Presentation Information
[3P90]Study on a pressure anomaly detection system applying machine learning for the SuperKEKB vacuum system -2-
*Yusuke Suetsugu1 (1. High Energy Accelerator Research Organization (KEK))
A pressure-anomaly detection system utilizing machine learning for the vacuum system of the SuperKEKB accelerator has been developed. The system identified abnormal pressure behaviors among approximately 600 vacuum gauges before triggering the conventional alarm system, facilitating the early implementation of countermeasures. In the system, by comparing the recent pressure behaviors of each vacuum gauge with the previous behaviors, the program detected anomalies using the decision boundary of a feed-forward neural network (FNN). Recently, another FNN was implemented in the system, capable of predicting possible causes of the anomalous behaviors. One feature of the system is its use of realistic regression models for pressure behaviors, enabling high-accuracy anomaly detection and cause prediction with a relatively small amount of data. The program, implemented in Python, has been operational since April 2024 for test.