Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Generative models”

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 307 records · Page 17
All sourcesNASA NTRSDOE OSTI
NASA NTRS2002

A Comparison of Model- and Satellite-Derived Aerosol Optical Depth and Reflectivity

The determination of an accurate quantitative understanding of the role of tropospheric aerosols in the earth's radiation budget is extremely important because forcing by anthropogenic aerosols presently represents one of the most uncertain aspects of climate models. Here the authors present a systematic comparison of three different analyses of satellite-retrieved aerosol optical depth based on the Advanced Very High Resolution Radiometer (AVHRR)- measured radiances with optical depths derived from six different models. Also compared are the model-derived clear-sky reflected shortwave radiation with satellite-measured reflectivities derived from the Earth Radiation Budget Experiment (ERBE) satellite. The three different satellite-derived optical depths differ by between -0.10 and 0.07 optical depth units in comparison to the average of the three analyses depending on latitude and month, but the general features of the retrievals are similar. The models differ by between -0.09 and +0.16 optical depth units from the average of the models. Differences between the average of the models and the average of the satellite analyses range over -0.11 to +0.05 optical depth units. These differences are significant since the annual average clear-sky radiative forcing associated with the difference between the average of the models and the average of the satellite analyses ranges between -3.9 and 0.7 W m(exp -2) depending on latitude and is -1.7 W m (exp -2) on a global average annual basis. Variations in the source strengths of dimethylsulfide (DMS)-derived aerosols and sea salt aerosols can explain differences between the models, and between the models and satellite retrievals of up to 0.2 optical depth units. The comparison of model-generated reflected shortwave radiation and ERBE-measured shortwave radiation is similar in character as a function of latitude to the analysis of modeled and satellite-retrieved optical depths, but the differences between the modeled clear-sky reflected flux and the ERBE clear-sky reflected flux is generally larger than that inferred from the difference between the models and the AVHRR optical depths, especially at high latitudes. The difference between the mean of the models and the ERBE-analyzed clear-sky flux is 1.6 W m(exp -2). The overall comparison indicates that the model-generated aerosol optical depth is systematically lower than that inferred from measurements between the latitudes of 10 and 30 deg S. It is not likely that the shortfall is due to small values of the sea salt optical depth because increases in this component would create modeled optical depths that are larger than those from satellites in the region north of 30 deg N and near 50 deg S. Instead, the source strengths for DMS and biomass aerosols in the models may be too low. Firm conclusions, however, will require better retrieval procedures for the satellites, including better cloud screening procedures, further improvement of the model's treatment of aerosol transport and removal, and a better determination of aerosol source strengths.

Penner, Joyce E.↗
NASA NTRS1998

Comparison of Computational, Model and Experimental, Example Trained Neural Networks for Processing Speckled Fringe Patterns

The responses of artificial neural networks to experimental and model-generated inputs are compared for detection of damage in twisted fan blades using electronic holography. The training-set inputs, for this work, are experimentally generated characteristic patterns of the vibrating blades. The outputs are damage-flag indicators or second derivatives of the sensitivity-vector-projected displacement vectors from a finite element model. Artificial neural networks have been trained in the past with computational-model- generated training sets. This approach avoids the difficult inverse calculations traditionally used to compare interference fringes with the models. But the high modeling standards are hard to achieve, even with fan-blade finite-element models.

Decker, A. J.↗
NASA NTRS1998

Comparison of Computational-Model and Experimental-Example Trained Neural Networks for Processing Speckled Fringe Patterns

The responses of artificial neural networks to experimental and model-generated inputs are compared for detection of damage in twisted fan blades using electronic holography. The training-set inputs, for this work, are experimentally generated characteristic patterns of the vibrating blades. The outputs are damage-flag indicators or second derivatives of the sensitivity-vector-projected displacement vectors from a finite element model. Artificial neural networks have been trained in the past with computational-model-generated training sets. This approach avoids the difficult inverse calculations traditionally used to compare interference fringes with the models. But the high modeling standards are hard to achieve, even with fan-blade finite-element models.

