ICAP2018 - Classification: E-2 Design, Optimization, Control

Paper	Title	Page
SAPAF01	Genetic Algorithm Enhanced by Machine Learning in Dynamic Aperture Optimization	8
	Y. Li, W.X. Cheng, R.S. Rainer, L. Yu BNL, Upton, Long Island, New York, USA
	Funding: This work was supported by Department of Energy Contract No. DE-SC0012704 With the aid of machine learning techniques, the genetic algorithm has been enhanced and applied to the multi-objective optimization problem presented by the dynamic aperture of the NSLS-II Ring. During the evolution employed by the genetic algorithm, the population is classified into different clusters. The clusters with top average fitness are given elite status. Intervention is implemented by repopulating some potentially competitive candidates based on the accumulated data. These candidates replace randomly selected candidates among the original data pool. The average fitness of the population is improved while diversity is not lost. The quality of the population increases and produces more competitive descendants accelerating the evolution process significantly. When identifying the distribution of optimal candidates, they appear to be located in isolated islands within the search space. Some of these optimal candidates have been experimentally confirmed at the NSLS-II storage ring. The machine learning techniques that exploit the genetic algorithm can also be used in other population-based optimization problems such as particle swarm algorithm.
	Slides SAPAF01 [6.696 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-ICAP2018-SAPAF01
About •	paper received ※ 15 October 2018 paper accepted ※ 24 October 2018 issue date ※ 26 January 2019
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

SAPAF02	Optimization of Heavy-Ion Synchrotrons Using Nature-Inspired Algorithms and Machine Learning	15
	S. Appel, W. Geithner, S. Reimann, M. Sapinski, R. Singh, D.M. Vilsmeier GSI, Darmstadt, Germany
	The application of machine learning and nature-inspired optimization methods, like for example genetic algorithms (GA) and particle swarm optimization (PSO) can be found in various scientific/technical areas. In recent years, those approaches are finding application in accelerator physics to a greater extent. In this report, nature-inspired optimization as well as the machine learning will be shortly introduced and their application to the accelerator facility at GSI/FAIR will be presented. For the heavy-ion synchrotron SIS18 at GSI, the multi-objective GA/PSO optimization resulted in a significant improvement of multi-turn injection performance and subsequent transmission for intense beams. An automated injection optimization with genetic algorithms at the CRYRING@ESR ion storage ring has been performed. The usage of machine learning for a beam diagnostic application, where reconstruction of space-charge distorted beam profiles from ionization profile monitors is performed, will also be shown. First results and the experience gained will be presented.
	Slides SAPAF02 [2.642 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-ICAP2018-SAPAF02
About •	paper received ※ 16 October 2018 paper accepted ※ 27 January 2019 issue date ※ 26 January 2019
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

SAPAF03	Comparison of Model-Based and Heuristic Optimization Algorithms Applied to Photoinjectors Using Libensemble	22
	N.R. Neveu IIT, Chicago, Illinois, USA S. T. P. Hudson, J.M. Larson ANL, Argonne, Illinois, USA L.K. Spentzouris Illinois Institute of Technology, Chicago, Illinois, USA
	Funding: U.S. DOE, OS, contract DE-AC02-06CH11357 and grant DE-SC0015479. Genetic algorithms are common and often used in the accelerator community. They require large amounts of computational resources and empirical adjustment of hyperparameters. Model based methods are significantly more efficient, but often labeled as unreliable for the nonlinear or unsmooth problems that can be found in accelerator physics. We investigate the behavior of both approaches using a photoinjector operated in the space charge dominated regime. All optimization runs are coordinated and managed by the Python library libEnsemble, which is developed at Argonne National Laboratory.
	Slides SAPAF03 [0.653 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-ICAP2018-SAPAF03
About •	paper received ※ 11 November 2018 paper accepted ※ 19 November 2018 issue date ※ 26 January 2019
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

