Engineering Papers⌕ Search
Home/Archive/DOE OSTI

DOE OSTI · 3030149

Deep learning-assisted modeling for χ (2) nonlinear optics

Hirschman, Jack [Stanford Univ., CA (United States); SLAC National Accelerator Laboratory (SLAC), Menlo Park, CA (United States)]·Abedi, Erfan [Univ. of California, Los Angeles, CA (United States)]·Wang, Minyang [Univ. of California, Los Angeles, CA (United States)]·Zhang, Hao [SLAC National Accelerator Laboratory (SLAC), Menlo Park, CA (United States)]·Borthakur, Abhimanyu [Univ. of California, Los Angeles, CA (United States)]·Baker, Justin [Univ. of California, Los Angeles, CA (United States)]·Bertozzi, Andrea L. [Univ. of California, Los Angeles, CA (United States)]·Lemons, Randy [SLAC National Accelerator Laboratory (SLAC), Menlo Park, CA (United States)]·Carbajo, Sergio [SLAC National Accelerator Laboratory (SLAC), Menlo Park, CA (United States); Univ. of California, Los Angeles, CA (United States)]
View original source ↗Read source document ↗Export metadata ↓

Abstract

Modeling second-order (χ(2)) nonlinear optical processes remains computationally expensive due to the need to resolve fast field oscillations and simulate wave propagation using methods such as the split-step Fourier method (SSFM). This can become a bottleneck in real-time applications, such as high-repetition-rate laser systems requiring rapid feedback and control. We present a long short-term memory-based surrogate model trained on SSFM simulations generated from a start-to-end model of the photocathode drive laser at SLAC National Accelerator Laboratory’s Linac Coherent Light Source II. The model achieves over 250× speedup while maintaining high fidelity, enabling future real-time optimization and laying the foundation for data-integrated modeling frameworks and digital twins of laser systems.

Explore related subjects

Accelerator Physics (physics.acc-ph)Data modelingDigital twinEducation and trainingFOS: Physical sciencesLaser systems engineeringModelingNonlinear opticsOptics (physics.optics)Second harmonic generationSimulationsStanford Linear ColliderSum frequency generationSystems modelingΧ(2) machine learning

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Hirschman, Jack [Stanford Univ., CA (United States); SLAC National Accelerator Laboratory (SLAC), Menlo Park, CA (United States)], Abedi, Erfan [Univ. of California, Los Angeles, CA (United States)], Wang, Minyang [Univ. of California, Los Angeles, CA (United States)], Zhang, Hao [SLAC National Accelerator Laboratory (SLAC), Menlo Park, CA (United States)], Borthakur, Abhimanyu [Univ. of California, Los Angeles, CA (United States)], Baker, Justin [Univ. of California, Los Angeles, CA (United States)], Bertozzi, Andrea L. [Univ. of California, Los Angeles, CA (United States)], Lemons, Randy [SLAC National Accelerator Laboratory (SLAC), Menlo Park, CA (United States)], Carbajo, Sergio [SLAC National Accelerator Laboratory (SLAC), Menlo Park, CA (United States); Univ. of California, Los Angeles, CA (United States)]. 2026-05-06. Deep learning-assisted modeling for χ (2) nonlinear optics. https://doi.org/10.1117/1.ap.8.3.036004

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related reports

DOE OSTI2026

Monochromation of pulsed electron beams with terahertz radiation at a planar mirror

Exquisite control of electron beam energy is required for many electron spectroscopy and imaging applications. For both continuous and pulsed beams, the beam energy spread is fundamentally limited by the electron source, and is typically a sizable fraction of an electron-volt. In this paper, we present a means to reduce electron beam energy spread after emission to the level of a few 10s of meV rms using femtosecond photoemission and an interaction with laser-derived single- to few-cycle terahertz (THz) radiation. We show analytically and in particle tracking simulations that this interaction can remove energy spread stored in both the transverse and longitudinal degrees of freedom. We analytically formulate the limit of energy spread that this technique can achieve, and map the non-ideal affects arising at high frequencies. The interaction is mediated by the beam's passage through a mirror which is reflective to terahertz radiation but allows transmission of the majority of the electron beam (e.g. a wire mesh). This method then only requires beam current losses of a few tens of percent, far smaller than what is achieved in prism and slit-based electron monochromators.

Accelerator Physics (physics.acc-ph)↗
DOE OSTI2026

Laboratory Tests of Laser Control of Electron Beams for Future Colliders

Laser-driven Compton backscattering (CBS) has been proposed as method for controlling the intensity of colliding bunches in the FCC-ee so as to avoid the flip-flop instability caused by intensity asymmetry in colliding bunches. Laser-based collimation has also been proposed as an indestructible collimator for high-intensity electron beams. We have initiated a laboratory-based test program of these concepts with the E344 experiment at FACET-II. In this paper, we describe simulations of laser-beam interactions at FACET-II and the relevant scaling for FCC-ee. We also describe the experimental setup and diagnostics that will be used to make the measurements at FACET-II.

Accelerator Physics (physics.acc-ph)↗
DOE OSTI2025

Coincident learning for beam-based rf station fault identification using phase information at the SLAC linac coherent light source

Anomalies in radio-frequency (rf) stations can result in unplanned downtime and performance degradation in linear accelerators such as SLAC’s Linac Coherent Light Source (LCLS). Detecting these anomalies is challenging due to the complexity of accelerator systems, high data volume, and scarcity of labeled fault data. Prior work identified faults using beam-based detection, combining rf amplitude and beam position monitor data. Due to the simplicity of the rf amplitude data, classical methods are sufficient to identify faults, but the recall is constrained by the low-frequency and asynchronous characteristics of the data. In this work, we leverage high-frequency, time-synchronous rf phase data to enhance anomaly detection in the LCLS accelerator. Due to the complexity of phase data, classical methods fail, and we instead train deep neural networks within the Coincident Anomaly Detection (CoAD) framework. We find that applying CoAD to phase data detects nearly 3 times as many anomalies as when applied to amplitude data, while achieving broader coverage across rf stations. Furthermore, the rich structure of phase data enables us to cluster anomalies into distinct physical categories. Through the integration of auxiliary system status bits, we link clusters to specific fault signatures, providing additional granularity for uncovering the root cause of faults. We also investigate interpretability via Shapley values, confirming that the learned models focus on the most informative regions of the data and providing insight for cases where the model makes mistakes. This work demonstrates that phase-based anomaly detection for rf stations improves both diagnostic coverage and root cause analysis in accelerator systems and that deep neural networks are essential for effective analysis.

Accelerator Physics (physics.acc-ph)↗