Paper | Title | Page |
---|---|---|
WEP05 |
Online Machine Learning Version of the SNS Differential Beam Current Monitor | |
|
||
Funding: This manuscript has been authored by UT-Battelle, LLC, under contract DE-AC05-00OR22725 with the US Department of Energy (DOE). We have duplicated the Spallation Neutron Source (SNS) Differential beam Current Monitor (DCM) and included Machine Learning algorithms to observe the beam condition by looking for errant beam event precursors. The new system runs in parallel to the existing operational system and receives the same beam current signals but can be modified without affecting Operations. The archived data from the operational DCM was used to prove the existence of precursors and to generate the models. The new system has implemented Siamese Twin models on the real-time OS and a Random Forest model in the FPGA of the system. The system can also stream all acquired data at full rates to a data server for archival and online analysis. We discuss the setup and initial results. |
||
![]() |
Poster WEP05 [1.258 MB] | |
Cite • | reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml) | |