17–22 May 2026
C.I.D
Europe/Zurich timezone

Machine Learning-Based Reconstruction of Longitudinal Phase Space for Electron Storage Rings

MOP6602
18 May 2026, 16:00
2h
C.I.D

C.I.D

Deauville, France
Poster Presentation MC6.D13: Instrumentation: Artificial Intelligence Poster session

Speaker

Dechong Zhu (Institute of High Energy Physics)

Description

This study addresses the challenge of directly measuring the longitudinal phase space in electron storage rings and proposes a novel reconstruction method based on machine learning. Precise reconstruction of the longitudinal phase space distribution is crucial for optimizing beam performance and suppressing instabilities. However, traditional methods used in proton accelerators are unsuitable for electron storage rings due to detector bandwidth limitations, which prevent their application to beams with picosecond-scale lengths.

To overcome this, the proposed method first establishes a mapping dataset through theoretical simulations, linking machine parameters and initial longitudinal phase space distributions to observed transient images during injection. Machine learning models are then trained on this dataset. Ultimately, the trained model can directly reconstruct the initial longitudinal phase space distribution and its key physical parameters from experimentally observed transient injection images. This approach provides a new diagnostic technique for the longitudinal phase space in electron storage rings, with the potential to support the fine-tuning and performance enhancement of next-generation accelerators.

In which format do you inted to submit your paper? Word

Author

Dechong Zhu (Institute of High Energy Physics)

Co-authors

Wan Zhang (Chinese Academy of Sciences) Yanfeng Sui (Institute of High Energy Physics) Taoguang Xu (Institute of High Energy Physics) Jun He (Institute of High Energy Physics) Junhui Yue (Institute of High Energy Physics, Chinese Academy of Sciences) Jianshe Cao (Institute of High Energy Physics)

Presentation materials

There are no materials yet.