SEARCH · Engineering Papers

Results for “Model Inference”

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 793 records · Page 44

All sources NASA NTRS DOE OSTI

DOE OSTI2021

On the Stochastic Stability of Deep Markov Models

Deep Markov models (DMM) are generative models which are scalable and expressive generalization of Markov models for representation, learning, and inference problems. However, the fundamental stochastic stability guarantees of such models have not been thoroughly investigated. In this paper, we present a novel stability analysis method and provide sufficient conditions of DMM's stochastic stability. The proposed stability analysis is based on the contraction of probabilistic maps modeled by deep neural networks. We make connections between the spectral properties of neural network's weights and different types of used activation function on the stability and overall dynamic behavior of DMMs with Gaussian distributions. Based on the theory, we propose a few practical methods for designing constrained DMMs with guaranteed stability. We empirically substantiate our theoretical results via intuitive numerical experiments using the proposed stability constraints.

Drgona, Jan↗

DOE OSTI2026

Learning turbulent flows with generative models for super resolution and sparse flow reconstruction

Neural operators are promising surrogates for dynamical systems but when trained with standard L 2 losses they tend to oversmooth fine-scale turbulent structures. Here, we show that combining operator learning with generative modeling overcomes this limitation. We consider three practical turbulent-flow challenges where conventional neural operators fail: spatio-temporal super-resolution, forecasting, and sparse flow reconstruction. For Schlieren jet super-resolution, an adversarially trained neural operator (adv-NO) reduces the energy-spectrum error by 15 × while preserving sharp gradients at neural operator-like inference cost. For 3D homogeneous isotropic turbulence, adv-NO trained on only 160 timesteps from a single trajectory forecasts accurately for five eddy-turnover times and offers 114 × wall-clock speed-up at inference than the baseline diffusion-based forecasters, enabling near-real-time rollouts. For reconstructing cylinder wake flows from highly sparse Particle Tracking Velocimetry-like inputs, a conditional generative model infers full 3D velocity and pressure fields with correct phase alignment and statistics. These advances enable accurate reconstruction and forecasting at low compute cost, bringing near-real-time analysis and control within reach in experimental and computational fluid mechanics.

Fluid dynamics↗

NASA NTRS1993

How well was total ozone abundance inferred with Mariner 9

Ozone is a key to understanding atmospheric chemistry on Mars. Over 20 photochemical models of the Martian atmosphere have been published, and O3 is often used as a benchmark for these models. O3 abundance has been inferred from instrumentation on several spacecraft, with the most complete coverage provided by Mariner 9. A radiative transfer computer model is used to re-examine the Mariner 9 UV spectra.

Lindner, B. L.↗

DOE OSTI2024

Constraining solar electron number density via neutrino flavor data at Borexino

Understanding the physics of the deep solar interior, and the more exotic environs of core-collapse supernovae (CCSN) and binary neutron-star (NS) mergers, is of keen interest in many avenues of research. To date, this physics is based largely on simulations via forward integration. While these simulations provide valuable constraints, it could be insightful to adopt the "inverse approach" as a point of comparison. Within this paradigm, parameters of the solar interior are not output based on an assumed model, but rather are inferred based on real data. Here, we take the specific case of solar electron number density, which historically is taken as output from the standard solar model. We show how one may arrive at an independent constraint on that density profile based on available neutrino flavor data from the Earth-based Borexino experiment. The inference technique's ability to offer a unique lens on physics can be extended to other datasets, and to analogous questions for CCSN and NS mergers, albeit with simulated data.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

NASA NTRS1989

Inference of geophysical parameters from multifrequency polarimetric radar observations and model inversion

In summer 1988 an intensive field experiment was conducted in the vicinity of the Pisgah lava flow in the Mojave Desert. Physical properties such as microtopography, composition, soil moisture, and dielectric constant were measured at five sites representing surfaces with rms heights varying from less than one centimeter to tens of centimeters. In addition, polarimetric radar images at P-band, L-band, and C-band were acquired at three different incidence angles with the NASA/JPL airborne imaging radar polarimeter. Using trihedral corner reflectors deployed in the area prior to imaging, the L- and C-band images were calibrated to provide sigma0 values for each resolution element in the scene. The results of inferring geophysical parameters such as rms surface height and correlation length of the measured surfaces by fitting the observed signatures with those predicted by the small-perturbation model are presented: for smoother surfaces, the rms height values inferred are in good agreement with in situ measurements.

