基于机器学习的电子储存环注入品质评估及异常检测

方子堃; 蒋天宇; 周逸媚; 冷用斌

doi:10.3724/j.0253-3219.2026.hjs.49.250102

您当前的位置：

首页 >

文章列表页 >

基于机器学习的电子储存环注入品质评估及异常检测

加速器技术、射线技术及应用 | 更新时间：2026-01-19

- 基于机器学习的电子储存环注入品质评估及异常检测
- Quality assessment and anomaly detection for electronic storage ring injection based on machine learning
- 核技术 2026年49卷第1期文章编号：010201
- 作者机构：
  
  1.中国科学技术大学国家同步辐射实验室合肥 230029
  2.中国科学院上海高等研究院上海 201210
- 作者简介：
  
  方子堃，男，2001年出生，2023年毕业于兰州大学，现为硕士研究生，研究领域为机器学习在束流诊断应用
  冷用斌，E-mail：lengyb@ustc.edu.cn
- 基金信息：
  
  中国科学院大科学装置维修改造项目“加速器束流测控通用信号处理平台”项目资助
- DOI：10.3724/j.0253-3219.2026.hjs.49.250102
  中图分类号： TL99
- 收稿：2025-03-10，
  
  修回：2025-06-21，
  
  纸质出版：2026-01-15
- 稿件说明：
移动端阅览
方子堃,蒋天宇,周逸媚等.基于机器学习的电子储存环注入品质评估及异常检测[J].核技术,2026,49(01):010201.

FANG Zikun,JIANG Tianyu,ZHOU Yimei,et al.Quality assessment and anomaly detection for electronic storage ring injection based on machine learning[J].NUCLEAR TECHNIQUES,2026,49(01):010201.
方子堃,蒋天宇,周逸媚等.基于机器学习的电子储存环注入品质评估及异常检测[J].核技术,2026,49(01):010201. DOI： 10.3724/j.0253-3219.2026.hjs.49.250102. CSTR： 32193.14.hjs.CN31-1342/TL.2026.49.250102.

FANG Zikun,JIANG Tianyu,ZHOU Yimei,et al.Quality assessment and anomaly detection for electronic storage ring injection based on machine learning[J].NUCLEAR TECHNIQUES,2026,49(01):010201. DOI： 10.3724/j.0253-3219.2026.hjs.49.250102. CSTR： 32193.14.hjs.CN31-1342/TL.2026.49.250102.

摘要

电子储存环注入过程中束团参数变化剧烈，通过逐束团三维位置测量技术可实时捕捉并分析这个过程，获得横向Beta振荡和纵向同步振荡相关的多个束流动力学参数，如何利用这些测量结果来评价并指导优化加速器的运行，是同步辐射光源装置需要研究的重要问题。本文提出了一种基于机器学习的电子储存环注入品质评估及异常检测方法。选择上海光源多年累积的注入瞬态过程数据作为样本库，对补注束团的动力学参数进行聚类分析，找到装置运行状态时间维度上的

演化规律并标记出不属于主群的异常数据样本。训练

https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=98524820&type=

https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=98524805&type=

1.60866666

2.28600001

-NN模型对异常数据进行了预测及校验，实验结果表明预测模型精确率好于90%，可作为装置运行性能在线自动评价的有效方法。

Abstract

Background

During electron storage ring injection

beam parameters change drastically. Bunch-by-bunch three-dimensional position measurements can capture this transient process in real time to obtain multiple beam dynamics parameters. However

how to effectively utilize these measurements to evaluate and optimize accelerator operation remains a critical challenge for synchrotron radiation facilities.

Purpose

This study aims to develop a machine learning-based method for injection quality assessment and anomaly detection in electron storage rings.

Methods

Firstly

injection transient data accumulated over multiple years at the Shanghai Synchrotron Radiation Facility (SSRF) from September 2021 to June 2024 were selected as the sample database

including longitudinal oscillation amplitude

synchrotron damping time

bunch charge

and other dynamic parameters of refilled bunches. Then

the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm was applied to perform cluster analysis on the longitudinal oscillation amplitude and synchrotron damping time

identifying the temporal evolution patterns of facility operation states and labeling anomalous data samples that deviated from the main clusters. Finally

-Nearest Neighbor (

-NN)

classification model was trained using the labeled dataset to predict anomalous injection events

with the optimal

value determined through comparative analysis.

Results

The clustering results reveal the evolution patterns of SSRF operating states over time

with Cluster 2 (containing data from 2023‒2024) demonstrating significantly improved matching between the storage ring and injector compared to earlier periods. The average longitudinal oscillation amplitude decreases from 101.21 ps (before September 2023) to 64.37 ps (after September 2023)

while the synchrotron damping time reduces from 3.01 ms to 1.70 ms. The optimized

-NN model with

=4 achieves a precision of 90%

recall of 66%

-score of 0.78

and PR-AUC of 0.84 in anomaly detection

outperforming other machine learning algorithms including Support Vector Machine

Decision Tree

Random Forest

Logistic Regression

and Naive Bayes.

