SEARCH · Search NASA

Results for “Architecture Study”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 55 records · Page 3

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Explainable AI for Multivariate Time Series Pattern Exploration: Latent Space Visual Analytics With Temporal Fusion Transformer and Variational Autoencoders in Power Grid Event Diagnosis

Detecting and analyzing complex patterns in multivariate time-series data is crucial for decision-making in urban and environmental system operations. However, challenges arise from the high dimensionality, intricate complexity, and interconnected nature of complex patterns, which hinder the understanding of their underlying physical processes. Existing AI methods often face limitations in interpretability, computational efficiency, and scalability, reducing their applicability in real-world scenarios. This paper proposes a novel visual analytics framework that integrates two generative AI models, Temporal Fusion Transformer (TFT) and Variational Autoencoders (VAEs), to reduce complex patterns into lower-dimensional latent spaces and visualize them in 2D using dimensionality reduction techniques such as PCA, t-SNE, and UMAP with DBSCAN. These visualizations, presented through coordinated and interactive views and tailored glyphs, enable intuitive exploration of complex multivariate temporal patterns, identifying patterns’ similarities and uncover their potential correlations for a better interpretability of the AI outputs. The framework is demonstrated through a case study on power grid signal data, where it identifies multi-label grid event signatures, including faults and anomalies with diverse root causes. Additionally, novel metrics and visualizations are introduced to validate the models and assess the performance, efficiency, and consistency of latent maps generated by VAE, which have been utilized in prior studies for latent space cartography and used as a benchmark in this study, and the emerging TFT architecture under various configurations. These analyses provide actionable insights for model parameter tuning and reliability improvements. Comparative results highlight that TFT achieves shorter run times and superior scalability to diverse time-series data shapes compared to VAE. This work advances fault diagnosis in multivariate time series, fostering explainable AI to support critical system operations.

Explainable AI↗

DOE OSTI2024

Use of carbon electrodes to reduce mobile ion concentration and improve reliability of metal halide perovskite photovoltaics

Ion migration is one of the prime reasons for the rapid degradation of metal halide perovskite solar cells (PSCs), and we report on a method for quantifying mobile ion concentration (No) using a transient dark current measurement. We perform both ex-situ and in-situ measurements on PSCs and study the evolution of No in films and devices under a range of temperatures. We also study the effect of device architecture, top electrode chemistry, and metal halide perovskite composition and dimensionality on No. Two-dimensional perovskites are shown to reduce the ion concentration along with inert C electrodes that do not react with halides by ~99% while also improving mechanical reliability by ~250%. We believe this work can provide design guidelines for the development of stable PSCs through the lens of minimizing mobile ions and their evolution over time under operational conditions.

14 SOLAR ENERGY↗

DOE OSTI2026

Heavy-Duty Nonroad Material Handler Electrification Part 2: Energy Efficiency and Performance Evaluation

Decarbonization efforts achieved through electrification in nonroad mobile machinery can realize a reduction in fuel consumption of more than 20%, thanks to concepts familiar to light-duty passenger vehicles. This case study compares the results of a hybrid-electric material handler to its conventional counterpart, utilizing machine-specific drive cycles presented in part one of this paper series. The hybrid prototype features an extended-range electric vehicle (EREV) powertrain that demonstrated substantial energy efficiency improvements. Specifically, there was a reduction in equivalent fuel consumption of 75% when operating in electric-only mode, and 33% when maintaining the battery by charging with an on-board generator. Together, the efficiency improvements can be extrapolated over a low-intensity, 8-h shift characterized by significant idle time and highly dynamic engine load for a 47% reduction in net energy consumption. Key technologies that led to this improvement included engine downsizing and decoupling, regenerative braking, and an electrohydraulic pump unit with advanced controls. Finally, his study explains details of the powertrain architecture and subsystems that were implemented on a demonstration vehicle, control strategies used to meet project goals, and an analysis of energy consumption from testing on a closed course. Also included in this study is a discourse on comparison metrics that can be used for quantifying the energy consumption differences between hybrid-electric and conventional diesel powertrains in nonroad mobile machinery.