Van Zyl, J. J.↗

DOE OSTI2024

A deep learning approach to fast analysis of collective Thomson scattering spectra

Fast analysis of collective Thomson scattering ion acoustic wave features using a deep convolutional neural network model is presented. The network was trained from spectra to predict the plasma parameters, including ion velocities, population fractions, and ion and electron temperatures. A fully kinetic particle-in-cell simulation was used to model a laboratory astrophysics experiment and simulate a diagnostic image of the ion acoustic wave feature. Network predictions were compared with Bayesian inference of the plasma model parameters for both the simulated and experimentally measured images. Both approaches were fairly accurate predicting the simulated image and the network predictions matched a good portion of the Bayesian results for the experimentally measured image. The Bayesian approach is more robust to noise and motivates future work to train deep learning models with realistic noise. The advantage of the deep learning model is making thousands of predictions in a few hundred milliseconds, compared to a few seconds to minutes per prediction for the optimization and Bayesian approaches presented here. The results demonstrate promising capabilities of deep learning models to analyze Thomson data orders of magnitude faster than conventional methods when using the neural network for standalone analysis. If more rigorous analysis is needed, neural network predictions can be used to quickly initialize other optimization methods and increase chances of success. This is especially useful when the dataset becomes very large or highly dimensional and manually refining initial conditions for the entire dataset are no longer tractable.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

NASA NTRS2015

Hybrid Glacier Inventory, Gravimetry and Altimetry (HIGA) Mass Balance Product for Greenland and the Canadian Arctic

We present a novel inversion algorithm that generates a mass balance field that is simultaneously consistent with independent observations of glacier inventory derived from optical imagery, cryosphere-attributed mass trends derived from satellite gravimetry, and ice surface elevation trends derived from airborne and satellite altimetry. We use this algorithm to assess mass balance across Greenland and the Canadian Arctic over the Sep-2003 to Oct- 2009 period at 26 kilometers resolution. We evaluate local algorithm-inferred mass balance against forty in situ point observations. This evaluation yields a root mean squared error (RMSE) of 0.15 mWE/a ( 0.15 meters (water equivalent) per annum), and highlights a paucity of in situ observations from regions of high dynamic mass loss and peripheral glaciers. We assess mass losses of 212 plus or minus 67 Gigatons per annum to the Greenland ice sheet proper, 38 Gigatons per annum to peripheral glaciers in Greenland, and 42 Gigatons per annum to glaciers in the Canadian Arctic. These magnitudes of mass loss are dependent on the gravimetry-derived spherical harmonic mass trend we invert. We spatially partition the transient glacier continuity equation by differencing algorithm inferred mass balance from modeled surface mass balance, in order to solve the horizontal divergence of ice flux as a residual. This residual ice dynamic field infers flux divergence (or submergent flow) in the ice sheet accumulation area and at tidewater margins, and flux convergence (or emergent flow) in land-terminating ablation areas, which is consistent with continuum mechanics theory.

Altimetry↗

DOE OSTI2021

Bayesian Inference for Integrating Yarrowia lipolytica Multiomics Datasets with Metabolic Modeling

Optimizing the metabolism of microbial cell factories for yields and titers is a critical step for economically viable production of bioproducts and biofuels. In this process, tuning the expression of individual enzymes to obtain the desired pathway flux is a challenging step, in which data from separate multiomics techniques must be integrated with existing biological knowledge to determine where changes should be made. Following a design-build-test-learn strategy, building on recent advances in Bayesian metabolic control analysis, we identify in this work key enzymes in the oleaginous yeast Yarrowia lipolytica that correlate with the production of itaconate by integrating a metabolic model with multiomics measurements. To this extent, we quantify the uncertainty for a variety of key parameters, known as flux control coefficients (FCCs), needed to improve the bioproduction of target metabolites and statistically obtain key correlations between the measured enzymes and boundary flux. Based on the top five significant FCCs and five correlated enzymes, our results show phosphoglycerate mutase, acetyl-CoA synthetase (ACSm), carbonic anhydrase (HCO3E), pyrophosphatase (PPAm), and homoserine dehydrogenase (HSDxi) enzymes in rate-limiting reactions that can lead to increased itaconic acid production.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2026

Multi-Modal Bayesian Neural Network Surrogates with Conjugate Last-Layer Estimation

As data collection and simulation capabilities advance, multi-modal learning, the task of learning from multiple modalities and sources of data, is becoming an increasingly important area of research. Surrogate models that learn from data of multiple auxiliary modalities to support the modeling of a highly expensive quantity of interest have the potential to aid outer loop applications such as optimization, inverse problems, or sensitivity analyses when multi-modal data are available. We develop two multi-modal Bayesian neural network surrogate models and leverage conditionally conjugate distributions in the last layer to estimate model parameters using stochastic variational inference (SVI). We provide a method to perform this conjugate SVI estimation in the presence of partially missing observations. Here, we demonstrate improved prediction accuracy and uncertainty quantification compared to unimodal surrogate models for both scalar and time series data.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS2004

