Research on neutron source accelerators visual-real fusion fault diagnosis with Digital Twin

To address the intelligent upgrading needs of neutron source accelerators, a digital twin visualization system was designed and developed based on digital twin technology. The virtual mapping model was constructed in the digital space using actual parameters of physical space equipment, establishing a mapping connection between the physical and digital spaces centered around twin data. And a multi-channel integrated learning loop diagnostic model with virtual-physical fusion was proposed. Focusing on the experimental operational data of the equipment, the diagnostic model features data monitoring and multi-channel diagnostic functions. Input thresholds for each diagnostic channel of the model were configured based on the data scale. The final diagnostic result with higher accuracy was obtained by comprehensively comparing the results of each channel, significantly improving the real-time performance and efficiency of the diagnostic process. The effectiveness of the multichannel diagnostic model was experimentally validated using vibration signals generated during equipment operation. Additionally, the digital twin visualization system implemented data information visualization and intelligent management of accelerator equipment.