25 ENERGY STORAGE↗

DOE OSTI2025

Control Oriented Models for Co-Design: Technical Overview of MT HVDC, MVDC, and Solid State Transformer Building Blocks

The electric power system is shifting toward a power electronics–enabled grid, where converter based “building blocks” (e.g., high voltage direct current (HVDC) links, multi terminal HVDC (MT HVDC) networks, medium voltage DC (MVDC) links, and solid state transformers (SSTs)) provide fast, precise control of power flows, voltage, and frequency. This report develops and applies publicly shareable electromagnetic transient (EMT) and phasor models to examine how such building blocks can be composed and coordinated to support offshore wind integration, inter area transfers, feeder support, and resilience. Section 2 documents a modular multilevel converter (MMC)–based MT HVDC modeling framework and two use cases: a compact WSCC/IEEE 9 bus test system and a 240 bus “mini WECC” case with five offshore wind plants (OWFs). Phasor to EMT transfer, initialization, and sanity checks are summarized, and neutral demonstrations of normal and contingency operation are reported. Section 3 frames the problem of wind plant inertial frequency response (IFR): shaping energy release and recovery to improve nadir while avoiding aerodynamic stall; representative simulations illustrate the issues without disclosing proprietary control. Section 4 develops MVDC concepts through an IEEE 16 bus loop and an Olympic Peninsula case study that compares AC vs. MVDC corridors and shows how feeder headroom can be pooled via DC couplers. Section 5 surveys SST architectures and identifies a gap: scalable, communication free coordination of multiple SSTs for islanded feeder networks. Across the report, novel methods and configurations under separate publication and IP review are not disclosed; only topic oriented, replicable setups and non proprietary results are shown. These models and use cases are intended as foundations for future publications and co design studies on architecture, control, and coordination of PE enabled grids.

24 POWER TRANSMISSION AND DISTRIBUTION↗

DOE OSTI2024

Enabling Parallel Performance and Portability of Solid Mechanics Simulations Across CPU and GPU Architectures

Efficiently simulating solid mechanics is vital across various engineering applications. As constitutive models grow more complex and simulations scale up in size, harnessing the capabilities of modern computer architectures has become essential for achieving timely results. This paper presents advancements in running parallel simulations of solid mechanics on multi-core CPUs and GPUs using a single-code implementation. This portability is made possible by the C++ matrix and array (MATAR) library, which interfaces with the C++ Kokkos library, enabling the selection of fine-grained parallelism backends (e.g., CUDA, HIP, OpenMP, pthreads, etc.) at compile time. MATAR simplifies the transition from Fortran to C++ and Kokkos, making it easier to modernize legacy solid mechanics codes. We applied this approach to modernize a suite of constitutive models and to demonstrate substantial performance improvements across different computer architectures. This paper includes comparative performance studies using multi-core CPUs along with AMD and NVIDIA GPUs. Results are presented using a hypoelastic–plastic model, a crystal plasticity model, and the viscoplastic self-consistent generalized material model (VPSC-GMM). The results underscore the potential of using the MATAR library and modern computer architectures to accelerate solid mechanics simulations.

Morgan, Nathaniel (ORCID:0000000276118449)↗

DOE OSTI2025

Vidyut3d: A Gpu Accelerated Fluid Solver for Non-Equilibrium Plasmas on Adaptive Grids

We present the numerical methods, programming methodology, verification, and performance assessment of a non-equilibrium plasma fluid solver that can effectively utilize current and upcoming central processing and graphics processing unit (CPU+GPU) architectures, in this work. Our plasma fluid model solves the coupled conservation equations for species transport, electrostatic Poisson and electron temperature on adaptive Cartesian grids. Our solver is written using performance portable adaptive-grid/particle management library, AMReX, and is portable over widely available vendor specific GPU architectures. We present verification of our solver using method of manufactured solutions that indicate formal second order accuracy with central diffusion and fifth-order weighted-essentially-non-oscillatory (WENO) advection scheme. We also verify our solver with published literature on capacitive discharges and atmospheric pressure streamer propagation. We demonstrate the use of our solver on two 3D simulation cases: an atmospheric streamer propagation in Ar-H2 mixtures and a low pressure twin electrode radio frequency reactor. Our performance studies on three different CPU+GPU architectures indicate approximately 150-400X speed-up using AMD and NVIDIA GPUs per time step compared to a single CPU core for a 4 million cell simulation with 15 species.

