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.

SourceResearch connectionFrom yearThrough year

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

At least 685 records · Page 38

All sources NASA NTRS DOE OSTI

DOE OSTI2024

SCIMON: Scientific Inspiration Machines Optimized for Novelty

We explore and enhance the ability of neu- ral language models to generate novel scien- tific directions grounded in literature. Work on literature-based hypothesis generation has traditionally focused on binary link prediction— severely limiting the expressivity of hypothe- ses. This line of work also does not focus on optimizing novelty. We take a dramatic depar- ture with a novel setting in which models use as input background contexts (e.g., problems, experimental settings, goals), and output natu- ral language ideas grounded in literature. We present SCIMON, a modeling framework that uses retrieval of “inspirations” from past scien- tific papers, and explicitly optimizes for novelty by iteratively comparing to prior papers and up- dating idea suggestions until sufficient novelty is achieved. Comprehensive evaluations reveal that GPT-4 tends to generate ideas with over- all low technical depth and novelty, while our methods partially mitigate this issue. Our work represents a first step toward evaluating and developing language models that generate new ideas derived from the scientific literature.

Ji, Heng↗

DOE OSTI2021

ADVANCED THERMAL-HYDRAULIC MODEL OF HEAT RECOVERY STEAM GENERATORS

Steam-side oxidation at elevated temperatures in heat recovery steam generators (HRSGs) can lead to overheating of local tubes, potential oxide scale exfoliation, and damage to critical equipment. Predicting the oxide growth distribution requires detailed knowledge of the distribution of tube metal temperatures. To address these challenges, a computational fluid dynamics (CFD) model of a high-pressure superheater (HPSH) was developed within the Siemens’ STAR-CCM+ CFD commercial software. The HPSH section of the steam generator was selected for modeling because of its high oxidation rates.The model development included several steps. In the first step, an oxidation model based on a parabolic oxidation relationship was implemented. The oxidation growth predicted by the model was then verified for a full-load boiler operation case. In the second step, a three-dimensional CFD model was developed to account for the thermal and flow physics involved in HRSGs. The CFD model was then verified and validated on a simplified geometry that contained six row-tubes which mimicked the tube arrangements of an HPSH section. Numerical results compared well with published experimental results. In the third step, the oxidation model was coupled to the HRSG’s CFD model, and a comparison study was performed for two cases: (a) regular finless tubes and (b) finned tubes. Preliminary results showed that the oxide thickness predicted by the case with tube fins was higher compared to that of the finless tube case, thus demonstrating the need to model the effects of the tube fins on oxide thickness.

Panicker, Nithin↗

DOE OSTI2022

Cubit for MCNP Unstructured Mesh Analysis of Oktavian Benchmarks

The Monte Carlo N-Particle (MCNP) transport code developed by Los Alamos National Laboratory (LANL) can be used to transport various particles across user defined three dimensional (3D) geometries. Traditionally, these geometries are created as constructive solid geometry (CSG), involving the use of Boolean operators on defined surfaces to create 3D regions known as cells. A newer method of geometry definition in the MCNP code is unstructured mesh (UM) embedded within a CSG cell using the universe and fill repeated-structure features. An MCNP UM calculation requires an MCNP input file and mesh geometry file. The MCNP code cannot be used to generate UM geometry models. A computer-aided design (CAD) software is typically used to construct a solid geometry model, and a CAD file is then imported into a mesh generation software to prepare and generate a mesh model. Some mesh generation software packages have the ability to create a solid geometry and thus a separate CAD software for creating a CAD model is not needed. In this work, Cubit, a geometry creation and meshing software developed by Sandia National Laboratories, is used to construct UM models for MCNP simulations.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2024

Understanding Generative AI Content with Embedding Models

The construction of high-quality numerical features is critical to any quantitative data analysis. Feature engineering has been historically addressed by carefully hand-crafting data representations based on domain expertise. This work views the internal representations of modern deep neural networks (DNNs), called embeddings, as an implicit form of traditional feature engineering. For trained DNNs, we show that these embeddings can reveal interpretable, high-level concepts in unstructured sample data. We use these embeddings in natural language and computer vision tasks to uncover both inherent heterogeneity in the underlying data and human-understandable explanations for it. In particular, we find empirical evidence that there is inherent separability between real data and those generated from AI models.

