Search NASA⌕ Search

SEARCH · Search NASA

Results for “generative diffusion model”

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.

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

At least 91 records · Page 5
All sourcesNASA NTRSDOE OSTI
NASA NTRSDate not supplied

Machine-Learning-Based Adaptive Thinning of CrIS Radiances to Improve Global Tropical Cyclone Analysis and Forecasts

This work is focused on optimizing the assimilation of hyperspectral infrared (IR) radiances from the Cross-track Infrared Sounder (CrIS) with the goal of improving the representation of tropical cyclones (TCs) in global analyses and forecasts. Current operational assimilation systems rely on subsampling IR radiances on a regular thinning grid. A new and improved adaptive methodology based on machine learning (ML) recognizes TCs from geostationary satellite imagery and is implemented in the Goddard Earth Observing System (GEOS) model and data assimilation framework. The ML methodology is extensively trained on existing TC data sets and creates for each TC a dynamic mask, based on the evolving shape and life cycle of that specific event. Once a TC mask is created, a switch is then activated in the data assimilation system to alter the thinning, ingesting more CrIS radiances within the moving mask, thus increasing the TC sampling. After the TC dissipates, the assimilation of CrIS radiances reverts to normal data density. Results of TC segmentation provided by a state-of-the-art generative machine learning model known as the Denoising Diffusion Probabilistic Model (DDPM) are compared to the previously used U-Net model. The new approach surpasses the performance of the previously developed one. The methodology is applied to both clear-sky and cloud-cleared radiances. Benefits from the latter methodology, particularly in improving the structure of TCs and the intensity forecasts, are presented.

Oreste Reale↗
DOE OSTI2024

Deep Generative Models in Energy System Applications: Review, Challenges, and Future Directions

In recent years, with the advent of mature machine learning products like ChatGPT, Stable Diffusion, and Sora, the world has witnessed tremendous changes driven by the rapid development of generative artificial intelligence (GAI). Beyond applications in text, speech, image, and video creation, deep generative models (DGMs) underpinning these cutting-edge technologies have also been employed by domain researchers to address scientific and engineering challenges. This paper aims to fill a gap in the research community by providing a systematic review of how DGMs have been utilized in energy system applications. After introducing four most popular DGMs, we review and categorize 196 research articles into five focus areas: data generation, forecasting, situational awareness, modeling, and optimal decision-making. Through this classification, we uncover trends in how DGMs are employed for each type of problem, highlighting GAI techniques that contribute to breakthroughs over traditional methods. We discuss limitations in existing literature, engineering challenges, and propose future directions, all tailored to the unique nature of problems in energy system engineering. Our goal is to offer insights for energy system domain researchers, providing a comprehensive view of existing studies and potential future opportunities.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2024

Report on the AAPM grand challenge on deep generative modeling for learning medical image statistics

Abstract Background The findings of the 2023 AAPM Grand Challenge on Deep Generative Modeling for Learning Medical Image Statistics are reported in this Special Report. Purpose The goal of this challenge was to promote the development of deep generative models for medical imaging and to emphasize the need for their domain‐relevant assessments via the analysis of relevant image statistics. Methods As part of this Grand Challenge, a common training dataset and an evaluation procedure was developed for benchmarking deep generative models for medical image synthesis. To create the training dataset, an established 3D virtual breast phantom was adapted. The resulting dataset comprised about 108 000 images of size 512 512. For the evaluation of submissions to the Challenge, an ensemble of 10 000 DGM‐generated images from each submission was employed. The evaluation procedure consisted of two stages. In the first stage, a preliminary check for memorization and image quality (via the Fréchet Inception Distance [FID]) was performed. Submissions that passed the first stage were then evaluated for the reproducibility of image statistics corresponding to several feature families including texture, morphology, image moments, fractal statistics, and skeleton statistics. A summary measure in this feature space was employed to rank the submissions. Additional analyses of submissions was performed to assess DGM performance specific to individual feature families, the four classes in the training data, and also to identify various artifacts. Results Fifty‐eight submissions from 12 unique users were received for this Challenge. Out of these 12 submissions, 9 submissions passed the first stage of evaluation and were eligible for ranking. The top‐ranked submission employed a conditional latent diffusion model, whereas the joint runners‐up employed a generative adversarial network, followed by another network for image superresolution. In general, we observed that the overall ranking of the top 9 submissions according to our evaluation method (i) did not match the FID‐based ranking, and (ii) differed with respect to individual feature families. Another important finding from our additional analyses was that different DGMs demonstrated similar kinds of artifacts. Conclusions This Grand Challenge highlighted the need for domain‐specific evaluation to further DGM design as well as deployment. It also demonstrated that the specification of a DGM may differ depending on its intended use.

