SEARCH · Engineering Papers

Results for “Samples”

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 775 records · Page 43

All sources NASA NTRS DOE OSTI

DOE OSTI2020

Sampling Subjective Polygons for Patch-Based Deep Learning Land-Use Classification in Satellite Images

Model generalization remains a key challenge in the analysis of large amounts of heterogeneous satellite image data. One major limiting factor in developing generalizable models, in the context of supervised learning, is the lack of high quality training datasets. A model's capacity to perform well on new data is often inhibited by imbalance and bias in the data that was used for training. This is especially a problem when using convolutional neural networks to classify urban land-use in satellite images. Notable dataset imbalance issues in this application include land-use type imbalance and image scene imbalance. To begin understanding these dataset imbalance problems in more detail, we develop and test a number of sampling methods for generating training image datasets from subjective training polygons for urban land-use classification. We investigate sampling at different point densities as a means to reduce content repetition and therefore content imbalance and bias in the training image dataset.

Arndt, Jacob↗

DOE OSTI2026

Optimal Client Sampling in Federated Learning with Client-level Heterogeneous Differential Privacy

Federated Learning with client-level differential privacy (DP) provides a promising framework for collaboratively training models while rigorously protecting clients’ privacy. However, classic approaches like DP-FedAvg struggle when clients have heterogeneous privacy requirements, as they must uniformly enforce the strictest privacy level across all clients, leading to excessive DP noise and significant degradation in model utility. Existing methods to improve the model utility in such heterogeneous privacy settings often assume a trusted server and are largely heuristic, resulting in suboptimal performance and lacking strong theoretical foundations. Here, in this work, we address these challenges under a practical attack model where both clients and the server are honest-but-curious. We propose GDPFed, which partitions clients into groups based on their privacy budgets and achieves client-level DP within each group to reduce the privacy budget waste and hence improve the model utility. Based on the privacy and convergence analysis of GDPFed, we find that the magnitude of DP noise depends on both model dimensionality and the per-group client sampling ratios. To further improve the performance of GDPFed, we introduce GDPFed+, which integrates model sparsification to eliminate unnecessary noise and optimizes per-group client sampling ratios to minimize convergence error. Extensive empirical evaluations on multiple benchmark datasets demonstrate the effectiveness of GDPFed+, showing substantial performance gains compared with state-of-the-art methods.

Xu, Jiahao [Univ. of Nevada, Reno, NV (United Stat↗

DOE OSTI2021

Encoding Frequency Constraints in Preventive Unit Commitment Using Deep Learning With Region-of-Interest Active Sampling

With the increasing penetration of renewable energy, frequency response and its security are of significant concerns for reliable power system operations. Frequency-constrained unit commitment (FCUC) is proposed to address this challenge. Despite existing efforts in modeling frequency characteristics in unit commitment (UC), current strategies can only handle oversimplified low-order frequency response models and do not consider wide-range operating conditions. This paper presents a generic data-driven framework for FCUC under high renewable penetration. Here, deep neural networks (DNNs) are trained to predict the frequency response using real data or high-fidelity simulation data. Next, the DNN is reformulated as a set of mixed-integer linear constraints to be incorporated into the ordinary UC formulation. In the data generation phase, all possible power injections are considered, and a region-of-interest active sampling is proposed to include power injection samples with frequency nadirs closer to the UFLC threshold, which enhances the accuracy of frequency constraints in FCUC. The proposed FCUC is investigated on the IEEE 39-bus system. Then, a full-order dynamic model simulation using PSS/E verifies the effectiveness of FCUC in frequency-secure generator commitments.

42 ENGINEERING↗

DOE OSTI2026

Smaller feature size impairs densification in digital light processing 3D printed ceramic samples

In this study, we investigate how geometric miniaturization affects the sintering densification behavior of alumina cylinders with various wall thicknesses (down to ∼500 µm in green body) fabricated via digital light processing (DLP). As the wall thickness decreased, the density of the sintered samples consistently declined. Microstructural characterization revealed the presence of a crack-rich surface skin that is ∼500 µm thick. As the wall thickness decreases, the skin's volume fraction rises, and its poor local densification increasingly dominates the part average, leading to an overall decrease in the density of the sintered samples. For the thinnest wall (500 µm) cylinder, macroscopic warping occurred during sintering. We suggest that warping further disrupts particle packing and induces through-thickness nonuniformity, further impairing densification in addition to the skin-fraction effect. These results highlight how geometric miniaturization may adversely influence the sintering behavior and final density of DLP-printed ceramics through distinct densification-limiting mechanisms.

