Speaker
Description
Fermilab FAST/IOTA is a research facility for advanced beam dynamics and a testbed for emerging technologies like AI/ML. As part of the recent IOTA proton injector (IPI) commissioning, we explored potential uses of various ML tools. End result was a novel methodology combining standard and ML algorithms into a comprehensive suite for simulation, software validation, staff training, performance optimization, and model calibration. Its’ three primary components were the differentiable virtual accelerator (beamline/instrumentation/control system), the Bayesian optimization/reinforcement learning suite, and the data management and orchestration layer. We will describe their implementation and integration, showcasing examples of various commissioning tasks like beam threading, phase space matching, and live real-to-sim model calibration. Overall, our experience has been highly positive. Work is now ongoing to expand and integrate ML tools into standard control room procedures.
| In which format do you inted to submit your paper? | LaTeX |
|---|