Radiology, Nuclear Medicine & Medical Imaging↗
NASA NTRS1990

A model for the distribution of material generating the soft X-ray background

The observational evidence relating to the soft X-ray diffuse background is discussed, and a simple model for its source and spatial structure is presented. In this simple model with one free parameter, the observed 1/4 keV X-ray intensity originates as thermal emission from a uniform hot plasma filling a cavity in the neutral material of the Galactic disk which contains the sun. Variations in the observed X-ray intensity are due to variations in the extent of the emission volume and therefore the emission measure of the plasma. The model reproduces the observed negative correlation between X-ray intensity and H I column density and predicts reasonable values for interstellar medium parameters.

Snowden, S. L.↗
NASA NTRS2015

A Linearized Prognostic Cloud Scheme in NASAs Goddard Earth Observing System Data Assimilation Tools

A linearized prognostic cloud scheme has been developed to accompany the linearized convection scheme recently implemented in NASA's Goddard Earth Observing System data assimilation tools. The linearization, developed from the nonlinear cloud scheme, treats cloud variables prognostically so they are subject to linearized advection, diffusion, generation, and evaporation. Four linearized cloud variables are modeled, the ice and water phases of clouds generated by large-scale condensation and, separately, by detraining convection. For each species the scheme models their sources, sublimation, evaporation, and autoconversion. Large-scale, anvil and convective species of precipitation are modeled and evaporated. The cloud scheme exhibits linearity and realistic perturbation growth, except around the generation of clouds through large-scale condensation. Discontinuities and steep gradients are widely used here and severe problems occur in the calculation of cloud fraction. For data assimilation applications this poor behavior is controlled by replacing this part of the scheme with a perturbation model. For observation impacts, where efficiency is less of a concern, a filtering is developed that examines the Jacobian. The replacement scheme is only invoked if Jacobian elements or eigenvalues violate a series of tuned constants. The linearized prognostic cloud scheme is tested by comparing the linear and nonlinear perturbation trajectories for 6-, 12-, and 24-h forecast times. The tangent linear model performs well and perturbations of clouds are well captured for the lead times of interest.

Precipitation↗
DOE OSTI2024

Effectiveness of denoising diffusion probabilistic models for fast and high-fidelity whole-event simulation in high-energy heavy-ion experiments

Artificial intelligence (AI) generative models, such as generative adversarial networks (GANs), variational autoencoders, and normalizing flows, have been widely used and studied as efficient alternatives for traditional scientific simulations. However, they have several drawbacks, including training instability and inability to cover the entire data distribution, especially for regions where data are rare. This is particularly challenging for whole-event, full-detector simulations in high-energy heavy-ion experiments, such as sPHENIX at the Relativistic Heavy Ion Collider and Large Hadron Collider experiments, where thousands of particles are produced per event and interact with the detector. This work investigates the effectiveness of denoising diffusion probabilistic models (DDPMs) as an AI-based generative surrogate model for the sPHENIX experiment that includes the heavy-ion event generation and response of the entire calorimeter stack. DDPM performance in sPHENIX simulation data is compared with a popular rival, GANs. Results show that both DDPMs and GANs can reproduce the data distribution where the examples are abundant (low-to-medium calorimeter energies). Nonetheless, DDPMs significantly outperform GANs, especially in high-energy regions where data are rare. Additionally, DDPMs exhibit superior stability compared to GANs. The results are consistent between both central and peripheral centrality heavy-ion collision events. Moreover, DDPMs offer a substantial speedup of approximately a factor of 100 compared to the traditional Geant4 simulation method.