Vargas, Max↗

DOE OSTI2021

Climate impacts on global agriculture emerge earlier in new generation of climate and crop models

Potential climate-related impacts on future crop yield are a major societal concern. Previous projections of the Agricultural Model Intercomparison and Improvement Project’s Global Gridded Crop Model Intercomparison based on the Coupled Model Intercomparison Project Phase 5 identified substantial climate impacts on all major crops, but associated uncertainties were substantial. Here we report new twenty-first-century projections using ensembles of latest-generation crop and climate models. Results suggest markedly more pessimistic yield responses for maize, soybean and rice compared to the original ensemble. Mean end-of-century maize productivity is shifted from +5% to -6% (SSP126) and from +1% to -24% (SSP585)—explained by warmer climate projections and improved crop model sensitivities. In contrast, wheat shows stronger gains (+9% shifted to +18%, SSP585), linked to higher CO2 concentrations and expanded high-latitude gains. The ‘emergence’ of climate impacts consistently occurs earlier in the new projections—before 2040 for several main producing regions. While future yield estimates remain uncertain, these results suggest that major breadbasket regions will face distinct anthropogenic climatic risks sooner than previously anticipated.

Jagermeyr, Jonas↗

DOE OSTI2022

INGRID: An interactive grid generator for 2D edge plasma modeling

A fusion boundary-plasma domain is defined by axisymmetric magnetic surfaces where the geometry is often complicated by the presence of one or more X-points; and modeling boundary plasmas usually relies on computational grids that account for the magnetic field geometry. The new grid generator INGRID (Interactive Grid Generator) presented here is a Python-based code for calculating grids for fusion boundary plasma modeling, for a variety of configurations with one or two X-points in the domain. INGRID first performs partitioning over the domain consisting of a small number of patches conforming to the magnetic field and wall geometry; then it generates a subgrid on each of the patches and joins them into a global grid. This domain partitioning strategy makes possible a uniform treatment of various configurations with one or two X-points in the domain. This includes single-null, double-null, and other configurations with two X-points in the domain. The INGRID design allows generating grids either interactively, via a parameterfile driven GUI, or using a non-interactive script-controlled workflow. Results of testing demonstrate that INGRID is a flexible, robust, and user-friendly grid-generation tool for fusion boundary-plasma modeling.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2020

Stochastic DC optimal power flow with reserve saturation

We propose an optimization framework for stochastic optimal power flow with uncertain loads and renewable generator capacity. Our model follows previous work in assuming that generator outputs respond to load imbalances according to an affine control policy, but introduces a model of saturation of generator reserves by assuming that when a generator’s target level hits its limit, it abandons the affine policy and produces at that limit. This is a particularly interesting feature in models where wind power plants, which have uncertain upper generation limits, are scheduled to provide reserves to balance load fluctuations. Here, the resulting model is a nonsmooth nonconvex two-stage stochastic program, and we use a stochastic approximation method to find stationary solutions to a smooth approximation. Computational results on 6-bus and 118-bus test instances demonstrate that by considering the effects of saturation, our model can yield solutions with lower expected generation costs (at the same target line violation probability level) than those obtained from a model that enforces the affine policy to stay within generator limits with high probability.

17 WIND ENERGY↗

DOE OSTI2022

INGRID: An interactive grid generator for 2D edge plasma modeling

A fusion boundary-plasma domain is defined by axisymmetric magnetic surfaces where the geometry is often complicated by the presence of one or more X-points; and modeling boundary plasmas usually relies on computational grids that account for the magnetic field geometry. The new grid generator INGRID (Interactive Grid Generator) presented in this work is a Python-based code for calculating grids for fusion boundary plasma modeling, for a variety of configurations with one or two X-points in the domain. INGRID first performs partitioning over the domain consisting of a small number of patches conforming to the magnetic field and wall geometry; then it generates a subgrid on each of the patches and joins them into a global grid. This domain partitioning strategy makes possible a uniform treatment of various configurations with one or two X-points in the domain. This includes single-null, double-null, and other configurations with two X-points in the domain. The INGRID design allows generating grids either interactively, via a parameter-file driven GUI, or using a non-interactive script-controlled workflow. Results of testing demonstrate that INGRID is a flexible, robust, and user-friendly grid-generation tool for fusion boundary-plasma modeling.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2026