InvestigationOrganizer: The Development and Testing of a Web-based Tool to Support Mishap Investigations

InvestigationOrganizer (IO) is a collaborative web-based system designed to support the conduct of mishap investigations. IO provides a common repository for a wide range of mishap related information, and allows investigators to make explicit, shared, and meaningful links between evidence, causal models, findings and recommendations. It integrates the functionality of a database, a common document repository, a semantic knowledge network, a rule-based inference engine, and causal modeling and visualization. Thus far, IO has been used to support four mishap investigations within NASA, ranging from a small property damage case to the loss of the Space Shuttle Columbia. This paper describes how the functionality of IO supports mishap investigations and the lessons learned from the experience of supporting two of the NASA mishap investigations: the Columbia Accident Investigation and the CONTOUR Loss Investigation.

Carvalho, Robert F.↗

DOE OSTI2026

Constraining Hamiltonians from chiral effective field theory with neutron-star data

Multi-messenger observations of neutron stars (NSs) and their mergers have placed strong constraints on the dense-matter equation of state (EOS). The EOS, in turn, depends on microscopic nuclear interactions that are described by nuclear Hamiltonians. These Hamiltonians are commonly derived within chiral effective field theory (EFT). Ideally, multi-messenger observations of NSs could be used to directly inform our understanding of EFT interactions, but such a direct inference necessitates millions of model evaluations. This is computationally prohibitive because each evaluation requires us to calculate the EOS from a Hamiltonian by solving the quantum many-body problem with methods such as auxiliary-field diffusion Monte Carlo (AFDMC), which provides very accurate and precise solutions but at a significant computational cost. Additionally, we need to solve the stellar structure equations for each EOS which further slows down each model evaluation by a few seconds. In this work, we combine emulators for AFDMC calculations of neutron matter, built using parametric matrix models, and for the stellar structure equations, built using multilayer perceptron neural networks, with the PyCBC data-analysis framework to enable a direct inference of coupling constants in an EFT Hamiltonian using multi-messenger observations of NSs. We find that astrophysical data can provide informative constraints on two-nucleon couplings despite the high densities probed in NS interiors.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2018

Correlated prompt fission data in transport simulations

Detailed information on the fission process can be inferred from the observation, modeling and theoretical understanding of prompt fission neutron and γ-ray observables. Beyond simple average quantities, the study of distributions and correlations in prompt data, e.g., multiplicity-dependent neutron and γ-ray spectra, angular distributions of the emitted particles, n -n, n - γ, and γ - γ correlations, can place stringent constraints on fission models and parameters that would otherwise be free to be tuned separately to represent individual fission observables. The FREYA and CGMF codes have been developed to follow the sequential emissions of prompt neutrons and γ rays from the initial excited fission fragments produced right after scission. Both codes implement Monte Carlo techniques to sample initial fission fragment configurations in mass, charge and kinetic energy and sample probabilities of neutron and γ emission at each stage of the decay. This approach naturally leads to using simple but powerful statistical techniques to infer distributions and correlations among many observables and model parameters. The comparison of model calculations with experimental data provides a rich arena for testing various nuclear physics models such as those related to the nuclear structure and level densities of neutron-rich nuclei, the γ-ray strength functions of dipole and quadrupole transitions, the mechanism for dividing the excitation energy between the two nascent fragments near scission, and the mechanisms behind the production of angular momentum in the fragments, etc. Beyond the obvious interest from a fundamental physics point of view, such studies are also important for addressing data needs in various nuclear applications. The inclusion of the FREYA and CGMF codes into the MCNP6.2 and MCNPX - PoliMi transport codes, for instance, provides a new and powerful tool to simulate correlated fission events in neutron transport calculations important in nonproliferation, safeguards, nuclear energy, and defense programs. Here, this review provides an overview of the topic, starting from theoretical considerations of the fission process, with a focus on correlated signatures. It then explores the status of experimental correlated fission data and current efforts to address some of the known shortcomings. Numerical simulations employing the FREYA and CGMF codes are compared to experimental data for a wide range of correlated fission quantities. The inclusion of those codes into the MCNP6.2 and MCNPX - PoliMi transport codes is described and discussed in the context of relevant applications. The accuracy of the model predictions and their sensitivity to model assumptions and input parameters are discussed. Lastly, a series of important experimental and theoretical questions that remain unanswered are presented, suggesting a renewed effort to address these shortcomings.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2024

