Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “model discovery”

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 649 records · Page 36
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Subsurface Characterization and Machine Learning Predictions at Brady Hot Springs Results

Geothermal power plants typically show decreasing heat and power production rates over time. Mitigation strategies include optimizing the management of existing wells - increasing or decreasing the fluid flow rates across the wells - and drilling new wells at appropriate locations. The latter is expensive, time-consuming, and subject to many engineering constraints, but the former is a viable mechanism for periodic adjustment of the available fluid allocations. Data and supporting literature from a study describing a new approach combining reservoir modeling and machine learning to produce models that enable strategies for the mitigation of decreased heat and power production rates over time for geothermal power plants. The computational approach used enables translation of sets of potential flow rates for the active wells into reservoir-wide estimates of produced energy and discovery of optimal flow allocations among the studied sets. In our computational experiments, we utilize collections of simulations for a specific reservoir (which capture subsurface characterization and realize history matching) along with machine learning models that predict temperature and pressure timeseries for production wells. We evaluate this approach using an "open-source" reservoir we have constructed that captures many of the characteristics of Brady Hot Springs, a commercially operational geothermal field in Nevada, USA. Selected results from a reservoir model of Brady Hot Springs itself are presented to show successful application to an existing system. In both cases, energy predictions prove to be highly accurate: all observed prediction errors do not exceed 3.68% for temperatures and 4.75% for pressures. In a cumulative energy estimation, we observe prediction errors that are less than 4.04%. A typical reservoir simulation for Brady Hot Springs completes in approximately 4 hours, whereas our machine learning models yield accurate 20-year predictions for temperatures, pressures, and produced energy in 0.9 seconds. This paper aims to demonstrate how the models and techniques from our study can be applied to achieve rapid exploration of controlled parameters and optimization of other geothermal reservoirs. Includes a synthetic, yet realistic, model of a geothermal reservoir, referred to as open-source reservoir (OSR). OSR is a 10-well (4 injection wells and 6 production wells) system that resembles Brady Hot Springs (a commercially operational geothermal field in Nevada, USA) at a high level but has a number of sufficiently modified characteristics (which renders any possible similarity between specific characteristics like temperatures and pressures as purely random). We study OSR through CMG simulations with a wide range of flow allocation scenarios. Includes a dataset with 101 simulated scenarios that cover the period of time between 2020 and 2040 and a link to the published paper about this project, where we focus on the Machine Learning work for predicting OSR's energy production based on the simulation data, as well as a link to the GitHub repository where we have published the code we have developed (please refer to the repository's readme file to see instructions on how to run the code). Additional links are included to associated work led by the USGS to identify geologic factors associated with well productivity in geothermal fields. Below are the high-level steps for applying the same modeling + ML process to other geothermal reservoirs: 1. Develop a geologic model of the geothermal field. The location of faults, upflow zones, aquifers, etc. need to be accounted for as accurately as possible 2. The geologic model needs to be converted to a reservoir model that can be used in a reservoir simulator, such as, for instance, CMG STARS, TETRAD, or FALCON 3. Using native state modeling, the initial temperature and pressure distributions are evaluated, and they become the initial conditions for dynamic reservoir simulations 4....

15 GEOTHERMAL ENERGY↗
DOE OSTI2026

FAIR Data and Interpretable AI Framework for Architectured Metamaterials (Final Report)

This research program established a transformative framework for the discovery and design of mechanical metamaterials, which are architected structures engineered to control physical phenomena like sound and vibration in ways natural materials cannot. To overcome the traditional reliance on trial-and-error, the project developed an interpretable Artificial Intelligence (AI) framework that moves beyond "black box" models to reveal the specific geometric patterns—such as "unit-cell templates"—that govern a material’s performance. A major breakthrough was the development of a hierarchical design method, which allows a single material to block vibrations across multiple frequency ranges simultaneously by layering patterns at different scales without them interfering with one another. This was further expanded to include irregular, graph-based designs that use spanning tree algorithms to ensure structural connectivity while allowing for customized, direction-dependent properties like stiffness and acoustic impedance. Beyond design, the project addressed the practicalities of real-world production by developing uncertainty quantification techniques that account for manufacturing defects and material variability, reducing the need for expensive physical testing by orders of magnitude. To speed up the discovery process, the team implemented Gaussian Process Regression and other surrogate models that provide accurate performance predictions at a fraction of the traditional computational cost. The AI-generated designs were successfully validated through fabrication of physical samples and wave propagation experiments, confirming their ability to accurately guide or reflect waves as predicted. By contributing these tools and high-quality FAIR benchmark datasets to the wider scientific community, this work provides a scalable foundation for advancing technologies in aerospace vibration control, medical imaging, and noise reduction.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Protein model quality assessment using rotation–equivariant transformations on point clouds