Conclusions

The method proposed in this study combining unsupervised learning and supervised learning effectively detects anomalous injection events with 90% precision

enabling real-time quality assessment of storage ring injection performance. The longitudinal oscillation amplitude and synchrotron damping time serve as reliable indicators of injection quality

while the machine learning approach provides early warning capabilities for facility operators to prevent performance degradation before observable operational failures occur.

关键词

Keywords

references

邓又铭 , 冷用斌 , 周逸媚 , 等 . 电子储存环逐束团诊断技术研究进展 [J ] . 核技术 , 2024 , 47 ( 10 ): 100201 . DOI: 10.11889/j.0253-3219.2024.hjs.47.100201. CSTR: 32193.14.hjs.CN31-1342/TL.2024.47.100201 http://dx.doi.org/10.11889/j.0253-3219.2024.hjs.47.100201.CSTR:32193.14.hjs.CN31-1342/TL.2024.47.100201 .

DENG Youming , Leng Yongbin , Zhou Yimei , et al . Current status and latest developments in electron storage ring bunch-by-bunch beam diagnostic techniques [J ] . Nuclear Techniques , 2024 , 47 ( 10 ): 100201 . DOI: 10.11889/j.0253-3219.2024.hjs.47.100201. CSTR: 32193.14.hjs.CN31-1342/TL.2024.47.100201 http://dx.doi.org/10.11889/j.0253-3219.2024.hjs.47.100201.CSTR:32193.14.hjs.CN31-1342/TL.2024.47.100201 .

张满洲 , 田顺强 , 王坤 , 等 . 上海光源储存环新注入方式 [J ] . 核技术 , 2017 , 40 ( 11 ): 110101 . DOI: 10.11889/j.0253-3219.2017.hjs.40.110101 http://dx.doi.org/10.11889/j.0253-3219.2017.hjs.40.110101 .

ZHANG Manzhou , TIAN Shunqiang , WANG Kun , et al . Study of new injection schemes for the SSRF storage ring [J ] . Nuclear Techniques , 2017 , 40 ( 11 ): 110101 . DOI: 10.11889/j.0253-3219.2017.hjs.40.110101 http://dx.doi.org/10.11889/j.0253-3219.2017.hjs.40.110101 .

Leng Y , Yang Y , Zhang N , et al . Bunch by bunch transverse beam position observation and analyze during injection at ssrf [J ] . Beijing: 13th International Beam Instrumentation Conference , 2013 : 746 – 748 .

Deng Y M , Leng Y B , Xu X Y , et al . Ultrahigh spatiotemporal resolution beam signal reconstruction with bunch phase compensation [J ] . Nuclear Science and Techniques , 2024 , 35 ( 5 ): 89 . DOI: 10.1007/s41365-024-01444-y http://dx.doi.org/10.1007/s41365-024-01444-y .

许兴懿 . 基于机器学习的逐束团信息提取及应用技术研究 [D ] . 上海 : 中国科学院上海应用物理研究所 , 2022 . DOI: 10.27585/d.cnki.gkshs.2022.000009 http://dx.doi.org/10.27585/d.cnki.gkshs.2022.000009 .

XU Xingyi . Research on information extraction and application technology of beam-by-beam cluster based on machine learning [D ] . Shanghai : Shanghai Institute of Applied Physics, Chinese Academy of Sciences , 2022 . DOI: 10.27585/d.cnki.gkshs.2022.000009 http://dx.doi.org/10.27585/d.cnki.gkshs.2022.000009 .

Xu X Y , Leng Y B , Gao B , et al . HOTCAP: a new software package for high-speed oscilloscope-based three-dimensional bunch charge and position measurement [J ] . Nuclear Science and Techniques , 2021 , 32 ( 11 ): 131 . DOI: 10.1007/s41365-021-00966-z http://dx.doi.org/10.1007/s41365-021-00966-z .

杨星 , 冷用斌 , 周逸媚 . 逐束团三维信息提取软件HOTCAP的实时性能优化 [J ] . 核技术 , 2024 , 47 ( 2 ): 020102 . DOI: 10.11889/j.0253-3219.2024.hjs.47.020102 http://dx.doi.org/10.11889/j.0253-3219.2024.hjs.47.020102 .

YANG Xing , LENG Yongbin , ZHOU Yimei . Real-time optimization of bunch-by-bunch 3D information extraction software HOTCAP [J ] . Nuclear Techniques , 2024 , 47 ( 2 ): 020102 . DOI: 10.11889/j.0253-3219.2024.hjs.47.020102 http://dx.doi.org/10.11889/j.0253-3219.2024.hjs.47.020102 .

Wang H S , Yang X , Leng Y B , et al . Bunch‑length measurement at a bunch‑by‑bunch rate based on time-frequency‑domain joint analysis techniques and its application [J ] . Nuclear Science and Techniques , 2024 , 35 ( 4 ): 80 . DOI: 10.1007/s41365-024-01443-z http://dx.doi.org/10.1007/s41365-024-01443-z .

