Research on AI Model Invocation Framework and Integration Platform for Plasma Control Systems
The increasing depth of nuclear fusion research and the rapid advancement of artificial intelligence (AI) applications have heightened the demand for intelligent plasma control systems. Currently, a new plasma control system is under development, which requires the integration of AI algorithms to enhance decision-making capabilities. To address this, this study has developed a comprehensive AI model invocation and integration solution tailored for plasma control systems.On one hand, an AI model invocation framework has been designed to enable the system to quickly call and execute models, facilitating real-time inference tasks. On the other hand, an AI model integration platform has been developed, which can automatically integrate models provided by developers into the system. This eliminates the need for users to deal with the intricate details of algorithm integration, thereby reducing unnecessary time consumption. Finally, the effectiveness of the invocation framework and integration platform has been demonstrated through a set of comparative experiments and an MHD mode recognition model integration case study.
Authors
- Qiping Yuan (ORCID: https://orcid.org/0000-0003-4292-1302)
- Zhenyuan Cai (ORCID: https://orcid.org/0000-0003-2066-2725)
- Junjie Huang (ORCID: https://orcid.org/0000-0002-6901-3374)
- Gen Xu
- Jianqiu Zhu
- Ruirui Zhang (ORCID: https://orcid.org/0000-0003-1898-1487)
- Zhongmin Huang (ORCID: https://orcid.org/0000-0003-3488-8864)
Institutions
- Chinese Academy of Sciences (CN)
- Hefei Institutes of Physical Science (CN)
- Institute of Plasma Physics (CN)
Publication Details
- Published
- 2025-03-21
- DOI
- https://doi.org/10.1109/iscait64916.2025.11010278
- Primary Topic
- Software Reliability and Analysis Research
- Type
- article
- Field-Weighted Citation Impact
- 0.00