Machine learning research concerning protein structure has seen a surge in popularity over the last years with promising advances for basic science and drug discovery. Working with macromolecular structure in a machine learning context requires an adequate numerical representation, and researchers have extensively studied representations such as graphs, discretized 3D grids, and distance maps. As part of CASP14, we explored a new and conceptually simple representation in a blind experiment: atoms as points in 3D, each with associated features. These features—initially just the basic element type of each atom—are updated through a series of neural network layers featuring rotation-equivariant convolutions. Starting from all atoms, we further aggregate information at the level of alpha carbons before making a prediction at the level of the entire protein structure. We find that this approach yields competitive results in protein model quality assessment despite its simplicity and despite the fact that it incorporates minimal prior information and is trained on relatively little data. As a result, its performance and generality are particularly noteworthy in an era where highly complex, customized machine learning methods such as AlphaFold 2 have come to dominate protein structure prediction.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2020

Solving Hard Problems with AI: Dramatically Accelerating Drug Discovery Through A Unique Public-Private Partnership

Dramatic disruptions in technologies can result in the creation of new markets, the emergence of new paradigms, and the displacement of entrenched approaches and business models. This can be especially pronounced when multiple technologies converge to create something new. Today, artificial intelligence (“AI”) is acting as an accelerant for such technology transformations. Still, the scope and depth of its impact will depend in part on our ability to address challenging problems that are surfacing without solutions. Industries will pursue AI for their corporate missions and to create shareholder value. But the timescales and applications can be incongruent with important public needs and with the longer research horizons required to make real progress. Public-Private partnerships are an important mechanism for tackling these rapidly emerging new, complex, challenging problems. Finally, this paper examines our experience gained by creating a unique public-private partnership to apply AI to one example of a hard problem that is particularly timely today-- dramatically accelerating drug discovery.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Generative models for simulation of KamLAND-Zen

Abstract The next generation of searches for neutrinoless double beta decay ($$0 \nu \beta \beta $$ 0 ν β β ) are poised to answer deep questions on the nature of neutrinos and the source of the Universe’s matter–antimatter asymmetry. They will be looking for event rates of less than one event per ton of instrumented isotope per year. To claim discovery, accurate and efficient simulations of detector events that mimic$$0 \nu \beta \beta $$ 0 ν β β is critical. Traditional Monte Carlo (MC) simulations can be supplemented by machine-learning-based generative models. This work describes the performance of generative models that we designed for monolithic liquid scintillator detectors like KamLAND to produce accurate simulation data without a predefined physics model. We present their current ability to recover low-level features and perform interpolation. In the future, the results of these generative models can be used to improve event classification and background rejection by providing high-quality abundant generated data.

Physics↗
DOE OSTI2024

Causes of and Solutions to Wind Speed Bias in NREL's 2020 Offshore Wind Resource Assessment for the California Pacific Outer Continental Shelf

