16–21 Aug 2026
Daejeon Convention Center
Asia/Seoul timezone

Physics-Informed Multi-Objective Genetic Optimization for High-Intensity RFQ Design

Not scheduled
2h
Daejeon Convention Center

Daejeon Convention Center

107 Expo-ro, Yuseong-gu, Daejeon (34125) South Korea
Poster Presentation MC1.A01: Beam Dynamics, beam simulations, beam transport Poster Session

Speaker

Yulin Ge (Sun Yat-sen University)

Description

RFQ design involves strong parameter coupling, multi-physics interactions, and multiple optimization constraints. To address these challenges, this work develops a physics-informed genetic optimization framework for RFQ beam dynamics and electromagnetic design.
By incorporating evolutionary rules associated with modulation factor, synchronous phase, and focusing strength, the framework improves optimization efficiency and physical reliability. The method has been validated for several RFQ configurations, including heavy-ion, proton, compact CW, and He²⁺ RFQs. The optimized LEAF, ADS Injector-II, PAFA, and SYSU-IFCEN HeRFQs achieved transmission efficiencies of 98.7%, 99.8%, 99.8%, and 97.4%, with RFQ lengths of 575.85 cm, 409.56 cm, 351.98 cm, and 143.42 cm, respectively. Typical optimization tasks required 80–300 core-hours using 20–70 generations with 100 individuals.
The framework was also applied to RFQ electromagnetic design using surrogate-model-assisted optimization. For a C⁴⁺ RFQ cavity, the optimized design achieved an approximately 10% increase in the Q value.
These results demonstrate an efficient approach for automated RFQ optimization and accelerator design.

Funding Agency

This work was supported by the National Natural Science Foundation of China (Grant Nos. 12505173 and 12575170)

I have read and accept the Privacy Policy Statement Yes

Author

Yulin Ge (Sun Yat-sen University)

Co-authors

Liang Lu (Sun Yat-sen University) Wei Ma (Sun Yat-sen University) Zeyang Zhang (Sun Yat-sen University)

Presentation materials

There are no materials yet.