Coyle L , Blanc F , Buffat X , et al . Detection and classification of collective beam behaviour in the LHC [C ] . 12th International Particle Accelerator Conference (IPAC'21 ), 2021 : 4318 – 4321 . DOI: 10.18429/JACoW-IPAC2021THPAB260 http://dx.doi.org/10.18429/JACoW-IPAC2021THPAB260 .

Wang G , Cheng W , Yang X , et al . Storage ring injection kickers alignment optimization in NSLS-II [C ] . 8th International Particle Accelerator Conference (IPAC' 17 ), 2017.

Lee S Y . Accelerator physics [M ] . Singapore : World Scientific Publishing Company , 2018 .

Nodes T , Gallagher N . Median filters: some modifications and their properties [J ] . IEEE Transactions on Acoustics, Speech, and Signal Processing , 1982 , 30 ( 5 ): 739 – 746 . DOI: 10.1109/TASSP.1982.1163951 http://dx.doi.org/10.1109/TASSP.1982.1163951 .

王骏 , 王士同 , 邓赵红 . 聚类分析研究中的若干问题 [J ] . 控制与决策 , 2012 , 27 ( 3 ): 321 – 328 . DOI: 10.13195/j.cd.2012.03.4.wangj.013 http://dx.doi.org/10.13195/j.cd.2012.03.4.wangj.013 .

WANG Jun , WANG Shitong , DENG Zhaohong . Survey on challenges in clustering analysis research [J ] . Control and Decision , 2012 , 27 ( 3 ): 321 – 328 . DOI: 10.13195/j.cd.2012.03.4.wangj.013 http://dx.doi.org/10.13195/j.cd.2012.03.4.wangj.013 .

Jain A K , Murty M N , Flynn P J . Data clustering [J ] . ACM Computing Surveys , 1999 , 31 ( 3 ): 264 – 323 . DOI: 10.1145/331499.331504 http://dx.doi.org/10.1145/331499.331504 .

孙吉贵 , 刘杰 , 赵连宇 . 聚类算法研究 [J ] . 软件学报 , 2008 , 19 ( 1 ): 48 – 61 . DOI: 10.3724/SP.J.1001.2008.00048 http://dx.doi.org/10.3724/SP.J.1001.2008.00048 .

SUN Jigui , LIU Jie , ZHAO Lianyu . Clustering algorithms research [J ] . Journal of Software , 2008 , 19 ( 1 ): 48 – 61 . DOI: 10.3724/SP.J.1001.2008.00048 http://dx.doi.org/10.3724/SP.J.1001.2008.00048 .

Macqueen J B . Some methods for classification and analysis of multivariate observations [C ] . Oakland, CA, USA : Proceedings of the fifth Berkeley symposium on mathematical statistics and probability , 1967 , 1 : 281 – 297 .

周志华 . 机器学习 [M ] . 北京 : 清华大学出版社 , 2016 .

ZHOU Zhihua . Machine learning [M ] . Beijing : Tsinghua University Press , 2016 .

Cover T , Hart P . Nearest neighbor pattern classification [J ] . IEEE Transactions on Information Theory , 1967 , 13 ( 1 ): 21 – 27 . DOI: 10.1109/TIT.1967.1053964 http://dx.doi.org/10.1109/TIT.1967.1053964 .

周逸媚 . 储存环逐束团三维位置测量技术研究 [D ] . 上海 : 中国科学院上海应用物理研究所 , 2020 . DOI: 10.27585/d.cnki.gkshs.2020.000011 http://dx.doi.org/10.27585/d.cnki.gkshs.2020.000011 .

ZHOU Yimei . Research on three-dimensional position measurement technology of storage ring beam by beam [D ] . Shanghai : Shanghai Institute of Applied Physics, Chinese Academy of Sciences , 2020 . DOI: 10.27585/d.cnki.gkshs.2020.000011 http://dx.doi.org/10.27585/d.cnki.gkshs.2020.000011 .

赵存秀 . 基于混淆矩阵的分类器性能评价指标比较 [J ] . 电子技术与软件工程 , 2020 ( 13 ): 146 – 147 . DOI: 10.20109/j.cnki.etse.2020.13.066 http://dx.doi.org/10.20109/j.cnki.etse.2020.13.066 .

ZHAO Cunxiu . Comparison of performance evaluation indexes of classifiers based on confusion matrix [J ] . Electronic Technology & Software Engineering , 2020 ( 13 ): 146 – 147 . DOI: 10.20109/j.cnki.etse.2020.13.066 http://dx.doi.org/10.20109/j.cnki.etse.2020.13.066 .

Kieu T , Yang B , Jensen C S . Outlier detection for multidimensional time series using deep neural networks [C ] . Aalborg, Denmark : 2018 19th IEEE International Conference on Mobile Data Management (MDM), IEEE , 2018 : 125 – 134 . DOI: 10.1109/MDM.2018.00029 http://dx.doi.org/10.1109/MDM.2018.00029 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于SVR-PSO的核安全级DCS关键芯片布局优化研究

基于机器学习预测膨润土中阳离子的分配系数

机器学习方法研究重离子熔合反应截面

深度学习在高能核物理中的前沿进展

基于连续型贝叶斯概率估计器预测原子核质量