This report provides the results of a detailed analysis into the causes of high wind speed bias in the 20-year wind resource data set for offshore California the National Renewable Energy Laboratory (NREL) released in 2020, herein called CA20. The data set was developed using the state-of-the-art Weather Research and Forecasting (WRF) model. Notably, no floating lidars were available at the time in offshore California to validate offshore hub-height wind speeds. In late 2020, the Pacific Northwest National Laboratory (PNNL) deployed two floating lidars in the California outer continental shelf (OCS), near the Bureau of Ocean Energy Management (BOEM) call areas of Humboldt and Morro Bay. Using these observations through 2021, NREL found considerable bias in modeled hub-height winds at both locations: up to +2 m/s at Humboldt over a 6-month period, and up to +1 m/s at Morro Bay over a one-year period. Upon the discovery of this bias, the Department of Energy (DOE) and BOEM funded NREL and PNNL to investigate the causes of, impacts of, and solutions to the bias in the CA20 data set. This report summarizes the findings of this research. We first investigated whether different WRF model setups could lead to reduced bias. We found that the choice of planetary boundary layer (PBL) scheme - which controls the vertical turbulent mixing of momentum, heat, and moisture in the lowermost part of the atmosphere - greatly affected hub-height wind speeds in the region. Specifically, switching from the Mellor-Yamada-Nakanishi-Niino (MYNN) scheme used in CA20 (and widely used across a range of operational and research weather models) to the less common Yonsei University (YSU) scheme nearly eliminated the bias at both the Humboldt and Morro Bay lidar locations. The large discrepancy between the MYNN- and YSU-modeled hub-height winds pointed towards the role of atmospheric stability. In general, PBL schemes agree well in conditions of high turbulence and mixing, normally referred to as "unstable" conditions. By contrast, PBL schemes start to diverge in "stable" conditions, where turbulence is low and thermal stratification (i.e., higher temperature air sitting on top of colder air) greatly suppresses vertical mixing. Under such conditions, winds aloft can decouple from surface effects and greatly accelerate, causing high wind speeds at hub-height and frequent low-level jets (LLJs). We determined that these stable conditions are in fact dominant in offshore California. The region is characterized by moderate-to-extreme stable stratification with a LLJ on average around 200 meters above sea-level. To our knowledge, no wind energy area globally has as strongly stable stratification as offshore California. Under these extreme conditions, we determined that the MYNN scheme models higher stability than YSU, resulting in less vertical turbulent mixing than YSU, allowing for the acceleration of hub-height winds, more intense LLJs, and higher-amplitude inertial oscillations. Using surface observations, we found that MYNN overestimates near-surface stability, whereas YSU tends to model stability better. We then considered several short-term case studies to assess additional meteorological drivers of the bias at Humboldt. We found that during synoptic scale northerly flows driven by the North Pacific High and inland thermal low, a coastal warm bias in the MYNN case studies contributes to the modeled wind speed bias by altering the boundary layer thermodynamics via a thermal wind mechanism. Given the strong performance of the YSU-based runs in offshore California, NREL has produced and published an updated version of the CA20 data set with YSU as the PBL scheme. This updated data set is now part of NREL's 2023 National Offshore Wind (NOW-23) data set, which covers all the U.S. offshore waters. The development and final validation of the NOW-23 data set in offshore California is documented in this report.

17 WIND ENERGY↗
DOE OSTI2024

Resonant multiscalar production in the generic complex singlet model in the multi-TeV region

We develop benchmarks for resonant discalar production in the generic complex singlet scalar extension of the Standard Model (SM) with no additional symmetries, which contains two new scalars. These benchmarks maximize discalar resonant production modes at future 𝑝⁢𝑝 colliders: 𝑝⁢𝑝 → ℎ 2 → ℎ 1⁡ ℎ 1 , 𝑝⁢𝑝 → ℎ 2 → ℎ 1 ⁡ℎ 3 , and 𝑝⁢𝑝 → ℎ 2 → ℎ 3 ⁡ℎ 3 , where ℎ 1 is the observed SM-like Higgs boson and ℎ 2,3 are new scalars. The decays ℎ 2 → ℎ 1 ⁡ℎ 3 and ℎ 2 → ℎ 3 ⁡ℎ 3 may be the only way to discover ℎ 3 , leading to a discovery of two new scalars at once. Current LHC and projected future collider (HL-LHC, FCC-ee+HL-LHC, IL⁢C⁢500+HL-LHC) constraints on this model are used to produce benchmarks at the HL-LHC for ℎ 2 masses between 250 GeV and 1 TeV and a future 𝑝⁢𝑝 collider (FCC-hh) for ℎ 2 masses between 250 GeV and 12 TeV. We update the current LHC bounds on the singlet-Higgs boson mixing angle for these benchmarks. As the mass of ℎ 2 approaches the multi-TeV region, certain limiting behaviors of the maximum rates are uncovered due to theoretical constraints on the parameters. These limits, which can be derived analytically, are BR⁡(ℎ 2 → ℎ 1⁡ ℎ 1 ) → 0.25, BR⁡(ℎ 2 → ℎ 3 ⁡ℎ 3 ) → 0.5, and BR⁡(ℎ 2 → ℎ 1⁡ ℎ 3 ) → 0. It can also be shown that the maximum rates of 𝑝⁢𝑝 → ℎ 2 → ℎ 1 ⁡ℎ 1 and 𝑝⁢𝑝 → ℎ 2 → ℎ 3⁡ ℎ 3 approach the same value. Hence, all three ℎ 2 → ℎ 𝑖 ⁡ℎ 𝑗 decays are promising discovery modes for ℎ 2 masses at and below 𝒪⁡(1 TeV), while above 𝒪⁡(1 TeV) the decays ℎ 2 → ℎ 1 ⁡ℎ 1 and ℎ 2 → ℎ 3 ⁡ℎ 3 are more encouraging. We choose benchmark masses for ℎ 3 to produce a large range of decay signatures including multi-𝑏, multivector boson, and multi-SM-like Higgs production. As we will show, the behavior of the maximum rates leads to the surprising conclusion that in the multi-TeV region this model may be discovered in the Higgs quartet production mode via ℎ 2 → ℎ 3 ⁡ℎ 3 → 4⁢ℎ 1 decays before Higgs triple production is observed. The maximum di- and four Higgs production rates are similar in the multi-TeV range.