Sitaraman, Hariswaran↗

DOE OSTI2026

Vidyut3d: A GPU accelerated fluid solver for non-equilibrium plasmas on adaptive grids

We present the numerical methods, programming methodology, verification, and performance assessment of a non-equilibrium plasma fluid solver that can effectively utilize current and upcoming central processing and graphics processing unit (CPU+GPU) architectures, in this work. Our plasma fluid model solves the coupled conservation equations for species transport, electrostatic Poisson and electron temperature on adaptive Cartesian grids. Our solver is written using performance portable adaptive-grid/particle management library, AMReX, and is portable over widely available vendor specific GPU architectures. We present verification of our solver using method of manufactured solutions that indicate formal second order accuracy with central diffusion and fifth-order weighted-essentially-non-oscillatory (WENO) advection scheme. We also verify our solver with published literature on capacitive discharges and atmospheric pressure streamer propagation. We demonstrate the use of our solver on two 3D simulation cases: an atmospheric streamer propagation in Ar-H2 mixtures and a low pressure three-electrode radio frequency reactor. Our performance studies on three different CPU+GPU architectures indicate ~ 150-400X speed-up using AMD and NVIDIA GPUs per time step compared to a single CPU core for a 4 million cell simulation with 15 species.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2024

Integrating multi-modal remote sensing, deep learning, and attention mechanisms for yield prediction in plant breeding experiments

In both plant breeding and crop management, interpretability plays a crucial role in instilling trust in AI-driven approaches and enabling the provision of actionable insights. The primary objective of this research is to explore and evaluate the potential contributions of deep learning network architectures that employ stacked LSTM for end-of-season maize grain yield prediction. A secondary aim is to expand the capabilities of these networks by adapting them to better accommodate and leverage the multi-modality properties of remote sensing data. In this study, a multi-modal deep learning architecture that assimilates inputs from heterogeneous data streams, including high-resolution hyperspectral imagery, LiDAR point clouds, and environmental data, is proposed to forecast maize crop yields. The architecture includes attention mechanisms that assign varying levels of importance to different modalities and temporal features that, reflect the dynamics of plant growth and environmental interactions. The interpretability of the attention weights is investigated in multi-modal networks that seek to both improve predictions and attribute crop yield outcomes to genetic and environmental variables. This approach also contributes to increased interpretability of the model's predictions. The temporal attention weight distributions highlighted relevant factors and critical growth stages that contribute to the predictions. The results of this study affirm that the attention weights are consistent with recognized biological growth stages, thereby substantiating the network's capability to learn biologically interpretable features. Accuracies of the model's predictions of yield ranged from 0.82-0.93 R 2 ref in this genetics-focused study, further highlighting the potential of attention-based models. Further, this research facilitates understanding of how multi-modality remote sensing aligns with the physiological stages of maize. The proposed architecture shows promise in improving predictions and offering interpretable insights into the factors affecting maize crop yields, while demonstrating the impact of data collection by different modalities through the growing season. By identifying relevant factors and critical growth stages, the model's attention weights provide valuable information that can be used in both plant breeding and crop management. The consistency of attention weights with biological growth stages reinforces the potential of deep learning networks in agricultural applications, particularly in leveraging remote sensing data for yield prediction. To the best of our knowledge, this is the first study that investigates the use of hyperspectral and LiDAR UAV time series data for explaining/interpreting plant growth stages within deep learning networks and forecasting plot-level maize grain yield using late fusion modalities with attention mechanisms.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2025

Backbone Dynamics of Bottlebrush Polymers Studied by Neutron Scattering

Bottlebrush polymers have a versatile architecture that is highly customizable due to the combination of a linear backbone and side chains. As a result of the chemical tethering of the side chains, both parts cannot be easily be separated. Huge effort is seen regarding the dynamical behavior of the side chains or the entire bottlebrush polymer, whereas few studies are available considering the backbone. Here, isotopic labeling in combination with quasi-elastic neutron scattering was used to compare the dynamical behavior of the bottlebrush’s backbone with the side chain dynamics. Keeping the side chains deuterated, (h-PNB)- g -(d-PPO), leads to the scattering signal dominated by backbone dynamics, while the fully protonated sample, (h-PNB)- g -(h-PPO), gives side chain dynamics. Both results reveal slower dynamics associated with the backbone with less heterogeneity, as seen for the side chains. Additionally, a plasticizer effect for the backbone dynamics is confirmed by extracting the glass transition temperature and comparing it with pure linear PNB.

