Genetic algorithm enhanced by machine learning in dynamic aperture optimization
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 National Synchrotron Light Source II (NSLS-II) Storage Ring. During the evolution processes employed by the genetic al...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
American Physical Society
2018-05-01
|
Series: | Physical Review Accelerators and Beams |
Online Access: | http://doi.org/10.1103/PhysRevAccelBeams.21.054601 |
id |
doaj-6bbfdd3c2f9d43419391912dbe51f40f |
---|---|
record_format |
Article |
spelling |
doaj-6bbfdd3c2f9d43419391912dbe51f40f2020-11-24T23:57:09ZengAmerican Physical SocietyPhysical Review Accelerators and Beams2469-98882018-05-0121505460110.1103/PhysRevAccelBeams.21.054601Genetic algorithm enhanced by machine learning in dynamic aperture optimizationYongjun LiWeixing ChengLi Hua YuRobert RainerWith 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 National Synchrotron Light Source II (NSLS-II) Storage Ring. During the evolution processes employed by the genetic algorithm, the population is classified into different clusters in the search space. The clusters with top average fitness are given “elite” status. Intervention on the population is implemented by repopulating some potentially competitive candidates based on the experience learned from the accumulated data. These candidates replace randomly selected candidates among the original data pool. The average fitness of the population is therefore improved while diversity is not lost. Maintaining diversity ensures that the optimization is global rather than local. 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.http://doi.org/10.1103/PhysRevAccelBeams.21.054601 |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Yongjun Li Weixing Cheng Li Hua Yu Robert Rainer |
spellingShingle |
Yongjun Li Weixing Cheng Li Hua Yu Robert Rainer Genetic algorithm enhanced by machine learning in dynamic aperture optimization Physical Review Accelerators and Beams |
author_facet |
Yongjun Li Weixing Cheng Li Hua Yu Robert Rainer |
author_sort |
Yongjun Li |
title |
Genetic algorithm enhanced by machine learning in dynamic aperture optimization |
title_short |
Genetic algorithm enhanced by machine learning in dynamic aperture optimization |
title_full |
Genetic algorithm enhanced by machine learning in dynamic aperture optimization |
title_fullStr |
Genetic algorithm enhanced by machine learning in dynamic aperture optimization |
title_full_unstemmed |
Genetic algorithm enhanced by machine learning in dynamic aperture optimization |
title_sort |
genetic algorithm enhanced by machine learning in dynamic aperture optimization |
publisher |
American Physical Society |
series |
Physical Review Accelerators and Beams |
issn |
2469-9888 |
publishDate |
2018-05-01 |
description |
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 National Synchrotron Light Source II (NSLS-II) Storage Ring. During the evolution processes employed by the genetic algorithm, the population is classified into different clusters in the search space. The clusters with top average fitness are given “elite” status. Intervention on the population is implemented by repopulating some potentially competitive candidates based on the experience learned from the accumulated data. These candidates replace randomly selected candidates among the original data pool. The average fitness of the population is therefore improved while diversity is not lost. Maintaining diversity ensures that the optimization is global rather than local. 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. |
url |
http://doi.org/10.1103/PhysRevAccelBeams.21.054601 |
work_keys_str_mv |
AT yongjunli geneticalgorithmenhancedbymachinelearningindynamicapertureoptimization AT weixingcheng geneticalgorithmenhancedbymachinelearningindynamicapertureoptimization AT lihuayu geneticalgorithmenhancedbymachinelearningindynamicapertureoptimization AT robertrainer geneticalgorithmenhancedbymachinelearningindynamicapertureoptimization |
_version_ |
1725455366874988544 |