Extensions of Higgs sector↗
DOE OSTI2020

GrAMME: Semisupervised Learning Using Multilayered Graph Attention Models

Modern data analysis pipelines are becoming increasingly complex due to the presence of multiview information sources. While graphs are effective in modeling complex relationships, in many scenarios, a single graph is rarely sufficient to succinctly represent all interactions, and hence, multilayered graphs have become popular. Though this leads to richer representations, extending solutions from the single-graph case is not straightforward. Consequently, there is a strong need for novel solutions to solve classical problems, such as node classification, in the multilayered case. In this article, we consider the problem of semisupervised learning with multilayered graphs. Though deep network embeddings, e.g., DeepWalk, are widely adopted for community discovery, we argue that feature learning with random node attributes, using graph neural networks, can be more effective. To this end, we propose to use attention models for effective feature learning and develop two novel architectures, GrAMME-SG and GrAMME-Fusion, that exploit the interlayer dependences for building multilayered graph embeddings. Using empirical studies on several benchmark data sets, we evaluate the proposed approaches and demonstrate significant performance improvements in comparison with the state-of-the-art network embedding strategies. The results also show that using simple random features is an effective choice, even in cases where explicit node attributes are not available.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Two distinct rotations of bithiazole DNA intercalation revealed by direct comparison of crystal structures of Co(III)•bleomycin A 2 and B 2 bound to duplex 5'-TAGTT sites

Bleomycins constitute a family of anticancer natural products that bind DNA through intercalation of a C-terminal tail/bithiazole moiety and hydrogen-bonding interactions between the remainder of the drug and the minor groove. The clinical utility of the bleomycins is believed to result from single- and double-strand DNA cleavage mediated by the HOO-Fe(III) form of the drug. The bleomycins also serve as a model system to understand the nature of complex drug-DNA interactions that may guide future DNA-targeted drug discovery. In this study, the impact of the C-terminal tail on bleomycin-DNA interactions was investigated. Toward this goal, we determined two crystal structures of HOO-Co(III)•BLMA 2 “green” (a stable structural analogue of the active HOO-Fe(III) drug) bound to duplex DNA containing 5'-TAGTT, one in which the entire drug is bound (fully bound) and a second with only the C-terminal tail/bithiazole bound (partially bound). The structures reported here were captured by soaking HOO-Co(III)•BLMA 2 into preformed host–guest crystals including a preferred DNA-binding site. While the overall structure of DNA-bound BLMA 2 was found to be similar to those reported earlier at the same DNA site for BLMB 2 , the intercalated bithiazole of BLMB 2 is “flipped” 180° relative to DNA-bound BLMA 2 . Further, this finding highlights an unidentified role for the C-terminal tail in directing the intercalation of the bithiazole. In addition, these analyses identified specific bond rotations within the C-terminal domain of the drug that may be relevant for its reorganization and ability to carry out a double-strand DNA cleavage event.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2023

MtrA modulates Mycobacterium tuberculosis cell division in host microenvironments to mediate intrinsic resistance and drug tolerance