42 ENGINEERING↗
DOE OSTI2026

Generative Thermodynamic Computing

Here, we introduce a generative modeling framework for thermodynamic computing, in which structured data are synthesized from noise by the natural time evolution of a physical system governed by Langevin dynamics. While conventional diffusion models use neural networks to perform denoising, here the information needed to generate structure from noise is encoded by the dynamics of a thermodynamic system. Training proceeds by maximizing the probability with which the computer generates the reverse of a noising trajectory, which ensures that the computer generates data with minimal heat emission. We demonstrate this framework within a digital simulation of a thermodynamic computer. If realized in analog hardware, such a system would function as a generative model that produces structured samples without the need for artificially injected noise or active control of denoising.

Whitelam, Stephen [Lawrence Berkeley National Labo↗
DOE OSTI2025

First-Principles Studies of Tritium Species Diffusivity Across the Interfaces of Ni-NiZr Alloy-Zircaloy-4

A tritium producing assembly, also known as tritium producing burnable absorber rods (TPBARs), consists of a metal getter tube located between the cladding and γ-LiAlO2 pellets. The metal getter tube is composed of a nickel (Ni) layer coated on Zircaloy-4. The getter assembly is used to capture tritium (3H) species (mainly 3H2 and 3H2O) generated from γ-LiAlO2 pellets during irradiation. Exploring 3H species (3H2, 3H2O) dissociation on the Ni surface and diffusion in the Ni layer and across the interface of Ni-plated Zircaloy-4 getters can provide insights on tritium transport and retention in the pellets and the getter materials. . In FY25, based on the ideal interface model of Ni-Zircaloy-4 generated in FY24, we will further explore the diffusion pathways of 3H across such Ni-ZrNi-alloy layer-Zircalory-4 interfaces under different conditions, including with oxide/hydroxide clusters on the Ni layer, and impurities located in ZrNi-alloy and Zircalory-4 layers. Due to the size limitation of DFT simulation, we will separate the 3-layer system into 3 sub-systems: Ni-ZrNi interface, ZrNi alloy layer, ZrNi-alloy-Zircaloy-4 interface, and simulate the diffusion barriers for tritium.

diffusion barrier↗
NASA NTRS1995

Stream interfaces and energetic ions 2: Ulysses test of Pioneer results

Ulysses measurements of energetic and solar wind particles taken near 5 AU between 20 and 30 degrees south latitude during a well-developed recurring corotating interaction region (CIR) show that the CIR's corotating energetic ion population (CEIP) associated with the trailing reverse shock starts within the CIR at the stream interface. This is consistent with an earlier result obtained by Pioneers 10 and 11 in the ecliptic plane between 4 and 6 AU. The Ulysses/Pioneer finding noteworthy since the stream interface is not magnetically connected to the reverse shock but lies 12-17 corotation hours from it. Thus, the finding to be inconsistent with the basic model that generates CEIP particles at the reverse shock and propagates them along field lines Eliminating the inconsistency probably entails an extension of the standard model. We consider two possible extensions cross-field diffusion and energetic particles generation closer to the sun in the gap between the stream interface and the reverse shock.

Intriligator, Devrie S.↗
NASA NTRS1976

Wind Tunnel Design Studies and Technical Evaluation of Advanced Cargo Aircraft Concepts

In support of aerodynamic studies relating to the design and performance prediction of the National Transonic Facility, the following main tasks were accomplished: (1) estimation of aerodynamic losses of the tunnel circuits, (2) refinement of the high-speed diffuser loss prediction method utilizing experimental data generated for the purpose; (3) model studies of flow in the second-turn and measurements of the fan inlet distortion and overall pressure loss; (4) development of a shortened fan nacelle configuration of improved aerodynamic performance; and (5) evolution through model studies of an efficient rapid-diffuser system as the key to a circuit-modification proposal to reduce volume and minimize liquid-nitrogen consumption, at the same time saving on the shell cost.