neutron scattering↗

DOE OSTI2025

Impacts of molecular architecture on the radiation-induced degradation and reaction kinetics of hydrophobic diglycolamides with the solvated electron and the dodecane radical cation

Given their proposed use as trivalent actinide–lanthanide separation ligands, the role of molecular architecture on the radiation robustness of diglycolamide (DGA) molecules has been investigated. This study examined three prototypical molecules with differences in their aliphatic chain architecture: N,N,N′,N′-tetra(n-octyl)diglycolamide (TODGA), N,N,N′,N′-tetra(2-ethylhexyl)diglycolamide (T2EHDGA), and N,N′-dimethyl-N,N′-dioctyldiglycolamide (DMDODGA). Rate coefficients and activation parameters are reported for the reactivity of each DGA with the solvated electron (e solv − ) and the corresponding dodecane radical cation (RH˙ + ) over the temperature range of 10.0 to 44.1 °C. These measurements indicate that DMDODGA is the most chemically reactive with both transient radicals, which may be attributed to this molecule's more accessible backbone. Complementary gamma dose accumulation studies (≤ 600 kGy) under envisioned process conditions—50 mM DGA in n-dodecane solvent—afforded dose constants for the loss of DGA of d = (3.41 ± 0.07) × 10 −3 , (4.19 ± 0.09) × 10 −3 , and (4.65 ± 0.09) × 10 −3 kGy −1 for T2EHDGA, DMDODGA, and TODGA, respectively. These dose constants indicate that varying DGA architecture affords subtle differences in chemical reactivity, leading to varying rates of radiolytic degradation under envisioned actinide–lanthanide separation conditions. However, more ambitious DGA frameworks, such as modifying the backbone, branching of the aliphatic chains, and/or changing the size of the chain may be required for larger gains in radiolytic longevity while optimizing actinide–lanthanide selectivity.

Arrhenius parameters↗

DOE OSTI2025

OR-AGENT framework – Architecting electrified heavy-duty drayage applications

The widespread adoption of zero-emission vehicles in heavy-duty (HD) commercial freight transportation faces considerable technoeconomic challenges. For heavy-duty trucks, ensuring high uptime, cost parity with diesel, and safety standards is especially critical as these vehicles operate over long distances with heavy loads, where any downtime or off-nominal behaviors significantly impacts logistics, productivity, and the total cost of ownership. Unlike traditional diesel refueling, BEV charging infrastructure must be co-optimized with vehicle deployment, operational demands, and grid capacity to ensure cost-effective and reliable freight operations. However, the lack of a standardized ownership and service model has led to a fragmented approach—where commercial vehicle operators may invest in, own, and maintain both vehicle/batteries and charging/energy infrastructure. This disconnect may exclude energy service providers from the equation, forcing fleet operators to explore ‘behind-the-fence’ energy solutions that increase capital investment, operational downtime, overhead costs, and, in some cases, net carbon emissions. To address these issues, this study introduces OR-AGENT (Optimal Regional Architecture Generation for Efficient National Transport), a comprehensive modeling framework that integrates powertrain architectures, charging infrastructures, and energy backbone systems into a cohesive strategy. In this paper, OR-AGENT is applied to develop an interconnected systems architecture for energy efficiency and resiliency enhancement of heavy-duty drayage vehicles at the Port of Savannah, GA. This framework showcases an interconnected systems approach to electrifying heavy-duty drayage vehicles at the Port of Savannah, GA. The study assessed BEVs with 400–1200 kWh battery capacities, accounting for seasonal variations in weather and freight routing. A diverse charging mix (150 kW–1250 kW) was evaluated alongside grid capacity constraints, cost, and carbon intensity analysis, leading to the development of a strategic microgrid/Distributed Energy Resources (DER) deployment architecture to ensure a reliable and sustainable transition. However, the findings also highlight the need for alternative zero-emission solutions for remaining trips, such as larger batteries, electrified roadways, hydrogen powertrains, or net-zero emission fuels. In conclusion, the findings are incorporated into a Total Cost of Ownership (TCO) model to identify optimal architectures for an interconnected electrified ecosystem.

