Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “deep well”

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 415 records · Page 23
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Graphene U-Net v1

Graphene U-Net is a library that provides a simplified platform for training deep neural networks for the task of microscopy image segmentation. It contains functions and classes that make the process of training a neural network simple such that non-ML experts can train and evaluate models on their own datasets. It uses Pytorch as a backend and can run on both CUDA-enable GPUs and CPUs. It contains the main library file Microscopy_Unet.py as well as unet.py which contains the UNET model used for this software. This can be replaced with any other fully convolutional deep learning architecture with relative ease.

Sadre, Robbie↗
DOE OSTI2020

Deep Generative Models that Solve PDEs: Distributed Computing for Training Large Data-Free Models

Recent progress in scientific machine learning (SciML) has opened up the possibility of training novel neural network architectures that solve complex partial differential equations (PDEs). Several (nearly data free) approaches have been recently reported that successfully solve PDEs, with examples including deep feed forward networks, generative networks, and deep encoder-decoder networks. However, practical adoption of these approaches is limited by the difficulty in training these models, especially to make predictions at large output resolutions (≥1024×1024). Here we report on a software framework for data parallel distributed deep learning that resolves the twin challenges of training these large SciML models - training in reasonable time as well as distributing the storage requirements. Our framework provides several out of the box functionality including (a) loss integrity independent of number of processes, (b) synchronized batch normalization, and (c) distributed higher-order optimization methods. We show excellent scalability of this framework on both cloud as well as HPC clusters, and report on the interplay between bandwidth, network topology and bare metal vs cloud. We deploy this approach to train generative models of sizes hitherto not possible, showing that neural PDE solvers can be viably trained for practical applications. We also demonstrate that distributed higher-order optimization methods are 2-3× faster than stochastic gradient-based methods and provide minimal convergence drift with higher batch-size.

PDEs↗
DOE OSTI2022

Differentiable, Learnable, Regionalized Process-Based Models With Multiphysical Outputs can Approach State-Of-The-Art Hydrologic Prediction Accuracy

Predictions of hydrologic variables across the entire water cycle have significant value for water resources management as well as downstream applications such as ecosystem and water quality modeling. Recently, purely data-driven deep learning models like long short-term memory (LSTM) showed seemingly insurmountable performance in modeling rainfall runoff and other geoscientific variables, yet they cannot predict untrained physical variables and remain challenging to interpret. Here, we show that differentiable, learnable, process-based models (called δ models here) can approach the performance level of LSTM for the intensively observed variable (streamflow) with regionalized parameterization. We use a simple hydrologic model HBV as the backbone and use embedded neural networks, which can only be trained in a differentiable programming framework, to parameterize, enhance, or replace the process-based model's modules. Without using an ensemble or post-processor, δ models can obtain a median Nash-Sutcliffe efficiency of 0.732 for 671 basins across the USA for the Daymet forcing data set, compared to 0.748 from a state-of-the-art LSTM model with the same setup. For another forcing data set, the difference is even smaller: 0.715 versus 0.722. Meanwhile, the resulting learnable process-based models can output a full set of untrained variables, for example, soil and groundwater storage, snowpack, evapotranspiration, and baseflow, and can later be constrained by their observations. Both simulated evapotranspiration and fraction of discharge from baseflow agreed decently with alternative estimates. The general framework can work with models with various process complexity and opens up the path for learning physics from big data.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

AI-Improved Resolution Projections of Population Characteristics and Imperviousness Can Improve Resolution and Accuracy of Urban Flood Predictions

Focal Area: (2) Predictive modeling through the use of AI techniques and AI-derived model components. Specifically, we call for the use of deep generative models from ML for providing high-resolution projections of impervious surface area, as well as neural network solvers for fast approximation of urban hydrodynamics to provide greatly enhanced forecasts of future urban flood dynamics.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Event Selection in the MicroBooNE Deep Learning Based Low Energy Excess Analysis Using Two-Body Scattering Criteria

The uniquely detailed neutrino event information from liquid argon time projection chambers allows reconstruction of a set of kinematic quantities that over-constrain the expectations for charged current quasielastic scattering (CCQE). MicroBooNE makes use of the CCQE consistency requirements in a deep-learning-based search for the MiniBooNE low energy excess analysis. This requirement rejects backgrounds as well as events with poorly reconstructed neutrino energy due to final state interactions of the outgoing proton. The results presented here demonstrate the quality of the selection of $\nu_e$ and $\nu_µ$ events. We show excellent agreement between the data and the simulation across many data sets. This positions us to be ready to unblind the MicroBooNE low energy excess analysis in the very near future.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Understanding the High Energy Higgs Sector with the CMS Experiment and Artificial Intelligence