Rao, D. M.↗
DOE OSTI2025

High‐Asymmetry Metasurface: A New Solution for Terahertz Resonance via Active Learning‐Augmented Diffusion Model

Terahertz (THz) metamaterials with high‐figure‐of‐merit (high‐FoM) performance resonance are essential for advancing sensors, detectors, and imagers. Conventional designs focus on symmetric or low‐asymmetry geometric structures, leaving high‐asymmetry designs largely unexplored due to the inefficiency of trial‐and‐error‐based rational design. Recent deep learning techniques offer automation and acceleration but are constrained by the need for large datasets inherent to their data‐driven nature. Here, a novel prior knowledge‐guided generative model augmented by a physics‐constrained active learning mechanism to design high‐asymmetry metamaterials. An advanced diffusion model learns features from a small set of classical structures with high‐FoM THz resonance and generates new high‐asymmetry structures. To mitigate the limited number of classical structures, the generated high‐asymmetry structures are actively selected and integrated into the initial training dataset based on their physical characteristics. Experimental results demonstrate the superior resonance performance of the generated high‐asymmetry metamaterials over classical designs, exhibiting improvements exceeding 30% in key resonance metrics. Remarkably, this performance is attained using only 68 classical structures as the initial training dataset, significantly reducing the data requirements for deep learning‐based metamaterial design. The proposed scheme for generating high‐asymmetry structures provides a new effective and efficient solution for high‐FoM resonance, expanding applications in high‐sensitivity THz metadevices.

diffusion model↗
NASA NTRS1974

Four-D global reference atmosphere technical description, part 1

An empirical atmospheric model was developed which generates values for pressure, density, temperature, and winds from surface levels to orbital altitudes. The output parameters consist of components for: (1) latitude, longitude, and altitude dependent monthly and annual means; (2) quasi-biennial oscillations; and (3) random perturbations to simulate partially the variability due to synoptic, diurnal, planetary wave, and gravity wave variations. Quasi-biennial and random variation perturbations are computed from parameters determined from various empirical studies and are added to the monthly mean values. This model has been developed as a computer program called PROFILE which can be used to generate altitude profiles of atmospheric parameters along any simulated trajectory through the atmosphere. The PROFILE program was developed for design applications in the space shuttle program. Other applications of the model are discussed, such as for global circulation and diffusion studies, and for generating profiles for comparison with other atmospheric measurement techniques, (e.g. satellite measured temperature profiles).

Justus, C. G.↗
NASA NTRS2003

Modeling Charge Collection in Detector Arrays

A detector array charge collection model has been developed for use as an engineering tool to aid in the design of optical sensor missions for operation in the space radiation environment. This model is an enhancement of the prototype array charge collection model that was developed for the Next Generation Space Telescope (NGST) program. The primary enhancements were accounting for drift-assisted diffusion by Monte Carlo modeling techniques and implementing the modeling approaches in a windows-based code. The modeling is concerned with integrated charge collection within discrete pixels in the focal plane array (FPA), with high fidelity spatial resolution. It is applicable to all detector geometries including monolithc charge coupled devices (CCDs), Active Pixel Sensors (APS) and hybrid FPA geometries based on a detector array bump-bonded to a readout integrated circuit (ROIC).

Hardage, Donna↗
NASA NTRS1983

Numerical experiments with a stochastic zonal climate model

A zonally averaged energy balance climate model is developed to generate zonal temperature variability through fluctuating meridional energy transports. Stochastic transport fluctuations are included in the model by multiplying the eddy diffusion coefficients by Gaussian random deviates. For a variability of eddy coefficients of 50 percent, the model is found to generate an interannual temperature variability of 0.03 K for the global temperature, and 0.04 and 0.05 K for the Northern and Southern Hemispheres, respectively. It is shown that the temperature variability level generated is linearly related to the transport variability level introduced. An increase in the level of model-generated temperature variability and a change in the shape of variance spectra of temperature anomaly time series are obtained by switching from the multiplicative noise model to an additive noise model. The results of these model studies are compared with a time series of central England temperatures as well as GCM generated climate variability.

