Predicting Performance of Hall Effect Ion Source Using Machine Learning
Hall Effect Ion Source In article number 2400555, Wonho Choe and co-workers introduce a machine learning-based approach to accurately predicting the performance of Hall thrusters, a critical technology for space propulsion and industrial ion beam sources. By utilizing an ensemble of neural networks trained on 18,000 simulation datasets validated by experiments, the model achieves high accuracy in thrust and discharge current predictions, enabling the rapid development of optimized, high-efficiency thrusters with shorter design cycles.
Authors
- Changmin Shin (ORCID: https://orcid.org/0000-0003-3015-277X)
- Wonho Choe (ORCID: https://orcid.org/0000-0002-8952-8252)
- Sanghoo Park (ORCID: https://orcid.org/0000-0002-4180-7455)
- Jaehong Park (ORCID: https://orcid.org/0000-0001-5722-0344)
- Young-chul Ghim (ORCID: https://orcid.org/0000-0003-4123-9416)
- Guentae Doh (ORCID: https://orcid.org/0000-0001-5908-2205)
- Su‐Jin Shin (ORCID: https://orcid.org/0000-0003-2331-2296)
- Dong‐Ho Lee (ORCID: https://orcid.org/0000-0002-7231-9008)
- Young-Ho Kim (ORCID: https://orcid.org/0000-0002-7954-9690)
Institutions
- Georgia Institute of Technology (US)
- Korea Advanced Institute of Science and Technology (KR)
- Korea Aerospace Research Institute (KR)
- Agency for Defense Development (KR)
Publication Details
- Journal
- Advanced Intelligent Systems
- Published
- 2025-03-01
- DOI
- https://doi.org/10.1002/aisy.202570011
- Citations
- 1
- Primary Topic
- Magnetic Field Sensors Techniques
- Type
- article
- Field-Weighted Citation Impact
- 0.73