This dissertation describes efforts towards understanding the Higgs boson at the highest energies humanly accessible, using the CMS experiment at the Large Hadron Collider and advances in artificial intelligence (AI) and machine learning (ML). We present searches for resonant and nonresonant Higgs-boson (H) pair production in the all-hadronic two beauty-quark and two vector boson (V) final state, using a novel strategy to measure the quartic HHVV coupling and search for new Higgs-like bosons. By targeting highly Lorentz-boosted Higgs pairs, we probe effects of potential new physics in the high energy Higgs sector, which could hold answers to fundamental mysteries of nature such as baryon asymmetry. To enable these and future searches, we introduce as well significant developments in AI/ML, including in the identification of boosted H$\rightarrow$VV decays with deep transformer networks and advances in AI-accelerated fast simulations of the CMS detector. The latter notably includes the development of the first, highly performant generative models for point-cloud data in high energy physics, which have the potential to improve CMS' computational efficiency by up to three orders of magnitude. We also highlight novel solutions to the important and challenging problems of calibrating and validating these ML techniques. Finally, we present new approaches to search for new physics in a model-agnostic manner, using physics-informed ML methods equivariant to Lorentz transformations. The quartic HHVV coupling is observed (expected) to be constrained to $[-0.04, 2.05]$ ($[0.05, 1.98]$) at the 95% confidence level relative to the standard model prediction, representing the second-most sensitive measurement of this coupling by CMS to date. Exclusion limits on the production cross section of new heavy resonances decaying to two Higgs-like bosons are expected to be as low as 0.3 fb for high resonance masses.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2022

Salinity and hydraulic retention time induce membrane phospholipid acyl chain remodeling in Halanaerobium congolense WG10 and mixed cultures from hydraulically fractured shale wells

Bacteria remodel their plasma membrane lipidome to maintain key biophysical attributes in response to ecological disturbances. For Halanaerobium and other anaerobic halotolerant taxa that persist in hydraulically fractured deep subsurface shale reservoirs, salinity, and hydraulic retention time (HRT) are important perturbants of cell membrane structure, yet their effects remain poorly understood. Membrane-linked activities underlie in situ microbial growth kinetics and physiologies which drive biogeochemical reactions in engineered subsurface systems. Hence, we used gas chromatography–mass spectrometry (GC–MS) to investigate the effects of salinity and HRT on the phospholipid fatty acid composition of H. congolense WG10 and mixed enrichment cultures from hydraulically fractured shale wells. We also coupled acyl chain remodeling to membrane mechanics by measuring bilayer elasticity using atomic force microscopy (AFM). For these experiments, cultures were grown in a chemostat vessel operated in continuous flow mode under strict anoxia and constant stirring. Our findings show that salinity and HRT induce significant changes in membrane fatty acid chemistry of H. congolense WG10 in distinct and complementary ways. Notably, under nonoptimal salt concentrations (7% and 20% NaCl), H. congolense WG10 elevates the portion of polyunsaturated fatty acids (PUFAs) in its membrane, and this results in an apparent increase in fluidity (homeoviscous adaptation principle) and thickness. Double bond index (DBI) and mean chain length (MCL) were used as proxies for membrane fluidity and thickness, respectively. These results provide new insight into our understanding of how environmental and engineered factors might disrupt the physical and biogeochemical equilibria of fractured shale by inducing physiologically relevant changes in the membrane fatty acid chemistry of persistent microbial taxa. GRAPHICAL ABSTRACT Salinity significantly alters membrane bilayer fluidity and thickness in Halanaerobium congolense WG10.

