10 results
Now showing1 - 10 of 10
- Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , A physics-informed AI model for on-demand design of complex optical coatings(SPIE, 2026-05-28); ; ; ; The inverse design of multilayer optical coatings is a computationally intensive, non-convex optimization problem. We address this using a neural network framework trained to directly generate physical layer structures from target spectra. Our approach couples a ‘design’ network with a parameter-free, differentiable Transfer Matrix Method (TMM) model. During training, the network proposes a layer stack, the TMM model calculates its optical response, and the network’s parameters are adjusted to minimize the difference between this response and the target. This end to-end process allows the network to learn the physics of optical design directly without relying on any prior design data. Using this approach, we have generated a broadband dispersive mirror with performance characteristics comparable to those obtained by established commercial optimization tools. Building on this validated approach, we now show that this framework can be used to create a general, pre-trained model that rapidly generates initial designs for new problems, functioning as a powerful tool for optical coating design. Furthermore, the model architecture is inherently adaptable; it can be quickly fine-tuned on specific, user-defined design problems to further enhance its performance. This work demonstrates a practical pathway towards fast, generalizable, and continually improving AI tools for optical system design.Publicationtype: Conference Paper48 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Data-model-guided framework for systemic emittance minimization in high brightness photoinjectors(American Physical Society (APS), 2026-05-08); ; ; ; Bunch shaping in electron linear accelerators is of critical importance, as it enables the realization of prescribed projection profiles along specific coordinates of the bunch distribution for linac-based x-ray free-electron laser (XFEL) applications. It is evident that temporally flattop electron bunches yield more uniform slice properties, reduced emittance growth, and enhanced FEL gain compared to conventional Gaussian profiles. In this work, we present a data- and model-guided framework that consists of a physics-informed model and an inverse emittance model for minimizing the bunch emittance through laser pulse shaping and the control of laser and accelerator parameters in high-brightness photoinjectors. With a temporal flattop laser shape selected, a flat slice emittance distribution along the bunch has been experimentally obtained at the European XFEL. A measured central slice emittance of about 0.4 μm at 250 pC is achieved using a 20 ps flattop laser pulse at the photocathode. Here we present the results of a commissioning experiment, marking a practical application of temporally flattop-shaped electron bunches for stable photon delivery in a user experiment week at the facility. Within limited tuning time, a pulse energy of 1.7 mJ has been achieved at photon energies of 7.1 and 7.6 keV for the two hard x-ray undulator beamlines, surpassing the user requirements. Such a data- and model-driven framework opens up new opportunities for end-to-end optimization of machine parameters, thereby enhancing the capabilities of large-scale FEL facilities.Publicationtype: Journal Article41 33 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Learning a latent pulse shape interface for photoinjector laser systems(2026-02-19); ; ; ; Controlling the longitudinal laser pulse shape in photoinjectors of Free-Electron Lasers is a powerful lever for optimizing electron beam quality, but systematic exploration of the vast design space is limited by the cost of brute-force pulse propagation simulations. We present a generative modeling framework based on Wasserstein Autoencoders to learn a differentiable latent interface between pulse shaping and downstream beam dynamics. Our empirical findings show that the learned latent space is continuous and interpretable while maintaining high-fidelity reconstructions. Pulse families such as higher-order Gaussians trace coherent trajectories, while standardizing the temporal pulse lengths shows a latent organization correlated with pulse energy. Analysis via principal components and Gaussian Mixture Models reveals a well behaved latent geometry, enabling smooth transitions between distinct pulse types via linear interpolation. The model generalizes from simulated data to real experimental pulse measurements, accurately reconstructing pulses and embedding them consistently into the learned manifold. Overall, the approach reduces reliance on expensive pulse-propagation simulations and facilitates downstream beam dynamics simulation and analysis.Publicationtype: Preprint64 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Physics-informed inverse design of ultrafast coatings: from direct optimization to generalizable fine-tuning(Optica Publishing Group, 