Accelerator beam orbit control method and system based on neural network inverse model
By analyzing beam position and correction signal data in real time and dynamically adjusting parameters and weights using an inverse mapping neural network model, the problem of mapping relationship drift in accelerator beam trajectory control was solved, ensuring the stability and accuracy of the beam trajectory.
Patent Information
- Application Number
- CN202610859255.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-15
- Publication Date
- 2026-07-17
AI Technical Summary
Existing accelerator beam trajectory control methods based on neural network inverse models cannot track in real time the dynamic changes in the mapping relationship caused by the drift of magnet excitation intensity in the storage ring and the misalignment of equipment installation, resulting in reduced beam trajectory stability.
By acquiring real-time data from the beam position detector and the power supply of the calibration magnet, the drift trend of the mapping relationship is analyzed using an inverse mapping neural network model. A mapping drift judgment identifier and a deviation judgment value are generated, and the model parameters and weights are dynamically adjusted to generate control commands to track the drift of the mapping relationship.
Real-time stable control of the beam track in the SSMB storage ring was achieved, maintaining residual track closure accuracy at the micrometer level, and enhancing control stability under long-term operation.
Smart Images
Figure CN122411077A_ABST