Rapid assessment of the creep rupture life of metals: A model enabling experimental design

Prediction of the creep rupture life of engineering metals is critical for qualification and design of new materials. The use of long-term creep tests and the need to quantify the performance variability in a priori similar systems hinder the rapid creep assessment of a given material. Therefore, it is essential to develop methods that can extrapolate the long-term performance of alloys and the associated variability from short-term experiments. To this end, this study introduces a new model which enables the estimation of the rupture life of a material for a given stress and temperature. This model relies on two components. First, a new relation for the minimum creep rate (MCR) of materials is introduced. It includes a stress dependent stress exponent allowing the model to capture the variation of MCR across a wide range of temperatures and stresses. Second, employing the Monkman-Grant (MG) law, we establish a relation between stress, temperature and creep rupture life. Together, these two elements yield a new closed-form mathematical expression for the Larson Miller parameter as a function of stress and temperature. This expression captures the creep rupture time for many metals (Gr91, Copper, Gr122 and 347H) and compares favorably with alternate empirical approaches. The model is then used to assess the minimum duration of creep rates necessary to qualify the material up to 100000h. Furthermore, it is found that depending on the material system, creep tests as few as five limited to 5000 h for steels (Gr91, Gr122, 347H) and 100 h for copper are sufficient to model creep lifetimes. Finally, using a Bayesian inference-based approach to calibrate the model, we demonstrate that variability in rupture life can be captured via the quantification of the uncertainty in the model parameters and extrapolated from a limited number of short to moderately short creep tests; thereby paving the way for accelerated creep testing.

36 MATERIALS SCIENCE↗

DOE OSTI2017

Protein structure determination using metagenome sequence data

Filling in the protein fold picture Fewer than a third of the 14,849 known protein families have at least one member with an experimentally determined structure. This leaves more than 5000 protein families with no structural information. Protein modeling using residue-residue contacts inferred from evolutionary data has been successful in modeling unknown structures, but it requires large numbers of aligned sequences. Ovchinnikov et al. augmented such sequence alignments with metagenome sequence data (see the Perspective by Söding). They determined the number of sequences required to allow modeling, developed criteria for model quality, and, where possible, improved modeling by matching predicted contacts to known structures. Their method predicted quality structural models for 614 protein families, of which about 140 represent newly discovered protein folds. Science , this issue p. 294 ; see also p. 248

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2016

Proposal of a Novel Approach to Developing Material Models for Micro-scale Composites Based on Testing and Modeling of Macro-scale Composites

The Los Alamos National Laboratory's Weapon Systems Engineering division's Advanced Engineering Analysis group employs material constitutive models of composites for use in simulations of components and assemblies of interest. Experimental characterization, modeling and prediction of the macro-scale (i.e. continuum) behaviors of these composite materials is generally difficult because they exhibit nonlinear behaviors on the meso- (e.g. micro-) and macro-scales. Furthermore, it can be difficult to measure and model the mechanical responses of the individual constituents and constituent interactions in the composites of interest. Current efforts to model such composite materials rely on semi-empirical models in which meso-scale properties are inferred from continuum level testing and modeling. The proposed approach involves removing the difficulties of interrogating and characterizing micro-scale behaviors by scaling-up the problem to work with macro-scale composites, with the intention of developing testing and modeling capabilities that will be applicable to the mesoscale. This approach assumes that the physical mechanisms governing the responses of the composites on the meso-scale are reproducible on the macro-scale. Working on the macro-scale simplifies the quantification of composite constituents and constituent interactions so that efforts can be focused on developing material models and the testing techniques needed for calibration and validation. Other benefits to working with macro-scale composites include the ability to engineer and manufacture—potentially using additive manufacturing techniques—composites that will support the application of advanced measurement techniques such as digital volume correlation and three-dimensional computed tomography imaging, which would aid in observing and quantifying complex behaviors that are exhibited in the macro-scale composites of interest. Ultimately, the goal of this new approach is to develop a meso-scale composite modeling framework, applicable to many composite materials, and the corresponding macroscale testing and test data interrogation techniques to support model calibration.