Decker, A. J.↗
DOE OSTI2024

CoarsenConf: Equivariant Coarsening with Aggregated Attention for Molecular Conformer Generation

Molecular conformer generation (MCG) is an important task in cheminformatics and drug discovery. The ability to efficiently generate low-energy 3D structures can avoid expensive quantum mechanical simulations, leading to accelerated virtual screenings and enhanced structural exploration. Several generative models have been developed for MCG, but many struggle to consistently produce high-quality conformers for meaningful downstream applications. To address these issues, we introduce CoarsenConf, which coarse-grains molecular graphs based on torsional angles and integrates them into an SE(3)-equivariant hierarchical variational autoencoder. Through equivariant coarse-graining, we aggregate the fine-grained atomic coordinates of subgraphs connected via rotatable bonds, creating a variable-length coarse-grained latent representation. Our model uses a novel aggregated attention mechanism to restore fine-grained coordinates from the coarse-grained latent representation, enabling efficient generation of accurate conformers. Furthermore, we evaluate the chemical and biochemical quality of our generated conformers on multiple downstream applications, including property prediction and large-scale oracle-based protein docking. Overall, CoarsenConf generates more accurate conformer ensembles compared to prior generative models.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

AIMSim : An accessible cheminformatics platform for similarity operations on chemicals datasets

The recent advances in deep learning, generative modeling, and statistical learning have ushered in a renewed interest in traditional cheminformatics tools and methods. Quantifying molecular similarity is essential in molecular generative modeling, exploratory molecular synthesis campaigns, and drug-discovery applications to assess how new molecules differ from existing ones. Further, most tools target advanced users and lack general implementations accessible to the larger community. In this work, we introduce Artificial Intelligence Molecular Similarity (AIMSim), an accessible cheminformatics platform for performing similarity operations on collections of molecules called molecular datasets. AIMSim provides a unified platform to perform similarity-based tasks on molecular datasets, such as diversity quantification, outlier and novelty analysis, clustering, dimensionality reduction, and inter-molecular comparisons. AIMSim implements all major binary similarity metrics and molecular fingerprints and is provided as a Python package that includes support for command-line use as well as a Graphical User Interface for code-free utilization with fully interactive plots.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Unifying simulation and inference with normalizing flows

There have been many applications of deep neural networks to detector calibrations and a growing number of studies that propose deep generative models as automated fast detector simulators. We show that these two tasks can be unified by using maximum likelihood estimation (MLE) from conditional generative models for energy regression. Unlike direct regression techniques, the MLE approach is prior independent and non-Gaussian resolutions can be determined from the shape of the likelihood near the maximum. Using an ATLAS-like calorimeter simulation, we demonstrate this concept in the context of calorimeter energy calibration. Published by the American Physical Society 2025

Hadronic calorimiters↗
NASA NTRS1974

Lift-induced wing-tip vortex attenuation

It has been shown by a new static airflow visualization method that a drogue device properly positioned downstream of the wing tip causes vortex breakdown. This same result has been obtained by mounting a jet engine simulator at the wing tip and directing the high-energy jet blast downstream into the vortex. These configurations, among others, are now under intensive investigation in the new Langley Vortex Research Facility. In this facility a balance mounted vortex generating model is propelled along the 1800-foot track while a second model trailed at 160 feet (scale distance of 1 mile) measures the far-field rolling moment induced by the vortex of the generating model.

Patterson, J. C., Jr.↗
DOE OSTI2024

RNA language models predict mutations that improve RNA function

Structured RNA lies at the heart of many central biological processes, from gene expression to catalysis. RNA structure prediction is not yet possible due to a lack of high-quality reference data associated with organismal phenotypes that could inform RNA function. We present GARNET (Gtdb Acquired RNa with Environmental Temperatures), a new database for RNA structural and functional analysis anchored to the Genome Taxonomy Database (GTDB). GARNET links RNA sequences to experimental and predicted optimal growth temperatures of GTDB reference organisms. Using GARNET, we develop sequence- and structure-aware RNA generative models, with overlapping triplet tokenization providing optimal encoding for a GPT-like model. Leveraging hyperthermophilic RNAs in GARNET and these RNA generative models, we identify mutations in ribosomal RNA that confer increased thermostability to the Escherichia coli ribosome. The GTDB-derived data and deep learning models presented here provide a foundation for understanding the connections between RNA sequence, structure, and function.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

Can Simple Machine Learning Tools Extend and Improve Temperature-Based Methods to Infer Streambed Flux?

Temperature-based methods have been developed to infer 1D vertical exchange flux between a stream and the subsurface. Current analyses rely on fitting physically based analytical and numerical models to temperature time series measured at multiple depths to infer daily average flux. These methods have seen wide use in hydrologic science despite strong simplifying assumptions including a lack of consideration of model structural error or the impacts of multidimensional flow or the impacts of transient streambed hydraulic properties. We performed a “perfect-model experiment” investigation to examine whether regression trees, with and without gradient boosting, can extract sufficient information from model-generated subsurface temperature time series, with and without added measurement error, to infer the corresponding exchange flux time series at the streambed surface. Using model-generated, synthetic data allowed us to assess the basic limitations to the use of machine learning; further examination of real data is only warranted if the method can be shown to perform well under these ideal conditions. We also examined whether the inherent feature importance analyses of tree-based machine learning methods can be used to optimize monitoring networks for exchange flux inference.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