Materials science↗

DOE OSTI2023

Mineralogy and petrology of fine‐grained samples recovered from the asteroid (162173) Ryugu

Abstract Samples returned from the carbonaceous asteroid (162173) Ryugu by the Hayabusa2 mission revealed that Ryugu is composed of materials consistent with CI chondrites and some types of space weathering. We report detailed mineralogy of the fine‐grained Ryugu samples allocated to our “Sand” team and report additional space weathering features found on the grains. The dominant mineralogy is composed of a fine‐grained mixture of Mg‐rich saponite and serpentine, magnetite, pyrrhotite, pentlandite, dolomite, and Fe‐bearing magnesite. These grains have mineralogy comparable to that of CI chondrites, showing severe aqueous alteration but lacking ferrihydrite and sulfate. These results are similar to previous works on large Ryugu grains. In addition to the major minerals, we also find many minerals that are rare or have not been reported among CI chondrites. Accessory minerals identified are hydroxyapatite, Mg‐Na phosphate, olivine, low‐Ca pyroxene, Mg‐Al spinel, chromite, manganochromite, eskolaite, ilmenite, cubanite, polydymite, transjordanite, schreibersite, calcite, moissanite, and poorly crystalline phyllosilicate. We also show scanning transmission electron microscope and scanning electron microscope compositional maps and images of some space‐weathered grains and severely heated and melted grains. Although our mineralogical results are consistent with that of millimeter‐sized grains, the fine‐grained fraction is best suited to investigate impact‐induced space weathering.

Geochemistry & Geophysics↗

DOE OSTI2023

Preparation and Characterization Methods of Thin Layer Samples for Standoff Detection

Detection of analytes deposited on surfaces is crucial for many applications: Development of methods to prepare thin layers (e.g. ~5 to 100 µm) is important for both system design and field studies. In this work, solid and liquid analytes were deposited on painted and bare substrates including aluminum, glass, plastic, and concrete using an ExactaCoat ultrasonic spray coater. Laboratory hemispherical reflectance (HRF) spectra were collected for samples with different layer thicknesses so as to characterize both the composition and layer thickness. Preliminary results demonstrate that to prepare homogenous layers on surfaces, parameters such as substrate type, analyte solubility, vapor pressure, paint color, surface porosity, and surface roughness are all important. Liquid chemicals posed several issues during deposition: Diisopropyl methyl phosphonate evaporated from surfaces more quickly than the other chemicals and was thus not detected in the HRF experiments. Less volatile liquids, such as tributylphosphate, remained on the surface for the duration of the test, but a uniform layer thickness could not be obtained as the liquid pooled to one side when mounted at an angle. The deposition of solids (e.g., acetaminophen, caffeine and methylphosphonic acid) from volatile solvents such as chloroform also proved problematic due to streaking caused by rapid solvent evaporation. Solids deposited from ethanol, however, worked well on bare substrates. For most samples plotting the integrated infrared band strength vs. surface thicknesses showed a linear relationship, confirming that the surface loading can be controlled by programming the concentration and the number of passes on the ultrasonic sprayer.

Thin layer, deposition, infrared standoff, Hemisph↗

DOE OSTI2025

An infrasound source analysis of the OSIRIS-REx sample return capsule hypersonic re-entry

The OSIRIS-REx sample return capsule's hypersonic re-entry into the atmosphere is a rare opportunity to test a variety of sonic boom source models since the projectile dimensions are well characterized. While the as-flown flight path is unknown, the predicted flight path enables a rough approximation of the source Mach number and location. Six infrasound microphones deployed in the boom carpet along the predicted flight path recorded impulsive signals from the OSIRIS-REx re-entry. Using a suite of atmosphere profiles and the geometric acoustics approximation, we estimate locations with uncertainty estimates along the flight path from which the signals were emitted. Acoustic overpressure and signal duration predictions from Whitham's far field theory, Carlson's simplified sonic boom prediction method, and a drag-dominated hypersonic model are analyzed with uncertainty estimates from the location estimate. While the Carlson simplified sonic boom prediction method could be accurate, our preference is for the drag-dominated source model. Using this source model with an inviscid Burgers's equation solver for propagation, we obtained an excellent match to the recorded data. In conclusion, these results will help better inform future sample return capsule re-entry observation campaigns as well as contribute to a better understanding of high altitude infrasonic sources.

58 GEOSCIENCES↗

DOE OSTI2023

Polycyclic aromatic hydrocarbons in samples of Ryugu formed in the interstellar medium

Polycyclic aromatic hydrocarbons (PAHs) contain ≲20% of the carbon in the interstellar medium. They are potentially produced in circumstellar environments (at temperatures ≳1000 kelvin), by (~10 kelvin) interstellar clouds, or by processing of carbon-rich dust grains. We report isotopic properties of PAHs extracted from samples of the asteroid Ryugu and the meteorite Murchison. The doubly- 13 C substituted compositions (Δ2× 13 C values) of the PAHs naphthalene, fluoranthene, and pyrene are 9 to 51‰ higher than values expected for a stochastic distribution of isotopes. The Δ2× 13 C values are higher than expected if the PAHs formed in a circumstellar environment, but consistent with formation in the interstellar medium. By contrast, the PAHs phenanthrene and anthracene in Ryugu samples have Δ2× 13 C values consistent with formation by higher-temperature reactions.