A curated benchmark for cofolding models on kinase conformational states

Abstract Protein kinases are critical drug targets, requiring therapeutics that can modulate their active and inactive conformational states. While cofolding models can generate global folds directly from kinase sequences and ligand SMILES strings, these models have not yet been tested on their ability to recover ligand-induced-fit conformational states of the kinase proteins. Here, we introduce KinConfBench, a curated benchmark of 2225 high-quality human kinase chains to evaluate the ability of four state-of-the-art cofolding models—Boltz-2, Chai-1, Protenix, and RoseTTAFold-All-Atom—to recover both canonical and rare conformational states. We show that geometric success metrics of a ligand pose in the active site do not correlate strongly with the correct kinase conformational state, motivating a new set of dynamical benchmarks for assessing cofolding models. While all four cofolding models achieve ~60–80% prediction accuracy for kinase conformational classification, they exhibit severe mode collapse when performing multiple inferences, show negligible structural diversity in sampling induced-fit motions, and display a prevalent “apo-drift” in which most cofolding models predominantly predict the kinase to be in its ligand-free state. Our results highlight that capturing ligand-induced protein conformational diversity, not just geometric fit, is critical for next-generation structure-based drug discovery.

Sun, Kunyang↗

DOE OSTI2024

Statistical data analysis of x-ray spectroscopy data enabled by neural network accelerated Bayesian inference

Bayesian inference applied to x-ray spectroscopy data analysis enables uncertainty quantification necessary to rigorously test theoretical models. However, when comparing to data, detailed atomic physics and radiation transfer calculations of x-ray emission from non-uniform plasma conditions are typically too slow to be performed in line with statistical sampling methods, such as Markov Chain Monte Carlo sampling. Furthermore, differences in transition energies and x-ray opacities often make direct comparisons between simulated and measured spectra unreliable. Here, we present a spectral decomposition method that allows for corrections to line positions and bound–bound opacities to best fit experimental data, with the goal of providing quantitative feedback to improve the underlying theoretical models and guide future experiments. In this work, we use a neural network (NN) surrogate model to replace spectral calculations of isobaric hot-spots created in Kr-doped implosions at the National Ignition Facility. The NN was trained on calculations of x-ray spectra using an isobaric hot-spot model post-processed with Cretin, a multi-species atomic kinetics and radiation code. The speedup provided by the NN model to generate x-ray emission spectra enables statistical analysis of parameterized models with sufficient detail to accurately represent the physical system and extract the plasma parameters of interest.

47 OTHER INSTRUMENTATION↗

DOE OSTI2020

Profiles of upcoming HPC Applications and their Impact on Reservation Strategies

With the expected convergence between HPC, BigData and AI, new applications with different profiles are coming to HPC infrastructures. Here, we aim at better understanding the features and needs of these applications in order to be able to run them efficiently on HPC platforms. The approach followed is bottom-up: we study thoroughly an emerging application from the neuroscience community (SLANT) to understand its behavior. Based on these observations, we derive a generic, yet simple, application model (namely, a linear sequence of stochastic jobs). We expect this model to be representative for a large set of upcoming applications that require the computational power of HPC clusters without fitting the typical behavior of large-scale traditional applications. In a second step, we show how one can manipulate this generic model in a scheduling framework. Specifically we consider the problem of making reservations (both time and memory) for an execution on an HPC platform. We derive solutions using the model of the first step of this work. We experimentally show the robustness of the model, even with very few data or with another application, to generate the model, and provide performance gains with regards to standard and more recent approaches used in the neuroscience community.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2024

Sensitivity of modelled passive margin stratigraphy to variations in sea level, sediment supply and subsidence

Abstract We produced a 10 Myr synthetic stratigraphic section using a forward stratigraphic model that generates marine deltaic stratigraphy over geological timescales. We recursively fit the model using a Bayesian inversion algorithm to test: (1) if it could be accurately reconstructed; (2) if the parameters used to create it could be recovered; and (3) the sensitivity of the model output to given model parameters and the attendant physical processes. The original synthetic stratigraphic section was produced with cyclical sea‐level variations of 40 and 30 m with 2.4 and 10 Myr periods respectively. Sediment was also supplied cyclically, in 2.4 and 10 Myr cycles with amplitudes of 30 and 80 tons/100 kyr, respectively, varying from a mean of 232 tons/100 kyr. Parameter values were sampled to fit the model using a Markov chain Monte Carlo algorithm, resulting in a ±5 m (1σ) variation between the experimental output and the original. Sea level varied by ±7 m (1σ) within the posterior distribution of parameters. As a result, both the 10 Myr and 2.4 Myr sea‐level cycles could be extracted from the original output. The variation in sediment supply was approximately ±38 tons/100 kyr (1σ) and, as a result, only the larger long‐term supply variations could be accurately recovered in refitting the model. The variation in thermal, flexural and total subsidence across those parameter sets is less than ±10 m (1σ). The original section experienced 150 m of total subsidence at the depocentre. Our results demonstrate the distinct and interpretable imprint of sea level and subsidence on continental margin stratigraphy can be quantified. Moreover, we conclude that sea‐level change produces a defined effect on the geometries of stratigraphic architecture, and that techniques applied for the purpose of delineating sea‐level variation from continental margin strata have a well‐founded conceptual basis.