Commercial vehicles↗

DOE OSTI2024

Exploring Saccharomycotina Yeast Ecology Through an Ecological Ontology Framework

Yeasts in the subphylum Saccharomycotina are found across the globe in disparate ecosystems. A major aim of yeast research is to understand the diversity and evolution of ecological traits, such as carbon metabolic breadth, insect association, and cactophily. This includes studying aspects of ecological traits like genetic architecture or association with other phenotypic traits. Genomic resources in the Saccharomycotina have grown rapidly. Ecological data, however, are still limited for many species, especially those only known from species descriptions where usually only a limited number of strains are studied. Moreover, ecological information is recorded in natural language format limiting high throughput computational analysis. To address these limitations, we developed an ontological framework for the analysis of yeast ecology. A total of 1,088 yeast strains were added to the Ontology of Yeast Environments (OYE) and analyzed in a machine-learning framework to connect genotype to ecology. This framework is flexible and can be extended to additional isolates, species, or environmental sequencing data. Widespread adoption of OYE would greatly aid the study of macroecology in the Saccharomycotina subphylum.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2026

RU-net for automatic characterization of TRISO fuel cross sections

During irradiation, phenomena such as kernel swelling and buffer densification may impact the performance of tristructural isotropic (TRISO) particle fuel. Post-irradiation microscopy is often used to identify these irradiation-induced morphologic changes. However, each fuel compact generally contains thousands of TRISO particles. Manually performing the work to get statistical information on these phenomena is cumbersome and subjective. Here, to reduce the subjectivity inherent in that process and to accelerate data analysis, we used convolutional neural networks (CNNs) to automatically segment cross-sectional images of microscopic TRISO layers. CNNs are a class of machine-learning algorithms specifically designed for processing structured grid data. They have gained popularity in recent years due to their remarkable performance in various computer vision tasks, including image classification, object detection, and image segmentation. In this research, we generated a large irradiated TRISO layer dataset with more than 2,000 microscopic images of cross-sectional TRISO particles and the corresponding annotated images. Based on these annotated images, we used different CNNs to automatically segment different TRISO layers. These CNNs include RU-Net (developed in this study), as well as three existing architectures: U-Net, Residual Network (ResNet), and Attention U-Net. The preliminary results show that the model based on RU-Net performs best in terms of Intersection over Union (IoU). Using CNN models, we can expedite the analysis of TRISO particle cross sections, significantly reducing the manual labor involved and improving the objectivity of the segmentation results.

11 - NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

DOE OSTI2026

Synthesis and Characterization of Layered Actinide (U, Np, Pu) Oxide and Hydroxide Phases

Systematic structural comparisons across the early actinides remain limited by the scarcity of well-defined transuranic layered oxide and oxyhydroxide phases. Here, we report the synthesis and single-crystal characterization of new layered actinide compounds spanning U, Np, and Pu obtained under mild hydrothermal conditions in concentrated alkali hydroxide media. These include hydrated oxides α-Cs 2 U 2 O 7 •0.5H 2 O and Rb 2 An 2 O 7 •0.5H 2 O (An = Np, Pu), oxy-hydroxides Rb 6 [(AnO 2 ) 6 O 8 (OH) 2 ]•xH 2 O x = 0, 0.5 (An = U, Np) and Rb 4 [(UO 2 ) 5 O 6 (OH) 2 ]•2H 2 O, as well as nitrate-intercalated compounds Cs 6 [(AnO 2 ) 3 O 4 (OH) 2 ](NO 3 ) 2 (An = Np, Pu). Single-crystal X-ray diffraction studies reveal extended two-dimensional architectures constructed from edge- and vertex-sharing actinyl polyhedra, with systematic evolution in equatorial coordination, hydration, and anionic sheet topology across the U–Np–Pu series. Incorporation of nitrate anions within the interlayer region of Cs 6 [(AnO 2 ) 3 O 4 (OH) 2 ](NO 3 ) 2 establishes a previously unobserved structural motif in layered transuranic oxyhydroxides, demonstrating an additional pathway for anion-mediated framework stabilization. Correlation of crystallographic metrics with single-crystal Raman spectroscopy provides new vibrational benchmarks linking differences in An═O yl bond lengths to the equatorial coordination and the interstitial cations. These findings expand the structural hierarchy of layered actinide materials and address clarifying periodic trends governing topology, bonding, and vibrational signatures in high-valent 5ƒ oxide systems.

