17–22 May 2026
C.I.D
Europe/Zurich timezone

AI-Enabled Adaptive Control of Beam Current in an Isochronous Cyclotron Using GA-Tuned PID

MOP6342
18 May 2026, 16:00
2h
C.I.D

C.I.D

Deauville, France
Poster Presentation MC6.D13: Instrumentation: Artificial Intelligence Poster session

Speakers

Dr Claudio Lopez Osses (University of California, Davis) Gabriel Soto (University of California, Davis)

Description

The 76-inch isochronous cyclotron at UC Davis is being modernized for digital control and data-driven optimization. A key challenge is stabilizing the extracted proton beam current, currently done by manually adjusting Trim Coil 10 (TC10) against magnetic drifts and source fluctuations.

This work develops a digital single-input single-output (SISO) controller that automates TC10 to regulate beam current. The system combines classical PID feedback with data-driven optimization, using a genetic algorithm (GA) to adaptively tune PID gains in the cyclotron’s non-stationary environment.

We have implemented a 200 Hz data-acquisition and control pipeline with beam diagnostics, digital logging, and a Python-based control layer for autonomous operation. Current efforts focus on characterizing the beam’s response to TC10, identifying operational regimes, and validating the GA-tuned PID controller under varying machine conditions. This adaptive system forms a practical foundation for future AI-based multivariable optimization of cyclotron operation

In which format do you inted to submit your paper? LaTeX

Authors

Dr Claudio Lopez Osses (University of California, Davis) Eric Prebys (University of California, Davis) Gabriel Soto (University of California, Davis) Hans-Gerd Berns (University of California, Davis) Irwin Lopez (University of California, Davis) Jesus Avalos (University of California, Davis) Mathew Novotny (University of California, Davis) Michael Backfish (University of California, Davis) Rafiaullah Sahedzada (University of California, Davis)

Presentation materials

There are no materials yet.