Adapted Swin Transformer-based real-time plasma shape detection and control in HL-3
In the field of magnetic confinement plasma control, the accurate feedback of plasma position and shape primarily relies on calculations derived from magnetic measurements through equilibrium reconstruction or matrix mapping method. However, under harsh conditions like high-energy neutron radiation...
Saved in:
Main Authors: | , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IOP Publishing
2025-01-01
|
Series: | Nuclear Fusion |
Subjects: | |
Online Access: | https://doi.org/10.1088/1741-4326/ada2fe |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
_version_ | 1841556234517348352 |
---|---|
author | Qianyun Dong Zhengwei Chen Rongpeng Li Zongyu Yang Feng Gao Yihang Chen Fan Xia Wulyu Zhong Zhifeng Zhao |
author_facet | Qianyun Dong Zhengwei Chen Rongpeng Li Zongyu Yang Feng Gao Yihang Chen Fan Xia Wulyu Zhong Zhifeng Zhao |
author_sort | Qianyun Dong |
collection | DOAJ |
description | In the field of magnetic confinement plasma control, the accurate feedback of plasma position and shape primarily relies on calculations derived from magnetic measurements through equilibrium reconstruction or matrix mapping method. However, under harsh conditions like high-energy neutron radiation and elevated temperatures, the installation of magnetic probes within the device becomes challenging. Relying solely on external magnetic probes can compromise the precision of EFIT in determining the plasma shape. To tackle this issue, we introduce a real-time, non-magnetic measurement method on the HL-3 tokamak, which diagnoses the plasma position and shape via imaging. Particularly, we put forward an adapted Swin Transformer model, the Poolformer Swin Transformer (PST), to accurately and fastly interpret the plasma shape from the Charge-Coupled Device Camera images. By adopting multi-task learning and knowledge distillation techniques, the model is capable of robustly detecting six shape parameters under visual interference conditions such as bright light from the divertor and gas injection, thereby avoiding cumbersome manual labeling. Specifically, the well-trained PST model capably infers R and Z within the mean average error below 1.1 cm and 1.8 cm, respectively, while requiring less than 2 ms for end-to-end feedback, an 80% improvement over the smallest Swin Transformer model, laying the foundation for real-time control. Finally, we deploy the PST model in the Plasma Control System using TensorRT, and achieve 500 ms stable PID feedback control based on the PST-computed horizontal displacement information. In conclusion, this research opens up new avenues for the practical application of image-computing plasma shape diagnostic methods in the realm of real-time feedback control. |
format | Article |
id | doaj-art-873fb7339124483da94d82496b6c4f74 |
institution | Kabale University |
issn | 0029-5515 |
language | English |
publishDate | 2025-01-01 |
publisher | IOP Publishing |
record_format | Article |
series | Nuclear Fusion |
spelling | doaj-art-873fb7339124483da94d82496b6c4f742025-01-07T12:20:17ZengIOP PublishingNuclear Fusion0029-55152025-01-0165202603110.1088/1741-4326/ada2feAdapted Swin Transformer-based real-time plasma shape detection and control in HL-3Qianyun Dong0https://orcid.org/0009-0007-6020-6122Zhengwei Chen1Rongpeng Li2https://orcid.org/0000-0003-4297-5060Zongyu Yang3https://orcid.org/0009-0000-4083-1552Feng Gao4Yihang Chen5Fan Xia6Wulyu Zhong7https://orcid.org/0000-0001-8217-9400Zhifeng Zhao8https://orcid.org/0000-0002-5479-7890Zhejiang University , Hangzhou 310058, ChinaSouthwestern Institute of Physics , Chengdu 610043, ChinaZhejiang University , Hangzhou 310058, ChinaSouthwestern Institute of Physics , Chengdu 610043, ChinaZhejiang Lab , Hangzhou 311500, ChinaSouthwestern Institute of Physics , Chengdu 