slides WED01 [13.825 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-FEL2019-WED01

About •

paper received ※ 28 August 2019 paper accepted ※ 06 November 2019 issue date ※ 05 November 2019

Export •

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

THP061

Bayesian Optimisation for Fast and Safe Parameter Tuning of SwissFEL

707

J. Kirschner, A. Krause, M. Mutný, M. Nonnenmacher
ETH, Zurich, Switzerland
A. Adelmann, N. Hiller, R. Ischebeck
PSI, Villigen PSI, Switzerland

Parameter tuning is a notoriously time-consuming task in accelerator facilities. As tool for global optimization with noisy evaluations, Bayesian optimization was recently shown to outperform alternative methods. By learning a model of the underlying function using all available data, the next evaluation can be chosen carefully to find the optimum with as few steps as possible and without violating any safety constraints. However, the per-step computation time increases significantly with the number of parameters and the generality of the approach can lead to slow convergence on functions that are easier to optimize. To overcome these limitations, we divide the global problem into sequential subproblems that can be solved efficiently using safe Bayesian optimization. This allows us to trade off local and global convergence and to adapt to additional structure in the objective function. Further, we provide slice-plots of the function as user feedback during the optimization. We showcase how we use our algorithm to tune up the FEL output of SwissFEL with up to 40 parameters simultaneously, and reach convergence within reasonable tuning times in the order of 30 minutes (< 2000 steps).

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-FEL2019-THP061

About •

paper received ※ 13 August 2019 paper accepted ※ 27 August 2019 issue date ※ 05 November 2019

Export •

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