1–6 Jun 2025
Taipei International Convention Center (TICC)
Asia/Taipei timezone

Prediction of electron beam parameters through diffraction images

THPM051
5 Jun 2025, 15:30
2h
Exhibiton Hall A _Magpie (TWTC)

Exhibiton Hall A _Magpie

TWTC

Poster Presentation MC6.T03 Beam Diagnostics and Instrumentation Thursday Poster Session

Speaker

Jiapeng Li (Huazhong University of Science and Technology)

Description

Achieving precise and real-time diagnostics of electron beam characteristics is critical for enhancing the performance of ultrafast electron diffraction (UED) and electron microscopy (UEM) techniques. Key parameters such as bunch size, emittance, energy spread, and spatial pointing jitter directly influence the quality and accuracy of experimental results. Traditional diagnostic methods often lack the ability to provide continuous, real-time, and non-intrusive monitoring, limiting their effectiveness. This work presents a machine learning (ML)-based approach that utilizes a small dataset of known beam parameters in combination with real-time diffraction image data recorded during experiments to predict electron beam characteristics for each run. This approach enables continuous optimization of beam stability without interfering with the experiment and facilitates real-time updates to UED parameters during data collection. As a result, it significantly improves the precision, reliability, and overall performance of UED and UEM experiments.

Region represented Asia
Paper preparation format Word

Author

Jiapeng Li (Huazhong University of Science and Technology)

Co-authors

Bingxi Liu (Huazhong University of Science and Technology) Cheng-Ying Tsai (Huazhong University of Science and Technology) Kuanjun Fan (Huazhong University of Science and Technology) Qiao Luo (Huazhong University of Science and Technology) Yaodan Hu (Huazhong University of Science and Technology) ZHENGYAN LI (Huazhong University of Science and Technology) Zhengzheng Liu (Huazhong University of Science and Technology)

Presentation materials

There are no materials yet.