Machine-learning approach for operating electron beam at KEK $e^-/e^+$ injector Linac
Authors:
Gaku Mitsuka,
Shinnosuke Kato,
Naoko Iida,
Takuya Natsui,
Masanori Satoh
Abstract:
In current accelerators, numerous parameters and monitored values are to be adjusted and evaluated, respectively. In addition, fine adjustments are required to achieve the target performance. Therefore, the conventional accelerator-operation method, in which experts manually adjust the parameters, is reaching its limits. We are currently investigating the use of machine learning for accelerator tu…
▽ More
In current accelerators, numerous parameters and monitored values are to be adjusted and evaluated, respectively. In addition, fine adjustments are required to achieve the target performance. Therefore, the conventional accelerator-operation method, in which experts manually adjust the parameters, is reaching its limits. We are currently investigating the use of machine learning for accelerator tuning as an alternative to expert-based tuning. In recent years, machine-learning algorithms have progressed significantly in terms of speed, sensitivity, and application range. In addition, various libraries are available from different vendors and are relatively easy to use. Herein, we report the results of electron-beam tuning experiments using Bayesian optimization, a tree-structured Parzen estimator, and a covariance matrix-adaptation evolution strategy. Beam-tuning experiments are performed at the KEK $e^-$/$e^+$ injector Linac to maximize the electron-beam charge and reduce the energy-dispersion function. In each case, the performance achieved is comparable to that of a skilled expert.
△ Less
Submitted 26 January, 2024;
originally announced January 2024.
Achievement of 200,000 hours of operation at KEK 7-GeV electron 4-GeV positron injector linac
Authors:
Kazuro Furukawa,
Mitsuo Akemoto,
Dai Arakawa,
Yoshio Arakida,
Yusei Bando,
Hiroyasu Ego,
Yoshinori Enomoto,
Toshiyasu Higo,
Hiroyuki Honma,
Naoko Iida,
Kazuhisa Kakihara,
Takuya Kamitani,
Hiroaki Katagiri,
Masato Kawamura,
Shuji Matsumoto,
Toshihiro Matsumoto,
Hideki Matsushita,
Katsuhiko Mikawa,
Takako Miura,
Fusashi Miyahara,
Hiromitsu Nakajima,
Takuya Natsui,
Yujiro Ogawa,
Satoshi Ohsawa,
Yuichi Okayasu
, et al. (17 additional authors not shown)
Abstract:
KEK electron positron injector LINAC initiated the injection operation into Photon Factory (PF) light source in 1982. Since then for 39 years, it has served for multiple projects, namely, TRISTAN, PF-AR, KEKB, and SuperKEKB. Its total operation time has accumulated 200 thousand hours on May 7, 2020. We are extremely proud of the achievement following continuous efforts by our seniors. The construc…
▽ More
KEK electron positron injector LINAC initiated the injection operation into Photon Factory (PF) light source in 1982. Since then for 39 years, it has served for multiple projects, namely, TRISTAN, PF-AR, KEKB, and SuperKEKB. Its total operation time has accumulated 200 thousand hours on May 7, 2020. We are extremely proud of the achievement following continuous efforts by our seniors. The construction of the injector LINAC started in 1978, and it was commissioned for PF with 2.5 GeV electron in 1982. In parallel, the positron generator linac was constructed for the TRISTAN collider project. The slow positron facility was also commissioned in 1992. After the KEKB asymmetric-energy collider project was commissioned in 1998 with direct energy injections, the techniques such as two-bunch acceleration and simultaneous injection were developed. As the soft structure design of the LINAC was too weak against the great east Japan earthquake, it took three years to recover. Then the construction and commissioning for the SuperKEKB project went on, and the simultaneous top-up injection into four storage rings contributes to the both elementary particle physics and photon science.
△ Less
Submitted 4 July, 2022;
originally announced July 2022.