610043, ChinaSouthwestern Institute of Physics , Chengdu 610043, ChinaSouthwestern Institute of Physics , Chengdu 610043, ChinaZhejiang Lab , Hangzhou 311500, ChinaIn the field of magnetic confinement plasma control, the accurate feedback of plasma position and shape primarily relies on calculations derived from magnetic measurements through equilibrium reconstruction or matrix mapping method. However, under harsh conditions like high-energy neutron radiation and elevated temperatures, the installation of magnetic probes within the device becomes challenging. Relying solely on external magnetic probes can compromise the precision of EFIT in determining the plasma shape. To tackle this issue, we introduce a real-time, non-magnetic measurement method on the HL-3 tokamak, which diagnoses the plasma position and shape via imaging. Particularly, we put forward an adapted Swin Transformer model, the Poolformer Swin Transformer (PST), to accurately and fastly interpret the plasma shape from the Charge-Coupled Device Camera images. By adopting multi-task learning and knowledge distillation techniques, the model is capable of robustly detecting six shape parameters under visual interference conditions such as bright light from the divertor and gas injection, thereby avoiding cumbersome manual labeling. Specifically, the well-trained PST model capably infers R and Z within the mean average error below 1.1 cm and 1.8 cm, respectively, while requiring less than 2 ms for end-to-end feedback, an 80% improvement over the smallest Swin Transformer model, laying the foundation for real-time control. Finally, we deploy the PST model in the Plasma Control System using TensorRT, and achieve 500 ms stable PID feedback control based on the PST-computed horizontal displacement information. In conclusion, this research opens up new avenues for the practical application of image-computing plasma shape diagnostic methods in the realm of real-time feedback control.https://doi.org/10.1088/1741-4326/ada2feshape detection and controlreal-timeSwin TransformerHL-3 tokamak |
spellingShingle | Qianyun Dong Zhengwei Chen Rongpeng Li Zongyu Yang Feng Gao Yihang Chen Fan Xia Wulyu Zhong Zhifeng Zhao Adapted Swin Transformer-based real-time plasma shape detection and control in HL-3 Nuclear Fusion shape detection and control real-time Swin Transformer HL-3 tokamak |
title | Adapted Swin Transformer-based real-time plasma shape detection and control in HL-3 |
title_full | Adapted Swin Transformer-based real-time plasma shape detection and control in HL-3 |
title_fullStr | Adapted Swin Transformer-based real-time plasma shape detection and control in HL-3 |
title_full_unstemmed | Adapted Swin Transformer-based real-time plasma shape detection and control in HL-3 |
title_short | Adapted Swin Transformer-based real-time plasma shape detection and control in HL-3 |
title_sort | adapted swin transformer based real time plasma shape detection and control in hl 3 |
topic | shape detection and control real-time Swin Transformer HL-3 tokamak |
url | https://doi.org/10.1088/1741-4326/ada2fe |
work_keys_str_mv | AT qianyundong adaptedswintransformerbasedrealtimeplasmashapedetectionandcontrolinhl3 AT zhengweichen adaptedswintransformerbasedrealtimeplasmashapedetectionandcontrolinhl3 AT rongpengli adaptedswintransformerbasedrealtimeplasmashapedetectionandcontrolinhl3 AT zongyuyang adaptedswintransformerbasedrealtimeplasmashapedetectionandcontrolinhl3 AT fenggao adaptedswintransformerbasedrealtimeplasmashapedetectionandcontrolinhl3 AT yihangchen adaptedswintransformerbasedrealtimeplasmashapedetectionandcontrolinhl3 AT fanxia adaptedswintransformerbasedrealtimeplasmashapedetectionandcontrolinhl3 AT wulyuzhong adaptedswintransformerbasedrealtimeplasmashapedetectionandcontrolinhl3 AT zhifengzhao adaptedswintransformerbasedrealtimeplasmashapedetectionandcontrolinhl3 |