Schmelz, William J.↗

DOE OSTI2023

Catalyst: Fast and flexible modeling of reaction networks

We introduce Catalyst.jl, a flexible and feature-filled Julia library for modeling and high-performance simulation of chemical reaction networks (CRNs). Catalyst supports simulating stochastic chemical kinetics (jump process), chemical Langevin equation (stochastic differential equation), and reaction rate equation (ordinary differential equation) representations for CRNs. Through comprehensive benchmarks, we demonstrate that Catalyst simulation runtimes are often one to two orders of magnitude faster than other popular tools. More broadly, Catalyst acts as both a domain-specific language and an intermediate representation for symbolically encoding CRN models as Julia-native objects. This enables a pipeline of symbolically specifying, analyzing, and modifying CRNs; converting Catalyst models to symbolic representations of concrete mathematical models; and generating compiled code for numerical solvers. Leveraging ModelingToolkit.jl and Symbolics.jl, Catalyst models can be analyzed, simplified, and compiled into optimized representations for use in numerical solvers. Finally, we demonstrate Catalyst’s broad extensibility and composability by highlighting how it can compose with a variety of Julia libraries, and how existing open-source biological modeling projects have extended its intermediate representation.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2022

Can we use antipredator behavior theory to predict wildlife responses to high-speed vehicles?

Animals seem to rely on antipredator behavior to avoid vehicle collisions. There is an extensive body of antipredator behavior theory that have been used to predict the distance/time animals should escape from predators. These models have also been used to guide empirical research on escape behavior from vehicles. However, little is known as to whether antipredator behavior models are appropriate to apply to an approaching high-speed vehicle scenario. We addressed this gap by (a) providing an overview of the main hypotheses and predictions of different antipredator behavior models via a literature review, (b) exploring whether these models can generate quantitative predictions on escape distance when parameterized with empirical data from the literature, and (c) evaluating their sensitivity to vehicle approach speed using a simulation approach wherein we assessed model performance based on changes in effect size with variations in the slope of the flight initiation distance (FID) vs. approach speed relationship. The slope of the FID vs. approach speed relationship was then related back to three different behavioral rules animals may rely on to avoid approaching threats: the spatial, temporal, or delayed margin of safety. We used literature on birds for goals (b) and (c). Our review considered the following eight models: the economic escape model, Blumstein’s economic escape model, the optimal escape model, the perceptual limit hypothesis, the visual cue model, the flush early and avoid the rush (FEAR) hypothesis, the looming stimulus hypothesis, and the Bayesian model of escape behavior. We were able to generate quantitative predictions about escape distance with the last five models. However, we were only able to assess sensitivity to vehicle approach speed for the last three models. The FEAR hypothesis is most sensitive to high-speed vehicles when the species follows the spatial (FID remains constant as speed increases) and the temporal margin of safety (FID increases with an increase in speed) rules of escape. The looming stimulus effect hypothesis reached small to intermediate levels of sensitivity to high-speed vehicles when a species follows the delayed margin of safety (FID decreases with an increase in speed). The Bayesian optimal escape model reached intermediate levels of sensitivity to approach speed across all escape rules (spatial, temporal, delayed margins of safety) but only for larger (> 1 kg) species, but was not sensitive to speed for smaller species. Overall, no single antipredator behavior model could characterize all different types of escape responses relative to vehicle approach speed but some models showed some levels of sensitivity for certain rules of escape behavior. We derive some applied applications of our findings by suggesting the estimation of critical vehicle approach speeds for managing populations that are especially susceptible to road mortality. Overall, we recommend that new escape behavior models specifically tailored to high-speeds vehicles should be developed to better predict quantitatively the responses of animals to an increase in the frequency of cars, airplanes, drones, etc. they will face in the next decade.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2024

Convolutional L2LFlows: generating accurate showers in highly granular calorimeters using convolutional normalizing flows