Science & Technology - Other Topics↗

DOE OSTI2025

Efficient Training of Deep Neural Operator Networks via Randomized Sampling

Neural operators (NOs) employ deep neural networks to learn the mappings between infinitedimensional function spaces. Deep operator network (DeepONet), a popular NO architecture, has demonstrated success in the real-time prediction of complex dynamics across various scientific and engineering applications. In this work, we introduce a random sampling technique to be adopted during the training of DeepONet, aimed at improving the generalization ability of the model, while significantly reducing the computational time. The proposed approach targets the trunk network of the DeepONet model that outputs the basis functions corresponding to the spatiotemporal locations of the bounded domain on which the physical system is defined. While constructing the loss function, DeepONet training traditionally considers a uniform grid of spatiotemporal points at which all the output functions are evaluated for each iteration. This approach leads to a larger batch size, resulting in poor generalization and increased memory demands, due to the limitations of the stochastic gradient descent (SGD) optimizer. The proposed random sampling over the inputs of the trunk net mitigates these challenges, improving generalization and reducing the memory requirements during training, resulting in significant computational gains. We validate our hypothesis through three benchmark examples, demonstrating substantial reductions in training time while achieving comparable or lower overall test errors relative to the traditional training approach. Our results indicate that incorporating randomization in the trunk network inputs during training enhances the efficiency and robustness of DeepONet, offering a promising avenue for improving the framework’s performance in modeling complex physical systems.

Karumuri, Sharmila [Department of Civil & Systems ↗

DOE OSTI2024

Multivariate Testing of Sampling Techniques to Address Class Imbalance in Building Use Type Classification

This study addresses the challenges inherent in building use type classification, particularly focusing on the issue of class imbalance in the training datasets for machine learning classifiers. We comprehensively analyze the efficacy of various class-balancing sampling techniques. Employing Monte Carlo simulations and Bayesian optimization, we evaluated the performance of multiple sampling methods, including Random Oversampling, Random Undersampling, SMOTE, Borderline-SMOTE, and ADASYN, across a dataset encompassing nine southeastern coastal states of the United States. Our findings reveal that simple random over- and undersampling techniques outperform more sophisticated methods. Additionally, we show inherent value in creating an imbalance in training data to effectively train a machine learning classifier for distinguishing between residential and nonresidential buildings. This study provides valuable guidance for future research on building use type classification research and lays essential groundwork for developing attribute-rich building stock datasets.

Adams, Daniel↗

DOE OSTI2025

Intelligent Sampling of Extreme-Scale Turbulence Datasets for Accurate and Efficient Spatiotemporal Model Training

With the end of Moore’s law and Dennard scaling, efficient training increasingly requires rethinking data volume. Can we train better models with significantly less data via intelligent subsampling? To explore this, we develop SICKLE, a sparse intelligent curation framework for efficient learning, featuring a novel maximum entropy (MaxEnt) sampling approach, scalable training, and energy benchmarking. We compare MaxEnt with random and phase-space sampling on large direct numerical simulation (DNS) datasets of turbulence. Evaluating SICKLE at scale on Frontier, we show that subsampling as a preprocessing step can, in many cases, improve model accuracy and substantially lower energy consumption, with observed reductions of up to 38×.

Brewer, Wes [ORNL] (ORCID:0000000236393956)↗

DOE OSTI2022

Direct-Sampling Beam Position Monitor (dsbpm) v1.0

Direct-Sampling Beam Position Monitor is an FPGA gateware and embedded firmware that run on evaluation kits (either Xilinx ZCU111 or Xilinx ZCU208 ) that digitizes signals up to 4GSPS and 4GHz analog bandwidth, which can be synchronized to the accelerator timing system, providing synchronous sampling frequency/phase digitization. In addition to these capabilities, a proof of concept for a beam position monitor is also present, providing an envelope detection algorithm capable of determining the transverse beam position for an electron circular accelerator, such as ALS/ALS-U.

