Speaker
Description
Achieving low emittance at the photoinjector is essential for meeting the performance targets of the European XFEL, particularly for high photon energies and future high-duty-cycle operation. Both the temporal structure of the drive-laser pulse and the RF-gun settings contribute significantly to the final beam quality, yet their optimization is complicated by strong nonlinearities in the laser system and complex gun response. We have developed a differentiable, physics-based model of NEPAL, the photoinjector laser of EuXFEL, that enables gradient-driven optimization of the temporal UV pulse shape. The model captures the relevant nonlinearities of the optical chain and allows direct optimization of spectral amplitude and phase to obtain target UV profiles at the photocathode. In parallel, a machine-learning surrogate model is being implemented to optimize the RF-gun operating parameters. Together, these tools provide an end-to-end control framework for emittance reduction at EuXFEL. Initial results demonstrate that the differentiable model enables accurate temporal UV pulse shaping at EuXFEL. Work is ongoing to integrate this approach with ML-assisted gun optimization within the proposed end-to-end control framework.
Funding Agency
German Federal Ministry of Research, Technology and Space within the project OPAL-FEL, FKZ: 13D22CH6
| In which format do you inted to submit your paper? | LaTeX |
|---|---|
| Preprint marking on your proceeding paper | I wish my paper to be marked as preprint. |