actinides↗

DOE OSTI2025

Scaling Laws of Graph Neural Networks for Atomistic Materials Modeling

Atomistic materials modeling is a critical task with wide-ranging applications, from drug discovery to materials science, where accurate predictions of the target material property can lead to significant advancements in scientific discovery. Graph Neural Networks (GNNs) represent the state-of-the-art approach for modeling atomistic material data thanks to their capacity to capture complex relational structures. While machine learning performance has historically improved with larger models and datasets, GNNs for atomistic materials modeling remain relatively small compared to large language models (LLMs), which leverage billions of parameters and terabyte-scale datasets to achieve remarkable performance in their respective domains. To address this gap, we explore the scaling limits of GNNs for atomistic materials modeling by developing a foundational model with billions of parameters, trained on extensive datasets in terabytescale. Our approach incorporates techniques from LLM libraries to efficiently manage large-scale data and models, enabling both effective training and deployment of these large-scale GNN models. This work addresses three fundamental questions in scaling GNNs: the potential for scaling GNN model architectures, the effect of dataset size on model accuracy, and the applicability of LLM-inspired techniques to GNN architectures. Specifically, the outcomes of this study include (1) insights into the scaling laws for GNNs, highlighting the relationship between model size, dataset volume, and accuracy, (2) a foundational GNN model optimized for atomistic materials modeling, and (3) a GNN codebase enhanced with advanced LLM-based training techniques. Our findings lay the groundwork for large-scale GNNs with billions of parameters and terabyte-scale datasets, establishing a scalable pathway for future advancements in atomistic materials modeling.

Li, Chaojian [ORNL] (ORCID:0000000340309777)↗

DOE OSTI2024

Pretraining Billion-Scale Geospatial Foundational Models on Frontier

As AI workloads increase in scope, generalization capability becomes challenging for small task-specific models and their demand for large amounts of labeled training samples increases. On the contrary, Foundation Models (FMs) are trained with internet-scale unlabeled data via self-supervised learning and have been shown to adapt to various tasks with minimal fine-tuning. Although large FMs have demonstrated significant impact in natural language processing and computer vision, efforts toward FMs for geospatial applications have been restricted to smaller size models, as pretraining larger models requires very large computing resources equipped with state-of-the-art hardware accelerators. Current satellite constellations collect 100+TBs of data a day, resulting in images that are billions of pixels and multimodal in nature. Such geospatial data poses unique challenges opening up new opportunities to develop FMs. We investigate billion scale FMs and HPC training profiles for geospatial applications by pretraining on publicly available data. We studied from end-to-end the performance and impact in the solution by scaling the model size. Our larger 3B parameter size model achieves up to 30% improvement in top1 scene classification accuracy when comparing a 100M parameter model. Moreover, we detail performance experiments on the Frontier supercomputer, America's first exascale system, where we study different model and data parallel approaches using PyTorch's Fully Sharded Data Parallel library. Specifically, we study variants of the Vision Transformer architecture (ViT), conducting performance analysis for ViT models with size up to 15B parameters. By discussing throughput and performance bottlenecks under different parallelism configurations, we offer insights on how to leverage such leadership-class HPC resources when developing large models for geospatial imagery applications.