03 NATURAL GAS↗
DOE OSTI2022

Deep learning-driven insights into super protein complexes for outer membrane protein biogenesis in bacteria

To reach their final destinations, outer membrane proteins (OMPs) of gram-negative bacteria undertake an eventful journey beginning in the cytosol. Multiple molecular machines, chaperones, proteases, and other enzymes facilitate the translocation and assembly of OMPs. These helpers usually associate, often transiently, forming large protein assemblies. They are not well understood due to experimental challenges in capturing and characterizing protein-protein interactions (PPIs), especially transient ones. Using AF2Complex, we introduce a high-throughput, deep learning pipeline to identify PPIs within the Escherichia coli cell envelope and apply it to several proteins from an OMP biogenesis pathway. Among the top confident hits obtained from screening ~1500 envelope proteins, we find not only expected interactions but also unexpected ones with profound implications. Subsequently, we predict atomic structures for these protein complexes. These structures, typically of high confidence, explain experimental observations and lead to mechanistic hypotheses for how a chaperone assists a nascent, precursor OMP emerging from a translocon, how another chaperone prevents it from aggregating and docks to a β-barrel assembly port, and how a protease performs quality control. This work presents a general strategy for investigating biological pathways by using structural insights gained from deep learning-based predictions.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2023

Energy Assets Transformation Web Mapping Application

This submission contains the link and geospatial materials used in the Energy Assets Transformation Web Mapping Application. The zip file contains 19 geospatial layers in a file geodatabase called EAT.gdb to be grouped in the following categories. 1. Industrial Assets: Coal Generation Units Retirements 2012-2040 (EIA); Examples of Repurposing Projects (32 projects in total); Abandoned Coal Mines (CORD, SkyTruth); Abandoned or Orphaned Wells (for ten states only). 2. Energy Transition Communities: 48C (e) Tax Credits - Designated Energy Communities (IRA); Index of Deep Disadvantage; Local Energy Action Program (LEAP); EJ Index for Proximity to Hazardous Waste (EPA). 3. Regional Landscape: State-Level Funding Programs (relevant to repurposing projects, for 2022 and 2023 only); Coal Flows from Mine to Plant 2021 (EIA), Variable Renewable Energy Shares (Wind and Solar, 2021, EIA). 4. Supporting Infrastructure: Railroads (HIFLD), Electric Power Transmission Lines (HIFLD), Major Highways (NHPN, DOT), Major Ports (National Atlas of the U.S.); Independent System Operators (HIFLD), NERC Regions and Subregions (HIFLD).

abandoned coal mines↗
DOE OSTI2022

Learning and Fast Adaptation for Grid Emergency Control via Deep Meta Reinforcement Learning

As power systems are undergoing a significant transformation with more uncertainties, less inertia and closer to operation limits, there is increasing risk of large outages. Thus, there is an imperative need to enhance grid emergency control to maintain system reliability and security. Towards this end, great progress has been made in developing deep reinforcement learning (DRL) based grid control solutions in recent years. However, existing DRL-based solutions have two main limitations: 1) they cannot handle well with a wide range of grid operation conditions, system parameters, and contingencies; 2) they generally lack the ability to fast adapt to new grid operation conditions, system parameters, and contingencies, limiting their applicability for real-world applications. Here, in this paper, we mitigate these limitations by developing a novel deep meta-reinforcement learning (DMRL) algorithm. The DMRL combines the meta strategy optimization together with DRL, and trains policies modulated by a latent space that can quickly adapt to new scenarios. We test the developed DMRL algorithm on the IEEE 300-bus system. We demonstrate fast adaptation of the meta-trained DRL polices with latent variables to new operating conditions and scenarios using the proposed method, which achieves superior performance compared to the state-of-the-art DRL and model predictive control (MPC) methods.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2021

Deep Search for Decaying Dark Matter with XMM-Newton Blank-Sky Observations

Sterile neutrinos with masses in the keV range are well-motivated extensions to the Standard Model that could explain the observed neutrino masses while also making up the dark matter (DM) of the universe. If sterile neutrinos are DM then they may slowly decay into active neutrinos and photons, giving rise to the possibility of their detection through narrow spectral features in astrophysical x-ray data sets. Here, we perform the most sensitive search to date for this and other decaying DM scenarios across the mass range from 5 to 16 keV using archival XMM-Newton data. We reduce 547 Ms of data from both the MOS and PN instruments using observations taken across the full sky and then use this data to search for evidence of DM decay in the ambient halo of the Milky Way. We determine the instrumental and astrophysical baselines with data taken far away from the Galactic Center, and use Gaussian process modeling to capture additional continuum background contributions. No evidence is found for unassociated x-ray lines, leading us to produce the strongest constraints to date on decaying DM in this mass range.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

CaMn 0.9 Ti 0.1 O 3 based redox catalysts for chemical looping – Oxidative dehydrogenation of ethane: Effects of Na 2 MoO 4 promoter and degree of reduction on the reaction kinetics