$\mathrm{D}$ark$\mathrm{F}$lux: A new tool to analyze indirect-detection spectra of next-generation dark matter models

Here we present DarkFlux, a software tool designed to analyze indirect-detection signatures for next-generation models of dark matter (DM) with multiple annihilation channels. Version 1.0 of this tool accepts user-generated models with 2→2 tree-level dark matter annihilation to pairs of Standard Model (SM) particles and analyzes DM annihilation to γ rays. The tool consists of three modules, which can be run in a loop in order to scan over DM mass if desired: The annihilation fraction module calls an internal installation of MadDM, a dark matter phenomenology plugin for the Monte Carlo event generator MadGraph5 _ aMC@NLO, to compute the thermally averaged cross section $\langle σv\rangle_i$ for each annihilation channel $χχ$ ($\bar{χ}$,$χ^†$)→ i $\∈{SM, SM}. . The module then computes the fractional annihilation rate (annihilation fraction) into each channel. The flux module combines the flux spectrum from each annihilation channel, weighted by the appropriate annihilation fractions, to compute the total flux at Earth due to DM annihilation. In DarkFlux v1.0, this module specifically computes the γ-ray flux for each channel using the publicly available PPPC4DMID tables. The analysis module compares the total flux to observational data and computes the upper limit at 95% confidence level (CL) on the total thermally averaged DM annihilation cross section. In DarkFlux v1.0, this module compares the total γ-ray flux to a joint-likelihood analysis of fifteen dwarf spheroidal galaxies (dSphs) analyzed by the Fermi-LAT collaboration. DarkFlux v1.0 automatically provides data tables and can plot the output of these three modules. In this manual, we briefly motivate this indirect-detection computer tool and review the essential DM physics. We then describe the several modules of DarkFlux in greater detail. Finally, we show how to install and run DarkFlux and provide two worked examples demonstrating its capabilities.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Accelerating accurate simulations of calorimeter showers with normalizing flows and probability density distillation

Recently, we introduced caloflow, a high-fidelity generative model for GEANT4 calorimeter shower emulation based on normalizing flows. Here, we present caloflow v2, an improvement on our original framework that speeds up shower generation by a further factor of 500 relative to the original. The improvement is based on a technique called probability density distillation, originally developed for speech synthesis in the machine learning literature, and which we develop further by introducing a set of powerful new loss terms. We demonstrate that caloflow v2 preserves the same high fidelity as the original using qualitative (average images, histograms of high-level features) and quantitative (classifier metric between geant4 and generated samples) measures. The result is a generative model for calorimeter showers that matches the state of the art in speed (a factor of 10 4 faster than GEANT4) and greatly surpasses the previous state of the art in fidelity.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2023

Ensemble flow reconstruction in the atmospheric boundary layer from spatially limited measurements through latent diffusion models

Due to costs and practical constraints, field campaigns in the atmospheric boundary layer typically only measure a fraction of the atmospheric volume of interest. Machine learning techniques have previously successfully reconstructed unobserved regions of flow in canonical fluid mechanics problems and two-dimensional geophysical flows, but these techniques have not yet been demonstrated in the three-dimensional atmospheric boundary layer. Here, we conduct a numerical analogue of a field campaign with spatially limited measurements using large-eddy simulation. We pose flow reconstruction as an inpainting problem, and reconstruct realistic samples of turbulent, three-dimensional flow with the use of a latent diffusion model. The diffusion model generates physically plausible turbulent structures on larger spatial scales, even when input observations cover less than 1% of the volume. Through a combination of qualitative visualization and quantitative assessment, we demonstrate that the diffusion model generates meaningfully diverse samples when conditioned on just one observation. These samples successfully serve as initial conditions for a large-eddy simulation code. We find that diffusion models show promise and potential for other applications for other turbulent flow reconstruction problems.

17 WIND ENERGY↗
NASA NTRS2017

Assessment of MERRA-2 Land Surface Energy Flux Estimates

In MERRA-2, observed precipitation is inserted in place of model-generated precipitation at the land surface. The use of observed precipitation was originally developed for MERRA-Land(a land-only replay of MERRA with model-generated precipitation replaced with observations).Previously shown that the land hydrology in MERRA-2 and MERRA-Land is better than MERRA. We test whether the improved land surface hydrology in MERRA-2 leads to the expected improvements in the land surface energy fluxes and 2 m air temperatures (T2m).