SAPAF04	Single Objective Genetic Optimization of an 85% Efficient Klystron	25
	A. Jensen, J.J. Petillo Leidos Corp, Billerica, MA, USA R.L. Ives, M.E. Read CCR, San Mateo, California, USA J. Neilson SLAC, Menlo Park, California, USA
	Overall efficiency is a critical priority for the next generation of particle accelerators as they push to higher and higher energies. In a large machine, even a small increase in efficiency of any subsystem or component can lead to a significant operational cost savings. The Core Oscillation Method (COM) and Bunch-Align-Compress (BAC) method have recently emerged as a means to greatly increase the efficiency of the klystron RF source for particle accelerators. The COM and BAC methods both work by uniquely tuning klystron cavity frequencies such that more particles from the anti-bunch are swept into the bunch before power is extracted from the beam. The single objective genetic algorithm from Sandia National Laboratory’s Dakota optimization library is used to optimize both COM and BAC based klystron designs to achieve 85% efficiency. The COM and BAC methods are discussed. Use of the Dakota optimization algorithm library from Sandia National Laboratory is discussed. Scalability of the optimization approach to High Performance Computing (HPC) is discussed. The optimization approach and optimization results are presented.
	Slides SAPAF04 [1.476 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-ICAP2018-SAPAF04
About •	paper received ※ 16 October 2018 paper accepted ※ 19 November 2018 issue date ※ 26 January 2019
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

SUPAF03	Optimization of Hadron Therapy Beamlines Using a Novel Fast Tracking Code for Beam Transport and Beam-Matter Interactions
	C. Hernalsteens, K. André CERN, Geneva, Switzerland V. Collignon, Q. Flandroy, B. Herregods IBA, Louvain-la-Neuve, Belgium R. Jungers, Z. Wang UCL, Louvain-la-Neuve, Belgium R. Tesse ULB - FSA - SMN, Bruxelles, Belgium
	The optimization of proton therapy beamlines challenges the traditional approach used in beam optics due to the very strict constraints on beam quality, especially for Pencil Beam Scanning, despite the large losses induced by the emittance increase coming from the energy degrader. In order to explore the performances of proton therapy beamlines, we proceed using a new fast beam tracking Python library coupled with a genetic algorithm. Global optimization algorithms such as the genetic algorithm or basin hopping schemes require numerous evaluations of the model and their practical implementations are limited by the computation time at each iteration. To overcome this limitation, while at the same time allowing an open-box user experience, a Python library has been developed, including transport models for the typical hadron therapy beamlines elements, as well as models for the computation of multiple Coulomb scattering. The Multi-Objective Genetic Algorithm (MOGA) allows to explore the parameter space in a global sense. This multi-objective algorithm enables the simultaneous optimization of complex constraints specific to proton therapy beamlines. Results for the IBA Proteus One system are presented and discussed in detail.
	Slides SUPAF03 [9.526 MB]
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPAG08	Uncertainty Quantification for the Fundamental Mode Spectrum of the European XFEL Cavities
	N. G. Georg, J. Corno, H. De Gersem, U. Römer, S. Schöps TEMF, TU Darmstadt, Darmstadt, Germany S. Gorgi Zadeh, U. van Rienen Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock, Germany A.A. Sulimov DESY, Hamburg, Germany
	Funding: The authors would like to acknowledge the support by the DFG (German Research Foundation) in the framework of the Scientific Network SCHM 3127/1,2 that provided the basis for this collaborative work. The fundamental mode spectrum of superconducting cavities is sensitive to small geometry deformations introduced by the manufacturing process. In this work we consider variations in the equatorial and iris radii of the 1.3 GHz TESLA cavities used at the European XFEL. The cavities with slightly perturbed geometry are simulated using a finite element based eigenvalue solver. Employing uncertainty quantification methods such as sparse-grids, statistical information about the fundamental mode spectrum can be efficiently calculated. Moreover, using global sensitivity analysis, in particular Sobol indices, the impact of the individual geometry parameters on the quantities of interest, i.e. resonance frequencies, field-flatness and the cell-to-cell coupling coefficient, can be computed. We will explain important aspects of the uncertainty quantification methodology and give numerical results for illustration.
	Slides TUPAG08 [0.672 MB]
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPAG14	Constrained Multi-Objective Shape Optimization of Superconducting RF Cavities to Counteract Dangerous Higher Order Modes	293
	M. Kranjcevic, P. Arbenz ETH, Zurich, Switzerland A. Adelmann PSI, Villigen PSI, Switzerland S. Gorgi Zadeh, U. van Rienen Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock, Germany
	High current storage rings, such as the Z operating mode of the FCC-ee, require superconducting radio frequency (RF) cavities that are optimized with respect to both the fundamental mode and the dangerous higher order modes. In order to optimize the shape of the RF cavity, a constrained multi-objective optimization problem is solved using a massively parallel implementation of an evolutionary algorithm. Additionally, a frequency-fixing scheme is employed to deal with the constraint on the frequency of the fundamental mode. Finally, the computed Pareto front approximation and an RF cavity shape with desired properties are shown.
	Slides TUPAG14 [3.001 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-ICAP2018-TUPAG14
About •	paper received ※ 19 October 2018 paper accepted ※ 10 December 2018 issue date ※ 26 January 2019
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

