Inverse modelling of a laser-plasma accelerator
Abstract
The electron beams of a laser-plasma accelerator depend non-linearly on the laser and gas settings. This dependence is explored using particle-in-cell (PIC) simulations, and machine-learning models trained on these simulations predict the beam from the settings. Here we address the reverse problem: given a measured beam, which settings produced it? We infer three of them: the gas pressure p1, the normalised vector potential a0 and the nitrogen concentration cN2. The solution is not guaranteed to be unique. We train an ensemble of mixture density networks that, for each beam, output several possible sets of settings with their probabilities and uncertainties, and test them on simulated beams with realistic measurement noise. On average, the predictions are accurate (R2 = 0.91), yet for about two-thirds of the beams the inferred distribution has two competing solutions. These share nearly the same value of a02 p1, which controls how strongly the laser focuses itself in the plasma and, in earlier PALLAS studies, when electrons start to be injected. Keeping this quantity fixed gives beams that are practically indistinguishable with realistic diagnostics. A forward model used as a check removes unphysical solutions but cannot choose between the two plausible ones. We then test two ways to resolve the ambiguity. Adding the full electron energy spectrum improves the mean prediction (R2 = 0.94), mostly through a0. It changes the weights of the two solutions but does not move them. A second measurement at a deliberately different plasma density is more effective. In paired PIC simulations, it reduces the error on a0 by a factor of 2.6, whereas repeating the measurement at the same density brings almost no gain.
Related
- Direct observation of electron shedding from a laser-plasma accelerator
- Physics-Informed Drift Diagnosis for Laser-Plasma Accelerator Operations
- Microcoulomb-level electron beam and multi-Joule hard X-rays driven by a high-efficiency laser-plasma accelerator
- Stable 50 MeV Beams from a 100 Hz Laser-Wakefield Accelerator Driven by an OPCPA Laser
- Electron Dynamics in a Wakefield Accelerator Driven by Laser Pulses Carrying Orbital Angular Momentum
BibTeX
@misc{lenivenko2026inverse,
title = {{Inverse modelling of a laser-plasma accelerator}},
author = {Mykyta Lenivenko and Arnaud Beck and Hayg Guler and Gueladio Kane and Sophie Kazamias and Viacheslav Kubytskyi and Bruno Lucas and Marceau Pintonato and Jana Serhal and Slava Smartsev and Kalyani Swain and Kevin Cassou},
howpublished = {HAL},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2611.00108},
eprint = {2611.00108},
archivePrefix = {arXiv},
url = {https://hal.science/hal-05840690v1},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Inverse modelling of a laser-plasma accelerator
AU - Mykyta Lenivenko
AU - Arnaud Beck
AU - Hayg Guler
AU - Gueladio Kane
AU - Sophie Kazamias
AU - Viacheslav Kubytskyi
AU - Bruno Lucas
AU - Marceau Pintonato
AU - Jana Serhal
AU - Slava Smartsev
AU - Kalyani Swain
AU - Kevin Cassou
PY - 2026
DA - 2026/10/27
PB - HAL
DO - 10.48550/arxiv.2611.00108
UR - https://hal.science/hal-05840690v1
AB - The electron beams of a laser-plasma accelerator depend non-linearly on the laser and gas settings. This dependence is explored using particle-in-cell (PIC) simulations, and machine-learning models trained on these simulations predict the beam from the settings. Here we address the reverse problem: given a measured beam, which settings produced it? We infer three of them: the gas pressure p_1, the normalised vector potential a_0 and the nitrogen concentration c_N_2. The solution is not guaranteed to be unique. We train an ensemble of mixture density networks that, for each beam, output several possible sets of settings with their probabilities and uncertainties, and test them on simulated beams with realistic measurement noise. On average, the predictions are accurate (R^2 = 0.91), yet for about two-thirds of the beams the inferred distribution has two competing solutions. These share nearly the same value of a_0^2 p_1, which controls how strongly the laser focuses itself in the plasma and, in earlier PALLAS studies, when electrons start to be injected. Keeping this quantity fixed gives beams that are practically indistinguishable with realistic diagnostics. A forward model used as a check removes unphysical solutions but cannot choose between the two plausible ones. We then test two ways to resolve the ambiguity. Adding the full electron energy spectrum improves the mean prediction (R^2 = 0.94), mostly through a_0. It changes the weights of the two solutions but does not move them. A second measurement at a deliberately different plasma density is more effective. In paired PIC simulations, it reduces the error on a_0 by a factor of 2.6, whereas repeating the measurement at the same density brings almost no gain.
ER -