Speaker
Description
As a critical infrastructure for advanced scientific research, the vacuum system of the China Spallation Neutron Source (CSNS) is essential for maintaining device performance and experimental reliability. Conventional vacuum system maintenance relies on expert experience and fixed threshold monitoring, leading to delayed fault detection and inaccurate parameter adjustments that fail to meet stringent stability requirements. This study integrates machine learning into the CSNS vacuum system's operational framework, developing a comprehensive dataset spanning multiple vacuum levels. Through rigorous data preprocessing and feature engineering, key diagnostic indicators are identified and a random forest-based fault diagnosis model is established. Validation using real operational data and simulation experiments demonstrates that the proposed machine learning approach significantly outperforms traditional methods in fault prediction accuracy. Results confirm that machine learning substantially enhances the intelligent operational maintenance capabilities of the CSNS vacuum system, providing a practical technical framework for auxiliary system upgrades in similar facilities.
Funding Agency
The National Natural Science Foundation of China (No. 12505184，12505183)
| In which format do you inted to submit your paper? | LaTeX |
|---|