36 MATERIALS SCIENCE↗

NASA NTRS1995

Estimating the urban bias of surface shelter temperatures using upper-air and satellite data. Part 1: Development of models predicting surface shelter temperatures

Multiple regression techniques were used to predict surface shelter temperatures based on the time period 1986-89 using upper-air data from the European Centre for Medium-Range Weather Forecasts (ECMWF) to represent the background climate and site-specific data to represent the local landscape. Global monthly mean temperature models were developed using data from over 5000 stations available in the Global Historical Climate Network (GHCN). Monthly maximum, mean, and minimum temperature models for the United States were also developed using data from over 1000 stations available in the U.S. Cooperative (COOP) Network and comparative monthly mean temperature models were developed using over 1150 U.S. stations in the GHCN. Three-, six-, and full-variable models were developed for comparative purposes. Inferences about the variables selected for the various models were easier for the GHCN models, which displayed month-to-month consistency in which variables were selected, than for the COOP models, which were assigned a different list of variables for nearly every month. These and other results suggest that global calibration is preferred because data from the global spectrum of physical processes that control surface temperatures are incorporated in a global model. All of the models that were developed in this study validated relatively well, especially the global models. Recalibration of the models with validation data resulted in only slightly poorer regression statistics, indicating that the calibration list of variables was valid. Predictions using data from the validation dataset in the calibrated equation were better for the GHCN models, and the globally calibrated GHCN models generally provided better U.S. predictions than the U.S.-calibrated COOP models. Overall, the GHCN and COOP models explained approximately 64%-95% of the total variance of surface shelter temperatures, depending on the month and the number of model variables. In addition, root-mean-square errors (rmse's) were over 3 C for GHCN models and over 2 C for COOP models for winter months, and near 2 C for GHCN models and near 1.5 C for COOP models for summer months.

Epperson, David L.↗

NASA NTRS2016

Uncertainty Reduction using Bayesian Inference and Sensitivity Analysis: A Sequential Approach to the NASA Langley Uncertainty Quantification Challenge

This paper presents a computational framework for uncertainty characterization and propagation, and sensitivity analysis under the presence of aleatory and epistemic un- certainty, and develops a rigorous methodology for efficient refinement of epistemic un- certainty by identifying important epistemic variables that significantly affect the overall performance of an engineering system. The proposed methodology is illustrated using the NASA Langley Uncertainty Quantification Challenge (NASA-LUQC) problem that deals with uncertainty analysis of a generic transport model (GTM). First, Bayesian inference is used to infer subsystem-level epistemic quantities using the subsystem-level model and corresponding data. Second, tools of variance-based global sensitivity analysis are used to identify four important epistemic variables (this limitation specified in the NASA-LUQC is reflective of practical engineering situations where not all epistemic variables can be refined due to time/budget constraints) that significantly affect system-level performance. The most significant contribution of this paper is the development of the sequential refine- ment methodology, where epistemic variables for refinement are not identified all-at-once. Instead, only one variable is first identified, and then, Bayesian inference and global sensi- tivity calculations are repeated to identify the next important variable. This procedure is continued until all 4 variables are identified and the refinement in the system-level perfor- mance is computed. The advantages of the proposed sequential refinement methodology over the all-at-once uncertainty refinement approach are explained, and then applied to the NASA Langley Uncertainty Quantification Challenge problem.

Uncertainty↗

DOE OSTI2017

Carbon cycle confidence and uncertainty: Exploring variation among soil biogeochemical models

Emerging insights into factors responsible for soil organic matter stabilization and decomposition are being applied in a variety of contexts, but new tools are needed to facilitate the understanding, evaluation, and improvement of soil biogeochemical theory and models at regional to global scales. To isolate the effects of model structural uncertainty on the global distribution of soil carbon stocks and turnover times we developed a soil biogeochemical testbed that forces three different soil models with consistent climate and plant productivity inputs. The models tested here include a first-order, microbial implicit approach (CASA-CNP), and two recently developed microbially explicit models that can be run at global scales (MIMICS and CORPSE). When forced with common environmental drivers, the soil models generated similar estimates of initial soil carbon stocks (roughly 1,400 Pg C globally, 0–100 cm), but each model shows a different functional relationship between mean annual temperature and inferred turnover times. Subsequently, the models made divergent projections about the fate of these soil carbon stocks over the 20th century, with models either gaining or losing over 20 Pg C globally between 1901 and 2010. Single-forcing experiments with changed inputs, temperature, and moisture suggest that uncertainty associated with freeze-thaw processes as well as soil textural effects on soil carbon stabilization were larger than direct temperature uncertainties among models. Finally, the models generated distinct projections about the timing and magnitude of seasonal heterotrophic respiration rates, again reflecting structural uncertainties that were related to environmental sensitivities and assumptions about physicochemical stabilization of soil organic matter. Here, by providing a computationally tractable and numerically consistent framework to evaluate models we aim to better understand uncertainties among models and generate insights about factors regulating the turnover of soil organic matter.

54 ENVIRONMENTAL SCIENCES↗