7–12 May 2023
Venice, Italy
Europe/Zurich timezone

Free Electron Laser online optimization method based on deep reinforcement learning

TUPL087
9 May 2023, 16:30
2h
Sala Laguna

Sala Laguna

Poster Presentation MC2.A06: Free Electron Lasers Tuesday Poster Session

Speaker

Jiacheng Wu (Shanghai Synchrotron Radiation Facility)

Description

The lateral position deviation between the electron beam and undulators will lead to an interaction area decrease in practical high-gain free electron laser (FEL) equipment. Corrector magnets can be modified in the FEL control system to regulate the electron beam trajectory and promote laser power. Tuning tasks are time-varying, drifting, and multi-dimensional, and manual tuning by operators takes lots of time and effort. This paper proposes an online optimization algorithm using a twin delayed deep deterministic policy gradient (TD3) to automatically optimize laser energy under ever-changing conditions.

I have read and accept the Privacy Policy Statement Yes

Primary author

Jiacheng Wu (Shanghai Synchrotron Radiation Facility)

Presentation materials

There are no materials yet.