2026); ; ; ; We present a physics-informed neural network for optical coating design. We demonstrate state-of-the-art chirped mirror optimization and introduce a generalist model with parameter-efficient fine-tuning to rapidly generate high-precision designs for arbitrary targets.Publicationtype: Conference Paper15 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Efficient optical coating design using an autoencoder-based neural network model(IOP Publishing, 2025-11-12); ; ; ; Optical thin-film coatings are integral to modern photonics and in particular to ultrafast lasers, providing precise control of dispersion and reflectivity, thus enabling tailored pulse shaping. Designing these coatings represents an inverse problem, requiring the mapping of desired optical properties to physical designs, a task that poses major challenges for traditional heuristic methods, which often rely on time-consuming, expert-guided iteration and can be constrained by the choice of an initial starting point. Here, we present an artificial intelligence (AI) framework for optical thin-film coating design that accelerates the design process, achieving excellent performance characteristics without expert intervention. We discuss our AI approach and demonstrate the capabilities of our algorithm by designing a complex broadband high-reflectivity mirror with state-of-the-art performance characteristics including a group delay dispersion covering a spectral range of 940 nm to 1120 nm.Publicationtype: Journal Article25 46 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Novel photoinjector laser providing advanced pulse shaping for FLASH and EuXFELWe recently commissioned the Next GenerAtion Photocathode Laser system (NEPAL) in Hamburg’s XFEL facilities (FLASH and EuXFEL) and at DESY’s Photoinjector Test Facility (PITZ). The system delivers deep UV pulse trains up to 1 ms long at repetition rates as high as 4.5 MHz, with temporal and spatial shaping capabilities and individual amplitude control for bunch charge manipulation. The shaping features enable the generation of exceptionally low emittance electron beams, essential for extending the EuXFEL X-ray photon energy beyond 25 keV and for future high duty cycle upgrades. Temporal shaping is achieved through a high-resolution spatial light modulator in the near-infrared driver laser, allowing precise spectral amplitude and phase control of UV pulses. We will present advanced control schemes that pre-compensate for laser nonlinearities and initial experimental results at EuXFEL. We generated UV flat-top pulse profiles with durations ranging from 10ps to 20ps and successfully transferred them onto the electron beam. This achievement represents a significant step toward emittance optimization at EuXFEL and will expand the facility's operational energy range in the near future.Publicationtype: Conference Paper33 34 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Temporal shaping of UV picosecond pulses for photoinjectors(2025); ; 22 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , AI-driven design of high-performance optical thin film coatings for ultrafast lasersUltrafast laser systems critically depend on optical thin film coatings to control dispersion and light propagation, enabling precise shaping of light. Optical thin film coating design represents a complex inverse problem traditionally relying on computationally intensive numerical optimization methods. These conventional approaches, exemplified by the widely used Needle Algorithm [1], require significant computational resources and often rely on expert intervention [2]. We here propose a new approach to these challenges and present a physics-informed machine learning framework based on an autoencoder architecture and use it to design an ultra-broadband dispersive mirror.Publicationtype: Conference Paper23 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Neural network method for dielectric optical coating design(EDP Sciences, 2024-10-18); ; ; ; We use neural networks to address the challenge of deriving dielectric coating designs from the desired optical properties. We show that our trained neural network can automatically design common laser mirror coatings types efficiently.Publicationtype: Journal Article50 46 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Characterization of low-emittance electron beams generated by a new photocathode drive laser system NEPAL at the European XFELAn ultrafast laser system for driving the photocathode RF gun at the European XFEL has been recently put into operation. The new laser system, NExt generation PhotocAthode Laser (NEPAL) is capable of providing drive laser pulses of variable pulse lengths and shapes, supporting the facility to extend its capabilities to operate in multiple user-desirable FEL modes. In this paper, we present a preliminary characterization of the low-emittance electron beams produced by NEPAL in the photoinjector. Both experimental and numerical results will be presented and discussed.Publicationtype: Conference Paper16 29