The success of Mycobacterium tuberculosis (Mtb) is largely attributed to its ability to physiologically adapt and withstand diverse localized stresses within host microenvironments. Here, we present a data-driven model (EGRIN 2.0) that captures the dynamic interplay of environmental cues and genome-encoded regulatory programs in Mtb. Analysis of EGRIN 2.0 shows how modulation of the MtrAB two-component signaling system tunes Mtb growth in response to related host microenvironmental cues. Disruption of MtrAB by tunable CRISPR interference confirms that the signaling system regulates multiple peptidoglycan hydrolases, among other targets, that are important for cell division. Further, MtrA decreases the effectiveness of antibiotics by mechanisms of both intrinsic resistance and drug tolerance. Together, the model-enabled dissection of complex MtrA regulation highlights its importance as a drug target and illustrates how EGRIN 2.0 facilitates discovery and mechanistic characterization of Mtb adaptation to specific host microenvironments within the host.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Discovering hidden geothermal signatures using non-negative matrix factorization with customized k-means clustering

Discovery of hidden geothermal resources is challenging. It requires the mining of large datasets with diverse data attributes representing subsurface hydrogeological and geothermal conditions. The commonly used play fairway analysis approach typically incorporates subject-matter expertise to analyze regional data to estimate geothermal characteristics and favorability. We demonstrate an alternative approach based on machine learning (ML) to process a geothermal dataset from southwest New Mexico (SWNM). The study region includes low- and medium-temperature hydrothermal systems. Several of these systems are not well characterized because of insufficient existing data and limited past explorative work. This study discovers hidden patterns and relations in the SWNM geothermal dataset to improve our understanding of the regional hydrothermal conditions and energy-production favorability. This understanding is obtained by applying an unsupervised ML algorithm based on non-negative matrix factorization coupled with customized k-means clustering (NMFk). NMFk can automatically identify (1) hidden signatures characterizing analyzed datasets, (2) the optimal number of these signatures, (3) the dominant data attributes associated with each signature, and (4) the spatial distribution of the extracted signatures. Here, in this study, NMFk is applied to analyze 18 geological, geophysical, hydrogeological, and geothermal attributes at 44 locations in SWNM. Using NMFk, we find data patterns and identify the spatial associations of hydrothermal signatures within two physiographic provinces (Colorado Plateau and Basin and Range) and two sub-regions of these provinces (the Mogollon-Datil volcanic field and the Rio Grande rift) in SWNM. The ML algorithm extracted five hydrothermal signatures in the SWNM datasets that differentiate between low (<90°C) and medium (90-150°C)-temperature hydrothermal systems. The algorithm also suggests that the Rio Grande rift and northern Mogollon-Datil volcanic field are the most favorable regions for future geothermal resource discovery. NMFk also identified critical attributes to identify medium-temperature hydrothermal systems in the study area. The resulting NMFk model can be applied to predict geothermal conditions and their uncertainties at new SWNM locations based on limited data from unexplored regions. The code to execute the performed analyses as well as the corresponding data can be found at https://github.com/SmartTensors/GeoThermalCloud.jl.

15 GEOTHERMAL ENERGY↗
DOE OSTI2020

Active- and transfer-learning applied to microscale-macroscale coupling to simulate viscoelastic flows

Active- and transfer-learning are applied to microscale dynamics of polymer flows for the multiscale discovery of effective constitutive approximations required in viscoelastic flow simulation. The result is macroscopic rheology directly connected to a microstructural model. Micro and macroscale simulations are adaptively coupled by means of Gaussian process regression (GPR) to run the expensive microscale computations only as necessary. This multiscale method is demonstrated with flows of a polymer solution as a model system. At the microscale level dissipative particle dynamics (DPD) is employed to model the fluid as a suspension of bead-spring micro-structures subjected to steady shear flow. The results yield the non-Newtonian viscosity and the first normal stress difference at strain rates as training data used in a GPR model. DPD parameters are calibrated with respect to experimental data for a real polymer solution. Compliance with these data requires adjustment of the DPD model's cutoff radius, which then becomes a function of the second invariant of the strain rate tensor. The FENE-P model is chosen for the macroscale description using the spectral element method (SEM) to simulate channel flow and flow past a circular cylinder. The DPD results at the lowest possible shear strain rate yield an estimate of the zero-shear rate viscosity, which allows the initiation of the macroscale flow by SEM as a Newtonian fluid. The resulting strain-rate field is surveyed to determine additional shear strain rate sampling points for the DPD system. This new information allows an initial fitting of parameters of the constitutive equation followed by new SEM simulations at the macroscale. Additionally, guided by active-learning GPR to select new sampling points, this process continues until convergence is achieved. The effectiveness of this new simulation paradigm for viscoelastic flows is tested with different macroscale operating conditions. The effective closure learned in the channel simulation is then transferred directly to the flow past a circular cylinder at low Reynolds number, where the results show that only two additional DPD simulations are required to achieve a satisfactory constitutive model. With an increase of the Reynolds number, the active-learning scheme automatically detects the inaccuracy of the learned constitutive model, and initiates additional DPD simulations for the extra data needed to once again close the microscale-macroscale coupled system. This new paradigm of active- and transfer-learning for multiscale modeling is readily applicable to other microscale-macroscale coupled simulations of complex fluids and other materials. Furthermore, the coupling between microscale and macroscale solvers can be seamlessly implemented with our open source multiscale universal interface (MUI) library.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2026

