Speaker
Description
Plasma-based acceleration is widely regarded as a highly promising candidate technology for next-generation linear colliders and light sources. In the blowout regime of plasma wakefield acceleration, an intense particle beam excites a nonlinear plasma wake. However, there is currently no theory that can fully predict the asymmetric nonlinear wakefields generated by elliptical beam drivers. Machine learning provides an alternative approach for predicting physical quantities from simulation data with improved accuracy and efficiency. In particular, physics-informed methods incorporate known physical laws into the learning process, enabling accurate and interpretable surrogate models for complex systems. In this work, we focus on developing a nonlinear theory for plasma wakefield produced by elliptical electron beams via data-driven physical modeling. We employ physics-informed machine learning to approximate components of the blowout theory that are analytically difficult to describe, enabling the discovery of governing equations for the wakefields. Preliminary results show good agreement with simulations in both the blowout radius and wakefields.
| Working group | WG3 |
|---|