Abstract In the quest to build generative surrogate models as computationally efficient alternatives to rule-based simulations, the quality of the generated samples remains a crucial frontier. So far, normalizing flows have been among the models with the best fidelity. However, as the latent space in such models is required to have the same dimensionality as the data space, scaling up normalizing flows to high dimensional datasets is not straightforward. The prior L2LFlows approach successfully used a series of separate normalizing flows and sequence of conditioning steps to circumvent this problem. In this work, we extend L2LFlows to simulate showers with a 9-times larger profile in the lateral direction. To achieve this, we introduce convolutional layers and U-Net-type connections, move from masked autoregressive flows to coupling layers, and demonstrate the successful modelling of showers in the ILD Electromagnetic Calorimeter as well as Dataset 3 from the public CaloChallenge dataset.

Instruments & Instrumentation↗

DOE OSTI2022

Limitations of reanalysis data for wind power applications

Wind energy resource estimates commonly depend on simulated wind speed profiles generated by reanalysis or weather models due to the lack of long time series measurements with sufficient coverage at relevant heights (roughly 90 m above ground). However, modeled data, including reanalyses, can be noisy and display a wide range of biases and errors, variously attributed to terrain effects, poor coverage of assimilated inputs, and model resolution. Wind generation records, if available at high temporal and geographical resolution, can provide a proxy for wind measurements and allow for evaluation of reanalyses and weather model wind time series. We use a 7-year-long data set of hourly, plant-level generation records from over 100 wind plants across Texas to evaluate two commonly used reanalysis data sets (MERRA2 and ERA5). Additionally, we use 1-year of records (2019) to evaluate an operational, high-resolution regional weather modeling product (HRRR v3). We find that across the region, and across all modeling products, the modeled representation of wind generation (i.e., wind speeds at hub heights passed through a power curve) has relatively small mean errors when aggregated daily, but that accuracy and hourly correlation have a strong diurnal sensitivity. Accuracy and correlation systematically decline through the evening and markedly improve after sunrise. These diurnal patterns persist even in the highest resolution model tested (HRRR v3). We hypothesize the nighttime decline in accuracy is mostly due to poorly represented boundary layer conditions, perhaps related to model representation of stability, while other uncertainties (such as wake effects) play a secondary role.

17 WIND ENERGY↗

DOE OSTI2024

Using Co-Simulation to Model Interconnect-Scale Power Systems from Loads to Generators

Co-simulation is a modeling technique that allows analysts to combine simulation tools and their corresponding models to exchange data during run-time, allowing the creation of larger and more complex models across heterogeneous domains. HELICS is a co-simulation platform developed over the past six years that has been shown to be effective for these multi-domain analysis. Recently, a HELICS-based analysis was completed where the ERCOT electrical interconnect in the United States was modeled in high detail from bulk power system generation to individual customer loads. This model was used to evaluate a flat-rate and transactive energy tariff with integrated wholesale and retail real-time and day-ahead energy markets. This modeling allows detailed analysis showing how the operations of the power system under these tariffs impact all actors in the power system, from individual customers to bulk power system operators.

co-simulation, HELICS, transactive energy system, ↗

DOE OSTI2025

Selection of Global Climate Model Data for Downscaling With Generative Machine Learning and Use in the Power Planning for Alignment of Climate and Energy Systems Project

The range of results from climate models and scenarios is important to the understanding of uncertainty in power planning analysis. A U.S. Department of Energy-funded analytic project called Power Planning for Alignment of Climate and Energy Systems is developing data and analytic methods to reflect the effects of climate change on key variables for power system planning, as part of the Grid Modernization Lab Consortium. This project will select and prepare global climate model results for use in power system planning models. A related report (Evaluation of Global Climate Models for Use in Energy Analysis) assesses the performance of various global climate models from the Coupled Model Intercomparison Project Phase 6 data archive for their historical skill with respect to energy system performance and for their future projections under multiple climate change scenarios. Building from that report, we describe the selection of a climate scenario (Shared Socioeconomic Pathway [SSP] 2-4.5) and five climate models: TaiESM1, EC-Earth3-CC, GFDL-CM4, EC-Earth3-Veg, and MPI-ESM1-2-HR. We describe the model selection criteria, which were based on the quality of the match between model results under historical conditions and on the representation of the range of future values for several variables. These results will be downscaled via an open-source generative machine learning method called Super-Resolution for Renewable Energy Resource Data with Climate Change Impacts.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