Norum, William↗

DOE OSTI2024

DISSIDE: Dynamic In Silico Sample Identification for Discrete Evaluation

DISSIDE uses novel unsupervised learning to select samples that best represent the strongest a priori discrete pattern in a given data set. It further removes major outliers and "noisy" samples that do not fit discrete patterns well or represent outliers in groups below a defined n value. It then estimates the fit and strength of the a priori discrete pattern using both unconstrained and constrained methods for raw and cleaned data

LeBrun, Erick↗

DOE OSTI2025

Multiscale Molecular Dynamics Simulations: Accelerating Conformational Sampling of Biomolecular Systems by Iterating All-Atom and Coarse-Grained Simulations

We developed the atomistic-coarse-grained multiscale MD simulation method in the OpenMM simulation package by iterating between the all-atom (AA) and coarse-grained (CG) MD simulations to enhance the sampling of biomolecular conformations. As the free energy surfaces are flattened during CG MD simulations, we can accelerate the transitions between different low-energy conformations. The AA-CG-AA cycles are repeated, facilitating the accelerated sampling of biomolecular conformations at a CG level, while the finer atomistic interactions are refined with AA simulators.

Do, Hung Nguyen↗

DOE OSTI2023

Multisensor Agile Adaptive Sampling (MAAS): A Methodology to Collect Radar Observations of Convective Cell Life Cycle

Abstract Multisensor Agile Adaptive Sampling (MAAS), a smart sensing framework, was adapted to increase the likelihood of observing the vertical structure (with little to no gaps), spatial variability (at subkilometer scale), and temporal evolution (at ∼2-min resolution) of convective cells. This adaptation of MAAS guided two mechanically scanning C-band radars (CSAPR2 and CHIVO) by automatically analyzing the latest NEXRAD data to identify, characterize, track, and nowcast the location of all convective cells forming in the Houston domain. MAAS used either a list of predetermined rules or real-time user input to select a convective cell to be tracked and sampled by the C-band radars. The CSAPR2 tracking radar was first tasked to collect three sector plan position indicator (PPI) scans toward the selected cell. Edge computer processing of the PPI scans was used to identify additional targets within the selected cell. In less than 2 min, both the CSAPR2 and CHIVO radars were able to collect bundles of three to six range–height indicator (RHI) scans toward different targets of interest within the selected cell. Bundles were successively collected along the path of cell advection for as long as the cell met a predetermined set of criteria. Between 1 June and 30 September 2022 over 315 000 vertical cross-section observations were collected by the C-band radars through ∼1300 unique isolated convective cells, most of which were observed for over 15 min of their life cycle. To the best of our knowledge, this dataset, collected primarily through automatic means, constitutes the largest dataset of its kind.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2021

Efficient Sampling Allocation Procedures for Optimal Quantile Selection

We propose a dynamic sampling allocation and selection paradigm for finding the alternative with the optimal quantile in a Bayesian framework. Myopic allocation policies (MAPs), analogous to existing methods in classic ranking and selection for selecting the alternative with the optimal mean, and computationally efficient selection policies are derived for selecting the alternative with the optimal quantile. Under certain conditions, we prove that the proposed MAPs and selection procedures are consistent, which means that the best quantile would be eventually correctly selected as the sample size goes to infinity. Numerical experiments demonstrate that the proposed schemes can significantly improve the performance.

Computer Science↗

DOE OSTI2026

Water4Energy Step-1 Band-M Ready-to-Train Samples for TVA Weeks-to-Years Prediction, Version 0

AI-ready Band-M (monthly) labelled training pack for the Water4Energy Genesis Task-1 project on weeks-to-years prediction of Tennessee Valley temperature and precipitation. The deposit includes leakage-aware issue-time samples (samples_M_v0.nc; N=486), train-only scalers, issue-time split table, supporting monthly panels, and Python generation scripts to recreate the pack from the companion Tier-1 raw observation collection (https://doi.org/10.13139/ORNLNCCS/3398576). Each sample pairs a 12-month lookback of teleconnection indices and SST box anomalies with TVA-mean ERA5 anomaly targets (t2m, tp, msl) at leads 1–3 months.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2022

Inferring small-sample-reactivity worth from oscillation experiment with auto-corrected local flux

To estimate reactivity changes as a function of time from the signal of detectors count rates, one usually uses point kinetic equations. However, when detectors are located inside the core, it is often necessary to compute a correction factor to account for the flux-shape deformation (i.e. the change of detector efficiency that may have been induced by neutron absorbers movements). Various simulation methods have been developed depending on transients observed. They range from the resolution of static equations up to full-time-dependent neutron transport in 3D. However, those methods require high computational-cost calculations of local reaction rates that rely on nuclear data evaluations. Therefore, a 'measured' reactivity is never determined through pure measurements. Taking advantage of the known oscillation movement of a small sample at the center of the MINERVE reactor, the innovative algorithm presented in this paper derives the correction factor needed to get no time drift on reactivity steps. Derived correction factors are in good agreement with those obtained independently with TRIPOLI-4 through sensitivity analysis. Low sensitivity to the sample composition and moderate sensitivity to the detectors position reinforce the reliability of those results. This application paves the way to almost pure nuclear data reactivity measurements. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