WEPAF04	Longitudinal Beam Dynamics in FRIB and ReA Linacs	330
	A.S. Plastun, P.N. Ostroumov, A.C.C. Villari, Q. Zhao FRIB, East Lansing, USA
	The Front-End and first three cryomodules of the Facility for Rare Isotope Beam (FRIB) at Michigan State University (MSU) commissioned in July, 2018. The paper describes the online tuning procedures of the longitudinal beam dynamics through the FRIB linac. These procedures include tuning of the accelerating field phases and amplitudes in the cavities. We developed an automated simulation-based tuning procedure for the multi-harmonic buncher. In order to tune the radio-frequency quadrupole (RFQ) we measured and calculated its threshold voltage and scanned its longitudinal acceptance. Tuning of the rebunchers and superconducting accelerating cavities is per-formed by means of the phase scans and Time-Of-Flight (TOF) beam energy measurements with beam position and phase monitors. While FRIB is being commissioned, the re-accelerator (ReA3) for rare isotope beams (RIBs) is being upgraded. We redesigned the ReA3 RFQ to improve its cooling system and provide reliable operation with 16.1 MHz prebunched ion beams with A/Q = 5. In order to provide matching of any ReA3 beam both to the following upgrade cryomodules and physics experiments’ requirements, room temperature rebuncher/debuncher is being designed. The design procedure includes the beam dynamics, electromagnetic, thermal and mechanical simulations and optimizations.
	Slides WEPAF04 [2.406 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-ICAP2018-WEPAF04
About •	paper received ※ 19 October 2018 paper accepted ※ 28 January 2019 issue date ※ 26 January 2019
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

Paper

Title

Page

Genetic Algorithm Enhanced by Machine Learning in Dynamic Aperture Optimization

8

Y. Li, W.X. Cheng, R.S. Rainer, L. Yu
BNL, Upton, Long Island, New York, USA

Funding: This work was supported by Department of Energy Contract No. DE-SC0012704
With the aid of machine learning techniques, the genetic algorithm has been enhanced and applied to the multi-objective optimization problem presented by the dynamic aperture of the NSLS-II Ring. During the evolution employed by the genetic algorithm, the population is classified into different clusters. The clusters with top average fitness are given elite status. Intervention is implemented by repopulating some potentially competitive candidates based on the accumulated data. These candidates replace randomly selected candidates among the original data pool. The average fitness of the population is improved while diversity is not lost. The quality of the population increases and produces more competitive descendants accelerating the evolution process significantly. When identifying the distribution of optimal candidates, they appear to be located in isolated islands within the search space. Some of these optimal candidates have been experimentally confirmed at the NSLS-II storage ring. The machine learning techniques that exploit the genetic algorithm can also be used in other population-based optimization problems such as particle swarm algorithm.

Slides SAPAF01 [6.696 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-ICAP2018-SAPAF01

About •

paper received ※ 15 October 2018 paper accepted ※ 24 October 2018 issue date ※ 26 January 2019

Export •

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

SAPAF02

Optimization of Heavy-Ion Synchrotrons Using Nature-Inspired Algorithms and Machine Learning

15

S. Appel, W. Geithner, S. Reimann, M. Sapinski, R. Singh, D.M. Vilsmeier
GSI, Darmstadt, Germany

The application of machine learning and nature-inspired optimization methods, like for example genetic algorithms (GA) and particle swarm optimization (PSO) can be found in various scientific/technical areas. In recent years, those approaches are finding application in accelerator physics to a greater extent. In this report, nature-inspired optimization as well as the machine learning will be shortly introduced and their application to the accelerator facility at GSI/FAIR will be presented. For the heavy-ion synchrotron SIS18 at GSI, the multi-objective GA/PSO optimization resulted in a significant improvement of multi-turn injection performance and subsequent transmission for intense beams. An automated injection optimization with genetic algorithms at the CRYRING@ESR ion storage ring has been performed. The usage of machine learning for a beam diagnostic application, where reconstruction of space-charge distorted beam profiles from ionization profile monitors is performed, will also be shown. First results and the experience gained will be presented.