Tuning the Functionalities of Porous Liquids for Emergent Gas-Capture Properties

Type 3 Porous Liquids (PLs) are a class of materials with the potential to revolutionize gas capture, storage, and utilization. These PLs are formed by suspending sorbent nanoparticles (e.g., metal−organic frameworks) in sterically excluded solvents, creating permanent porosity for gas capture in a processable, low-viscosity phase. Herein, a computational study revealed sorbent surface functionalization strategies to enhance CO 2 sorption, and the molecular structural signatures underpinning the enhancements in gas uptake. PLs composed of a ZIF-8 surface functionalized with 3-amino-1,2,4-triazole (Atz) in glyceryl triacetate were targeted for emergent CO 2 capture, exceeding that of the unfunctionalized ZIF-8 PL. ZIF-8 was surface functionalized with Atz at various surface coverage fractions (f), and classical molecular dynamics simulations predicted an increase in CO 2 sorption capacity with increasing f, up to f = 0.75. Additionally, detailed structural analyses revealed that solvent orientational order, derived from the solvent triplet-angle distribution, can identify the gas-capture potential of a PL without requiring computationally expensive direct modeling of the CO 2 sorption. Combined with experimental validation, initial computational screening of PL compositions promises to accelerate the discovery of PL compositions for novel gas separation materials platforms.

gas capture↗
DOE OSTI2020

Phosphorus-rich stars with unusual abundances are challenging theoretical predictions

Almost all chemical elements have been made by nucleosynthetic reactions in various kind of stars and have been accumulated along our cosmic history. Among those elements, the origin of phosphorus is of extreme interest because it is known to be essential for life such as we know on Earth. However, current models of (Galactic) chemical evolution under-predict the phosphorus we observe in our Solar System. Here we report the discovery of 15 phosphorus-rich stars with unusual overabundances of O, Mg, Si, Al, and Ce. Phosphorus-rich stars likely inherit their peculiar chemistry from another nearby stellar source but their intriguing chemical abundance pattern challenge the present stellar nucleosynthesis theoretical predictions. Specific effects such as rotation or advanced nucleosynthesis in convective-reactive regions in massive stars represent the most promising alternatives to explain the existence of phosphorus-rich stars. The phosphorus-rich stars progenitors may significantly contribute to the phosphorus present on Earth today.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Highly transferable atomistic machine-learning potentials from curated and compact datasets across the periodic table

Machine learning atomistic potentials trained using density functional theory (DFT) datasets allow for the modeling of complex material properties with near-DFT accuracy while imposing a fraction of its computational cost. The curation of the DFT datasets can be extensive in size and time-consuming to train and refine. In this study, we focus on addressing these barriers by developing minimalistic and flexible datasets for many elements in the periodic table regardless of their mass, electronic configuration, and ground state lattice. These DFT datasets have on average, ~4000 different structures and 27 atoms per structure, which we found sufficient to maintain the predictive accuracy of DFT properties and notably with high transferability. We envision these highly curated training sets as starting points for the community to expand, modify, or use with other machine learning atomistic potential models, whatever may suit individual needs, further accelerating the utilization of machine learning as a tool for material design and discovery.