Reduction kinetics and stability of 20 wt% Na 2 MoO 4 -promoted CaMn 0.9 Ti 0.1 O 3 were investigated for its applications in Chemical Looping – Oxidative Dehydrogenation (CL-ODH) of ethane, a potential alternative for ethylene production with higher efficiency and lower emissions. Here, the present work reports a kinetics model and parameters for a Na 2 MoO 4 -promoted, Ti-doped CaMnO 3 (CaMn 0.9 Ti 0.1 O 3 ) redox catalyst under H 2 and C 2 H 4 . A first-order reaction model provides the best fit for the reduction of Na 2 MoO 4 /CaMn 0.9 Ti 0.1 O 3 under H 2 , while the C 2 H 4 reduction is well described by an Avrami–Erofe’ev model. The activation energy for C 2 H 4 oxidation is approximately three times higher than that for H 2 conversion, showing that the activation of C 2 H 4 is significantly more difficult on the surface of the redox catalyst. The reduction rate of Na 2 MoO 4 /CaMn 0.9 Ti 0.1 O 3 under H 2 at 750 °C is more than two orders of magnitude greater than that under C 2 H 4 , while the reduction rate of unpromoted CaMn 0.9 Ti 0.1 O 3 is comparable under H 2 and C 2 H 4 , showing that the addition of Na 2 MoO 4 effectively suppresses C 2 H 4 combustion relative to H 2 oxidation. The kinetics results for Na 2 MoO 4 /CaMn 0.9 Ti 0.1 O 3 confirm its excellent selectivity towards hydrogen combustion, making it a promising candidate under CL-ODH. Additionally, the stability of the CaMn 0.9 Ti 0.1 O 3 @ Na 2 MoO 4 core-shell structure, which was the underlying reason for the excellent selectivity, was examined under both shallow and deep reductions. It was determined that deep reduction of the redox catalyst, e.g. higher than 80% solid conversion, would lead to loss of sodium and hence to decreased selectivity for hydrogen combustion. In contrast, the core-shell structure was well-maintained, exhibiting excellent performance after 50 redox cycles when deep reduction of the redox catalyst was avoided. This study offers a basis for both the CL-ODH reactor design and redox catalyst optimizations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Learning physics at future $e^–e^+$ colliders with machine

Information deformation and loss in jet clustering are one of the major limitations for precisely measuring hadronic events at future $e^–e^+$ colliders. Because of their dominance in data, the measurements of such events are crucial for advancing the precision frontier of Higgs and electroweak physics in the next decades. We show that this difficulty can be well-addressed by synergizing the event-level information into the data analysis, with the techniques of deep neutral network. In relation to this, we introduce a CMB-like observable scheme, where the event-level kinematics is encoded as Fox-Wolfram (FW) moments at leading order and multi-spectra of spherical harmonics at higher orders. Then we develop a series of jet-level (w/ and w/o the FW moments) and event-level classifiers, and analyze their sensitivity performance comparatively with two-jet and four-jet events. As an application, we analyze measuring Higgs decay width at $e^–e^+$ colliders with the data of 5ab –1 @240GeV. The precision obtained is significantly better than the baseline ones presented in documents. We expect this strategy to be applied to many other hadronic- event measurements at future $e^–e^+$ colliders, and to open a new angle for evaluating their physics capability.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2022

Dissipation and Bathymetric Sensitivities in an Unstructured Mesh Global Tidal Model

Abstract The mechanisms and geographic distribution of global tidal dissipation in barotropic tidal models are examined using a high resolution unstructured mesh finite element model. Mesh resolution varies between 2 and 25 km and is especially focused on inner shelves and steep bathymetric gradients. Tidal response sensitivities to bathymetric changes are examined to put into context response sensitivities to frictional processes. We confirm that the Ronne Ice Shelf dramatically affects Atlantic tides but also find that bathymetry in the Hudson Bay system is a critical control. We follow a sequential frictional parameter optimization process and use TPXO9 data‐assimilated tidal elevations as a reference solution. From simulated velocities and depths, dissipation within the global model is estimated and allows us to pinpoint dissipation at high resolution. Boundary layer dissipation is extremely focused with 1.4% of the ocean accounting for 90% of the total. Internal tide friction is much more distributed with 16.7% of the ocean accounting for 90% of the total. Often highly regional dissipation can impact basin‐scale and even ocean wide tides. Optimized boundary layer friction parameters correlate very well with the physical characteristics of the locality with high friction factors associated with energetic tidal regions, deep ocean island chains, and ice covered areas. Global complex M 2 tide errors are 1.94 cm in deep waters. Total global boundary layer and internal tide dissipation are estimated, respectively, at 1.83 and 1.49 TW. This continues the trend in the literature toward attributing more dissipation to internal tides.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Revealing the closed pore formation of waste wood-derived hard carbon for advanced sodium-ion battery