Slides SAPAF02 [2.642 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-ICAP2018-SAPAF02

About •

paper received ※ 16 October 2018 paper accepted ※ 27 January 2019 issue date ※ 26 January 2019

Export •

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

SAPAF03

Comparison of Model-Based and Heuristic Optimization Algorithms Applied to Photoinjectors Using Libensemble

22

N.R. Neveu
IIT, Chicago, Illinois, USA
S. T. P. Hudson, J.M. Larson
ANL, Argonne, Illinois, USA
L.K. Spentzouris
Illinois Institute of Technology, Chicago, Illinois, USA

Funding: U.S. DOE, OS, contract DE-AC02-06CH11357 and grant DE-SC0015479.
Genetic algorithms are common and often used in the accelerator community. They require large amounts of computational resources and empirical adjustment of hyperparameters. Model based methods are significantly more efficient, but often labeled as unreliable for the nonlinear or unsmooth problems that can be found in accelerator physics. We investigate the behavior of both approaches using a photoinjector operated in the space charge dominated regime. All optimization runs are coordinated and managed by the Python library libEnsemble, which is developed at Argonne National Laboratory.

Slides SAPAF03 [0.653 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-ICAP2018-SAPAF03

About •

paper received ※ 11 November 2018 paper accepted ※ 19 November 2018 issue date ※ 26 January 2019

Export •

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

SAPAF04

Single Objective Genetic Optimization of an 85% Efficient Klystron

25

A. Jensen, J.J. Petillo
Leidos Corp, Billerica, MA, USA
R.L. Ives, M.E. Read
CCR, San Mateo, California, USA
J. Neilson
SLAC, Menlo Park, California, USA

Overall efficiency is a critical priority for the next generation of particle accelerators as they push to higher and higher energies. In a large machine, even a small increase in efficiency of any subsystem or component can lead to a significant operational cost savings. The Core Oscillation Method (COM) and Bunch-Align-Compress (BAC) method have recently emerged as a means to greatly increase the efficiency of the klystron RF source for particle accelerators. The COM and BAC methods both work by uniquely tuning klystron cavity frequencies such that more particles from the anti-bunch are swept into the bunch before power is extracted from the beam. The single objective genetic algorithm from Sandia National Laboratory’s Dakota optimization library is used to optimize both COM and BAC based klystron designs to achieve 85% efficiency. The COM and BAC methods are discussed. Use of the Dakota optimization algorithm library from Sandia National Laboratory is discussed. Scalability of the optimization approach to High Performance Computing (HPC) is discussed. The optimization approach and optimization results are presented.

Slides SAPAF04 [1.476 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-ICAP2018-SAPAF04

About •

paper received ※ 16 October 2018 paper accepted ※ 19 November 2018 issue date ※ 26 January 2019

Export •

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

SUPAF03

Optimization of Hadron Therapy Beamlines Using a Novel Fast Tracking Code for Beam Transport and Beam-Matter Interactions

C. Hernalsteens, K. André
CERN, Geneva, Switzerland
V. Collignon, Q. Flandroy, B. Herregods
IBA, Louvain-la-Neuve, Belgium
R. Jungers, Z. Wang
UCL, Louvain-la-Neuve, Belgium
R. Tesse
ULB - FSA - SMN, Bruxelles, Belgium

The optimization of proton therapy beamlines challenges the traditional approach used in beam optics due to the very strict constraints on beam quality, especially for Pencil Beam Scanning, despite the large losses induced by the emittance increase coming from the energy degrader. In order to explore the performances of proton therapy beamlines, we proceed using a new fast beam tracking Python library coupled with a genetic algorithm. Global optimization algorithms such as the genetic algorithm or basin hopping schemes require numerous evaluations of the model and their practical implementations are limited by the computation time at each iteration. To overcome this limitation, while at the same time allowing an open-box user experience, a Python library has been developed, including transport models for the typical hadron therapy beamlines elements, as well as models for the computation of multiple Coulomb scattering. The Multi-Objective Genetic Algorithm (MOGA) allows to explore the parameter space in a global sense. This multi-objective algorithm enables the simultaneous optimization of complex constraints specific to proton therapy beamlines. Results for the IBA Proteus One system are presented and discussed in detail.