42 ENGINEERING↗
DOE OSTI2023

Electron scattering and neutrino physics

Abstract A thorough understanding of neutrino–nucleus scattering physics is crucial for the successful execution of the entire US neutrino physics program. Neutrino–nucleus interaction constitutes one of the biggest systematic uncertainties in neutrino experiments—both at intermediate energies affecting long-baseline deep underground neutrino experiment, as well as at low energies affecting coherent scattering neutrino program—and could well be the difference between achieving or missing discovery level precision. To this end, electron–nucleus scattering experiments provide vital information to test, assess and validate different nuclear models and event generators intended to test, assess and validate different nuclear models and event generators intended to be used in neutrino experiments. Similarly, for the low-energy neutrino program revolving around the coherent elastic neutrino–nucleus scattering (CEvNS) physics at stopped pion sources, such as at ORNL, the main source of uncertainty in the evaluation of the CEvNS cross section is driven by the underlying nuclear structure, embedded in the weak form factor, of the target nucleus. To this end, parity-violating electron scattering (PVES) experiments, utilizing polarized electron beams, provide vital model-independent information in determining weak form factors. This information is vital in achieving a percent level precision needed to disentangle new physics signals from the standard model expected CEvNS rate. In this white paper, we highlight connections between electron- and neutrino–nucleus scattering physics at energies ranging from 10 s of MeV to a few GeV, review the status of ongoing and planned electron scattering experiments, identify gaps, and lay out a path forward that benefits the neutrino community. We also highlight the systemic challenges with respect to the divide between the nuclear and high-energy physics communities and funding that presents additional hurdles in mobilizing these connections to the benefit of neutrino programs.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2020

To TDE or not to TDE: the luminous transient ASASSN-18jd with TDE-like and AGN-like qualities

ABSTRACT We present the discovery of ASASSN-18jd (AT 2018bcb), a luminous optical/ultraviolet(UV)/X-ray transient located in the nucleus of the galaxy 2MASX J22434289–1659083 at z = 0.1192. Over the year after discovery, Swift UltraViolet and Optical Telescope (UVOT) photometry shows the UV spectral energy distribution of the transient to be well modelled by a slowly shrinking blackbody with temperature $T \sim 2.5 \times 10^{4} \, {\rm K}$, a maximum observed luminosity of $L_{\rm max} = 4.5^{+0.6}_{-0.3}\times 10^{44} \, {\rm erg \,s}^{-1}$, and a radiated energy of $E = 9.6^{+1.1}_{-0.6} \times 10^{51} \, {\rm erg}$. X-ray data from Swift X-Ray Telescope (XRT) and XMM–Newton show a transient, variable X-ray flux with blackbody and power-law components that fade by nearly an order of magnitude over the following year. Optical spectra show strong, roughly constant broad Balmer emission and transient features attributable to He ii, N iii–v, O iii, and coronal Fe. While ASASSN-18jd shares similarities with tidal disruption events (TDEs), it is also similar to the newly discovered nuclear transients seen in quiescent galaxies and faint active galactic nuclei (AGNs).

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

Mitigating Catastrophic Forgetting in Deep Learning in a Streaming Setting Using Historical Summary

Recent advancements in scientific equipment and the adaptation of electronics and the Internet of Things (IoT) in our everyday lives resulted in large and complex data production at a high rate. Making meaningful and timely knowledge discovery at a modest cost from this big data is difficult for computing power and storage limitations. Training deep learning models incrementally in a streaming setting can help us with overcoming these limitations. However, in a well-known phenomenon named catastrophic forgetting, incrementally trained models increasingly perform poorly on the past data. To mitigate catastrophic forgetting in training in a streaming setting, we propose constructing a historical summary over time and use the summary with newly arrived data during incremental training. We propose various data summarization techniques such as random sampling, micro clustering, coreset computation, and Auto Encoders to counteract catastrophic forgetting. We built a pipeline for incremental training with a historical summary for training deep learning models for streaming data. We demonstrate the effectiveness of historical summary in mitigating catastrophic forgetting using three case studies involving three different deep learning applications: an Artificial Neural Network (ANN) for classification task on MNIST dataset, a language model (RNN-LM) on the WikiText2 dataset, and a Convolutional Neural Network (CNN), ResNet50 to classify the ImageNet dataset. Through the training of the models, we observe that catastrophic forgetting is evident in ANN and CNN but not in an RNN. For the first task, our method recovers up to 47.9% lost accuracy due to catastrophic forgetting. For the third task, the historical summary recovers classification accuracy by up to 25%. For the second task, though there is not proof of catastrophic forgetting, the training performance (PPL) improves by up to 26% with historical summary.

Dash, Sajal↗