Schneider, S. H.↗
DOE OSTI2026

Conditional diffusion machine-learning framework for mapping valence electron distribution from convergent beam electron diffraction

Quantitative convergent beam electron diffraction (CBED) enables determination of aspherical valence electron distributions through refinement of low-order structure factors, which are highly sensitive to chemical bonding and charge density variations. However, conventional quantitative CBED (QCBED) requires solving a highly nonlinear inverse problem with many coupled parameters, and computationally intensive dynamical diffraction calculations, making it time-consuming and difficult to apply to complex systems. More broadly, reconstructing charge density and orbital electron distribution from diffraction data has long been a central challenge in both x-ray and electron crystallography. Here, in this study, we introduce an artificial-intelligence (AI)-based framework that replaces traditional refinement with a data-driven inverse solver. Using a large synthetic CBED dataset generated by Bloch-wave simulations, we train a conditional diffusion model to directly infer crystal structural parameters and multipole density formalism parameters, and hence valence electron distributions, from CBED patterns alone. By learning from forward simulations across realistic parameter space, the model effectively solves the inverse problem. Compared with direct regression approaches, the diffusion-based framework provides posterior parameter distributions for rigorous uncertainty quantification while preserving quantitative fidelity and reducing analysis time by orders of magnitude. By eliminating the need for external single-crystal x-ray diffraction data and complex nonlinear refinement, this approach enables practical, high-throughput, and in situ quantitative CBED, enabling real-time mapping of valence electron distributions and their correlation with functional responses in quantum and energy materials.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
NASA NTRS1996

Diffusion lengths in irradiated N/P InP-on-Si solar cells

Indium phosphide (InP) solar cells were made on silicon (Si) wafers (InP/Si) by to take advantage of both the radiation-hardness properties of the InP solar cell and the light weight and low cost of Si wafers. The InP/Si cell application is for long duration and/or high radiation orbit space missions. Spire has made N/P InP/Si cells of sizes up to 2 cm by 4 cm with beginning-of-life (BOL) AM0 efficiencies over 13% (one-sun, 28C). These InP/Si cells have higher absolute efficiency and power density after a high radiation dose than gallium arsenide (GaAs) or silicon (Si) solar cells after a fluence of about 2e15 1 MeV electrons/sq. cm. In this work, we investigate the minority carrier (electron) base diffusion lengths in the N/P InP/Si cells. A quantum efficiency model was constructed for a 12% BOL AM0 N/P InP/Si cell which agreed well with the absolutely measured quantum efficiency and the sun-simulator measured AM0 photocurrent (30.1 mA/sq. cm). This model was then used to generate a table of AM0 photocurrents for a range of base diffusion lengths. AM0 photocurrents were then measured for irradiations up to 7.7e16 1 MeV electrons/sq. cm (the 12% BOL cell was 8% after the final irradiation). By comparing the measured photocurrents with the predicted photocurrents, base diffusion lengths were assigned at each fluence level. A damage coefficient K of 4e-8 and a starting (unirradiated) base electron diffusion length of 0.8 microns fits the data well. The quantum efficiency was measured again at the end of the experiment to verify that the photocurrent predicted by the model (25.5 mA/sq. cm) agreed with the simulator-measured photocurrent after irradiation (25.7 mA/sq. cm).

Wojtczuk, Steven↗
DOE OSTI2026

D$e^+e^-$ffusion: Capturing the Beam-Beam Physics of $e^+e^-$ Collisions with Diffusion Models

Beam-induced backgrounds at high-luminosity $e^+e^-$ colliders, such as the FCC-ee, are dominated by incoherent pair creation (IPC), and require computationally expensive simulations with dedicated Monte Carlo (MC) event generators. Reliable detector and machine-detector interface studies necessitate event samples that are several orders of magnitude larger than what is practically attainable with existing MC. To alleviate this bottleneck, we present D$e^+e^-$ffusion, a denoising diffusion probabilistic model that operates as a permutation-equivariant, set-valued surrogate for fast IPC simulation. Trained on a small GuineaPig++ sample, D$e^+e^-$ffusion faithfully reproduces the marginal and joint kinematic, angular, and positional distributions of all three IPC production processes. In addition, we assess the fidelity at the detector level by propagating both Geant4 and D$e^+e^-$ffusion events through a Geant4 simulation of the CLD vertex detector and by training a transformer-based two-sample classifier; the classifier achieves an area under the ROC curve of $0.553 \pm 0.016$. The trained model generates events nearly four orders of magnitude faster than Geant4, paving the way for a fast-simulation surrogate for FCC-ee design studies.

Chahine, Antonio [Imperial Coll., London]↗