Slides SUPAF03 [9.526 MB]

Export •

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

TUPAG08

Uncertainty Quantification for the Fundamental Mode Spectrum of the European XFEL Cavities

N. G. Georg, J. Corno, H. De Gersem, U. Römer, S. Schöps
TEMF, TU Darmstadt, Darmstadt, Germany
S. Gorgi Zadeh, U. van Rienen
Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock, Germany
A.A. Sulimov
DESY, Hamburg, Germany

Funding: The authors would like to acknowledge the support by the DFG (German Research Foundation) in the framework of the Scientific Network SCHM 3127/1,2 that provided the basis for this collaborative work.
The fundamental mode spectrum of superconducting cavities is sensitive to small geometry deformations introduced by the manufacturing process. In this work we consider variations in the equatorial and iris radii of the 1.3 GHz TESLA cavities used at the European XFEL. The cavities with slightly perturbed geometry are simulated using a finite element based eigenvalue solver. Employing uncertainty quantification methods such as sparse-grids, statistical information about the fundamental mode spectrum can be efficiently calculated. Moreover, using global sensitivity analysis, in particular Sobol indices, the impact of the individual geometry parameters on the quantities of interest, i.e. resonance frequencies, field-flatness and the cell-to-cell coupling coefficient, can be computed. We will explain important aspects of the uncertainty quantification methodology and give numerical results for illustration.

Slides TUPAG08 [0.672 MB]

Export •

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

TUPAG14

Constrained Multi-Objective Shape Optimization of Superconducting RF Cavities to Counteract Dangerous Higher Order Modes

293

M. Kranjcevic, P. Arbenz
ETH, Zurich, Switzerland
A. Adelmann
PSI, Villigen PSI, Switzerland
S. Gorgi Zadeh, U. van Rienen
Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock, Germany

High current storage rings, such as the Z operating mode of the FCC-ee, require superconducting radio frequency (RF) cavities that are optimized with respect to both the fundamental mode and the dangerous higher order modes. In order to optimize the shape of the RF cavity, a constrained multi-objective optimization problem is solved using a massively parallel implementation of an evolutionary algorithm. Additionally, a frequency-fixing scheme is employed to deal with the constraint on the frequency of the fundamental mode. Finally, the computed Pareto front approximation and an RF cavity shape with desired properties are shown.

Slides TUPAG14 [3.001 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-ICAP2018-TUPAG14

About •

paper received ※ 19 October 2018 paper accepted ※ 10 December 2018 issue date ※ 26 January 2019

Export •

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

WEPAF04

Longitudinal Beam Dynamics in FRIB and ReA Linacs

330

A.S. Plastun, P.N. Ostroumov, A.C.C. Villari, Q. Zhao
FRIB, East Lansing, USA

The Front-End and first three cryomodules of the Facility for Rare Isotope Beam (FRIB) at Michigan State University (MSU) commissioned in July, 2018. The paper describes the online tuning procedures of the longitudinal beam dynamics through the FRIB linac. These procedures include tuning of the accelerating field phases and amplitudes in the cavities. We developed an automated simulation-based tuning procedure for the multi-harmonic buncher. In order to tune the radio-frequency quadrupole (RFQ) we measured and calculated its threshold voltage and scanned its longitudinal acceptance. Tuning of the rebunchers and superconducting accelerating cavities is per-formed by means of the phase scans and Time-Of-Flight (TOF) beam energy measurements with beam position and phase monitors. While FRIB is being commissioned, the re-accelerator (ReA3) for rare isotope beams (RIBs) is being upgraded. We redesigned the ReA3 RFQ to improve its cooling system and provide reliable operation with 16.1 MHz prebunched ion beams with A/Q = 5. In order to provide matching of any ReA3 beam both to the following upgrade cryomodules and physics experiments’ requirements, room temperature rebuncher/debuncher is being designed. The design procedure includes the beam dynamics, electromagnetic, thermal and mechanical simulations and optimizations.

Slides WEPAF04 [2.406 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-ICAP2018-WEPAF04

About •

paper received ※ 19 October 2018 paper accepted ※ 28 January 2019 issue date ※ 26 January 2019

Export •

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