Abstract Although the closed pore structure plays a key role in contributing low-voltage plateau capacity of hard carbon anode for sodium-ion batteries, the formation mechanism of closed pores is still under debate. Here, we employ waste wood-derived hard carbon as a template to systematically establish the formation mechanisms of closed pores and their effect on sodium storage performance. We find that the high crystallinity cellulose in nature wood decomposes to long-range carbon layers as the wall of closed pore, and the amorphous component can hinder the graphitization of carbon layer and induce the crispation of long-range carbon layers. The optimized sample demonstrates a high reversible capacity of 430 mAh g −1 at 20 mA g −1 (plateau capacity of 293 mAh g −1 for the second cycle), as well as good rate and stable cycling performances (85.4% after 400 cycles at 500 mA g −1 ). Deep insights into the closed pore formation will greatly forward the rational design of hard carbon anode with high capacity.

25 ENERGY STORAGE↗
DOE OSTI2023

Time-Resolved Line Shapes of Single Quantum Emitters via Machine Learned Photon Correlations

Solid-state single-photon emitters (SPEs) are quantum light sources that combine atomlike optical properties with solid-state integration and fabrication capabilities. SPEs are hindered by spectral diffusion, where the emitter’s surrounding environment induces random energy fluctuations. Timescales of spectral diffusion span nanoseconds to minutes and require probing single emitters to remove ensemble averaging. Photon correlation Fourier spectroscopy (PCFS) can be used to measure time-resolved single emitter line shapes, but is hindered by poor signal-to-noise ratio in the measured correlation functions at early times due to low photon counts. Here, we develop a framework to simulate PCFS correlation functions directly from diffusing spectra that match well with experimental data for single colloidal quantum dots. We use these simulated datasets to train a deep ensemble autoencoder machine learning model that outputs accurate, noiseless, and probabilistic reconstructions of the noisy correlations. Using this model, we obtain reconstructed time-resolved single dot emission line shapes at timescales as low as 10 ns, which are otherwise completely obscured by noise. This enables PCFS to extract optical coherence times on the same timescales as Hong-Ou-Mandel two-photon interference, but with the advantage of providing spectral information in addition to estimates of photon indistinguishability. Further, our machine learning approach is broadly applicable to different photon correlation spectroscopy techniques and SPE systems, offering an enhanced tool for probing single emitter line shapes on previously inaccessible timescales.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2020

Biosystems Design to Accelerate C 3 -to-CAM Progression

Global demand for food and bioenergy production has increased rapidly, while the area of arable land has been declining for decades due to damage caused by erosion, pollution, sea level rise, urban development, soil salinization, and water scarcity driven by global climate change. In order to overcome this conflict, there is an urgent need to adapt conventional agriculture to water-limited and hotter conditions with plant crop systems that display higher water-use efficiency (WUE). Crassulacean acid metabolism (CAM) species have substantially higher WUE than species performing C 3 or C 4 photosynthesis. CAM plants are derived from C 3 photosynthesis ancestors. However, it is extremely unlikely that the C 3 or C 4 crop plants would evolve rapidly into CAM photosynthesis without human intervention. Currently, there is growing interest in improving WUE through transferring CAM into C 3 crops. However, engineering a major metabolic plant pathway, like CAM, is challenging and requires a comprehensive deep understanding of the enzymatic reactions and regulatory networks in both C 3 and CAM photosynthesis, as well as overcoming physiometabolic limitations such as diurnal stomatal regulation. Recent advances in CAM evolutionary genomics research, genome editing, and synthetic biology have increased the likelihood of successful acceleration of C 3 -to-CAM progression. Here, we first summarize the systems biology-level understanding of the molecular processes in the CAM pathway. Then, we review the principles of CAM engineering in an evolutionary context. Lastly, we discuss the technical approaches to accelerate the C 3 -to-CAM transition in plants using synthetic biology toolboxes.

59 BASIC BIOLOGICAL SCIENCES↗