GLEA↗
NASA NTRS1999

Neural Networks Used to Compare Designed and Measured Time-Average Patterns

Electronic time-average holograms are convenient for comparing the measured vibration modes of fan blades with those calculated by finite-element models. At the NASA Lewis Research Center, neural networks recently were trained to perform what had been a simple visual comparison of the predictions of the design models with the measurements. Finite-element models were used to train neural networks to recognize damage and strain information encoded in subtle changes in the time-average patterns of cantilevers. But the design-grade finite element models were unable to train the neural networks to detect damage in complex blade shapes. The design-model-generated patterns simply did not agree well enough with the measured patterns. Instead, hybrid-training records, with measured time-average patterns as the input and model-generated strain information as the output, were used to effect successful training.

Decker, Arthur J.↗
NASA NTRS1992

Advanced techniques in reliability model representation and solution

The current tendency of flight control system designs is towards increased integration of applications and increased distribution of computational elements. The reliability analysis of such systems is difficult because subsystem interactions are increasingly interdependent. Researchers at NASA Langley Research Center have been working for several years to extend the capability of Markov modeling techniques to address these problems. This effort has been focused in the areas of increased model abstraction and increased computational capability. The reliability model generator (RMG) is a software tool that uses as input a graphical object-oriented block diagram of the system. RMG uses a failure-effects algorithm to produce the reliability model from the graphical description. The ASSURE software tool is a parallel processing program that uses the semi-Markov unreliability range evaluator (SURE) solution technique and the abstract semi-Markov specification interface to the SURE tool (ASSIST) modeling language. A failure modes-effects simulation is used by ASSURE. These tools were used to analyze a significant portion of a complex flight control system. The successful combination of the power of graphical representation, automated model generation, and parallel computation leads to the conclusion that distributed fault-tolerant system architectures can now be analyzed.

Palumbo, Daniel L.↗
DOE OSTI2021

Challenges in simulating economic effects of climate change on global agricultural markets

Abstract Many approaches have been used to investigate climate change impacts on agriculture. However, several caveats remain in this field: (i) analyses focus only on a few major crops, (ii) large differences in yield impacts are observed between projections from site-based crops models and Global Gridded Crop Models (GGCMs), (iii) climate change impacts on livestock are rarely quantified, and (iv) several causal relations among biophysical, environmental, and socioeconomic aspects are usually not taken into account. We investigate how assumptions about these four aspects affect agricultural markets, food supply, consumer well-being, and land use at global level by deploying a large-scale socioeconomic model of the global economy with detailed representation of the agricultural sector. We find global welfare impacts several times larger when climate impacts all crops and all livestock compared to a scenario with impacts limited to major crops. At the regional level, food budget can decrease by 10 to 25% in developing countries, challenging food security. The role of land area expansion as a major source of adaptation is highlighted. Climate impacts on crop yields from site-based process crop models generate more challenging socioeconomic outcomes than those from GGCMs. We conclude that the agricultural research community should expand efforts to estimate climate impacts on many more crops and livestock. Also, careful comparison of the GGCMs and traditional site-based process crop models is needed to understand their major implications for agricultural and food markets.

Gurgel, Angelo C. (ORCID:0000000183319508)↗
DOE OSTI2020

Learning to Optimize Variational Quantum Circuits to Solve Combinatorial Problems

Quantum computing is a computational paradigm with the potential to outperform classical methods for a variety of problems. Proposed recently, the Quantum Approximate Optimization Algorithm (QAOA) is considered as one of the leading candidates for demonstrating quantum advantage in the near term. QAOA is a variational hybrid quantum-classical algorithm for approximately solving combinatorial optimization problems. The quality of the solution obtained by QAOA for a given problem instance depends on the performance of the classical optimizer used to optimize the variational parameters. In this paper, we formulate the problem of finding optimal QAOA parameters as a learning task in which the knowledge gained from solving training instances can be leveraged to find high-quality solutions for unseen test instances. To this end, we develop two machine-learning-based approaches. Our first approach adopts a reinforcement learning (RL) framework to learn a policy network to optimize QAOA circuits. Our second approach adopts a kernel density estimation (KDE) technique to learn a generative model of optimal QAOA parameters. In both approaches, the training procedure is performed on small-sized problem instances that can be simulated on a classical computer; yet the learned RL policy and the generative model can be used to efficiently solve larger problems. Furthermore, extensive simulations using the IBM Qiskit Aer quantum circuit simulator demonstrate that our proposed RL- and KDE-based approaches reduce the optimality gap by factors up to 30.15 when compared with other commonly used off-the-shelf optimizers.

97 MATHEMATICS AND COMPUTING↗