Tsaris, Aristeidis (aris)↗

DOE OSTI2026

RU Net for Automatic Characterization of TRISO Fuel Cross Sections

TRistructural ISOtropic (TRISO) particle fuel is a type of nuclear fuel known for its high-temperature and high-burnup performance. Each sub-millimeter diameter TRISO particle consists of uranium-oxycarbide (UCO) or UO2 fuel kernel, coated with buffer, inner pyrolytic carbon (IPyC), silicon carbide (SiC), and outer pyrolytic carbon (OPyC) layers. The SiC layer acts as the main containment barrier for the TRISO particle to retain the fission products, while the IPyC and OPyC layers provide additional barriers to the release of fission products, especially fission gases. During irradiation, phenomena like kernel swelling, buffer densification, and IPyC fracture may impact fuel performance. Post-irradiation microscopy on entire compact cross sections or samples of individual particles deconsolidated from compacts is often used to identify these irradiation-induced changes in morphology. However, each fuel compact generally contains thousands of TRISO particles. To get statistical information on these phenomena, it is cumbersome work if done manually. For example, to get information about swelling/densification behaviors of different layers or kernels after irradiation, researchers previously manually measured the perimeter of each TRISO layer in hundreds of particles after four rounds of iterative grinding and polishing encompassing more than 2000 cross-section images for a total of four fuel compacts. To attempt to reduce the subjectivity inherent in that process and accelerate data analysis, we conducted a study on the automatic TRISO layer segmentation on cross-sectional microscopic images using Convolutional Neural Networks (CNNs). CNNs are a class of machine learning algorithms specifically designed for processing structured grid data that have gained popularity in recent years due to their remarkable performance in various computer vision tasks, including image classification, object detection, and image segmentation. In this research, we have generated the large irradiated TRISO layer dataset with more than 2000 cross-section TRISO microscopic images and the corresponding annotated images. Based on these annotated images, we have employed different CNNs for automatic segmentation of different TRISO layers. These include RU-Net (developed in this study), as well as three existing architectures: U-Net, Residual Network (ResNet), and Attention U-Net. The preliminary results show that the model based on RU-Net has the best performance in terms of intersection-over-union (IoU). Through the aid of these CNN models, we can expedite the analysis of TRISO particle cross-sections, significantly reducing the manual labor involved and improving the objectivity of the segmentation results.

Convolutional Neural Networks↗

DOE OSTI2025

Analyzing and Exploring Training Recipes for Large-Scale Transformer-Based Weather Prediction

Abstract The rapid rise of deep learning (DL) in numerical weather prediction (NWP) has led to a proliferation of models which forecast atmospheric variables with comparable or superior skill than traditional physics-based NWP. However, among these leading DL models, there is a wide variance in both the training settings and architecture used. Further, the lack of thorough ablation studies makes it hard to discern which components are most critical to success. In this work, we show that it is possible to attain high forecast skill even with relatively off-the-shelf architectures, simple training procedures, and moderate compute budgets. Specifically, we train a minimally modified Swin Transformer V2 (SwinV2) on ERA5 data and find that it attains superior skill in terms of mean-square errors of deterministic forecasts when compared against the European Centre for Medium-Range Weather Forecasts’ Integrated Forecasting System (IFS). Almost all DL–NWP systems share a core set of hyperparameters and design decisions. To aid and expedite future DL–NWP research, we present an in-depth, systematic exploration of different loss functions, model sizes and depths, patch sizes, and multistep training objectives. We also examine the model performance with metrics beyond the typical accuracy (ACC) and RMSE and investigate how the performance scales with model size. Through our open-source code, scoring pipelines, and models, we share our findings on key aspects of the training pipeline. These ablations reduce the necessity for expensive hyperparameter tuning and lower the barrier to entry for future DL–NWP research. Significance Statement This study investigates the potential of using large-scale transformer-based models for weather prediction, showing that it is possible to achieve high forecast accuracy with simpler, off-the-shelf architectures. By training a minimally modified SwinV2 transformer on ERA5 data, we show that the model achieves competitive forecast skill in terms of mean-square error for key variables, outperforming the European Centre for Medium-Range Weather Forecasts’ Integrated Forecasting System (IFS) at all lead times. Our findings suggest that effective training strategies, such as multistep fine-tuning and channel-weighted losses, significantly enhance the model’s performance. However, we also highlight that these improvements come with trade-offs in other areas, such as ensemble spread and high-frequency spatial detail. This work highlights the promise of deep learning in improving weather forecasts, which could lead to better preparedness and response to weather events, ultimately benefiting society by providing more reliable weather predictions.

Willard, Jared D. [Lawrence Berkeley National Labo↗