Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “level set methods”

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 325 records · Page 18
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Beyond pinball loss: Quantile methods for calibrated uncertainty quantification

Amongthemanywaysofquantifying uncertainty in a regression setting, specifying the full quantile function is attractive, as quantiles are amenable to interpretation and evaluation. A model that predicts the true conditional quantiles for each input, at all quantile levels, presents a correct and efficient representation of the underlying uncertainty. To achieve this, many current quantile-based methods focus on optimizing the pinball loss. However, this loss restricts the scope of applicable regression models, limits the ability to target many desirable properties (e.g. calibration, sharpness, centered intervals), and may produce poor conditional quantiles. In this work, we develop new quantile methods that address these shortcomings. In particular, we propose methods that can apply to any class of regression model, select an explicit balance between calibration and sharpness, optimize for calibration of centered intervals, and produce more accurate conditional quantiles. We provide a thorough experimental evaluation of our methods, which includes a high dimensional uncertainty quantification task in nuclear fusion.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Iterative Linearization for Phasor-Defined Optimal Power Dispatch

Optimal power flow (OPF) problems, which dispatch power targets to controllable generating units across a network, must generally account for non-convex constraints on power flow. Furthermore, adapting those problems so as to make them solvable with convex optimization techniques is an area of much academic and operational interest. In this paper, we present a method for solving OPF as a quadratic program by iteratively refining and re-initializing a linearized model of power flow based on the outputs of an associated nonlinear solver. The linear model on which we demonstrate this method is an adapted version of an approximation designed for use with unbalanced distribution networks. As an important benefit, the model allows for the explicit inclusion of nodal voltage phasor values in both the OPF problem's objective and its constraints, which opens the door to the idea of phasor-based control (PBC) design. We show in simulations on the IEEE 13-node test feeder that our method quickly converges to a set of phasor targets that are sufficiently precise for use in operations at the distribution level.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2025

Artificial Intelligence Benchmarking

AI benchmarking is a method for evaluating the effectiveness of an AI model using a set of standardized metrics, for example, high school-level math exams. These benchmarks and their results will enable ranking various AI models based on their effectiveness in performing a specific task.

Krishnan, Anjay [Fermilab]↗
DOE OSTI2017

Single–Column Model Simulations of Subtropical Marine Boundary–Layer Cloud Transitions Under Weakening Inversions

Results are presented of the GASS/EUCLIPSE single-column model inter-comparison study on the subtropical marine low-level cloud transition. A central goal is to establish the performance of state-of-the-art boundary-layer schemes for weather and climate models for this cloud regime, using large-eddy simulations of the same scenes as a reference. A novelty is that the comparison covers four different cases instead of one, in order to broaden the covered parameter space. Three cases are situated in the North-Eastern Pacific, while one reflects conditions in the North-Eastern Atlantic. A set of variables is considered that reflects key aspects of the transition process, making use of simple metrics to establish the model performance. Using this method some longstanding problems in low level cloud representation are identified. Considerable spread exists among models concerning the cloud amount, its vertical structure and the associated impact on radiative transfer. Here, the sign and amplitude of these biases differ somewhat per case, depending on how far the transition has progressed. After cloud breakup the ensemble median exhibits the well-known “too few too bright” problem. The boundary layer deepening rate and its state of decoupling are both underestimated, while the representation of the thin capping cloud layer appears complicated by a lack of vertical resolution. Encouragingly, some models are successful in representing the full set of variables, in particular the vertical structure and diurnal cycle of the cloud layer in transition. An intriguing result is that the median of the model ensemble performs best, inspiring a new approach in subgrid parameterization.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Non-Gaussian estimates of tensions in cosmological parameters

We discuss how to efficiently and reliably estimate the level of agreement and disagreement on parameter determinations from different experiments, fully taking into account non-Gaussianities in the parameter posteriors. We develop two families of scalable algorithms that allow us to perform this type of calculations in increasing number of dimensions and for different levels of tensions. One family of algorithms rely on kernel density estimates of posterior distributions while the other relies on machine learning modeling of the posterior distribution with normalizing flows. We showcase their effectiveness and accuracy with a set of benchmark examples and find both methods agree with each other and the true tension within 0.5σ in difficult cases and generally to 0.2σ or better. This allows us to study the level of internal agreement between different measurements of the clustering of cosmological structures from the Dark Energy Survey and their agreement with measurements of the Cosmic Microwave Background from the Planck satellite.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Bias-Adjustment Methods for Future Subdaily Precipitation Extremes Consistent Across Durations

Model output from climate projections often requires bias-adjustment to compensate for systematic model errors. A bias-adjustment method for extreme precipitation intensity is proposed that preserves the scaling equation for different accumulation levels from hourly to daily, using intensity-duration-frequency (IDF) modeling. A validation is performed within a pseudo-reality setting, based on hourly precipitation from 28 regional climate model projections of the EURO-CORDEX ensemble over Belgium. The scaling-based adjustment methods improve upon previous methods, an optimal method is identified, and, analytical quantile mapping methods must be avoided due to three identified problems. The ensemble mean of the adjusted extreme precipitation intensity obeys the above-mentioned scale-invariance property, which is consistent with observed extreme intensities. We thus show that IDF modeling provides added value in the context of bias-adjustment, and, that the particular IDF model proposed balances well between accuracy and the preservation of desired properties such as scale invariance and consistency among rainfall durations.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Few-shot leaf segmentation

In this work, we use few-shot learning to segment the body and vein architecture of Populus trichocarpa leaves from high-resolution scans obtained in the UC Davis common garden. Leaf and vein segmentation are formulated as separate tasks, in which convolutional neural networks (CNNs) are used to iteratively expand partial segmentations until reaching stopping criteria. Our leaf and vein segmentation approaches use just 50 and 8 manually traced images for training, respectively, and are applied to a set of 2,634 top and bottom leaf scans. We show that both methods achieve high segmentation accuracy, in some cases exceeding even human-level segmentation. The leaf and vein segmentations are subsequently used to extract 68 morphological traits using traditional open-source image processing tools, which are validated using real-world physical measurements. For a biological perspective, we perform a genome-wide association study using the "vein density" trait to discover novel genetic architectures associated with multiple physiological processes relating to leaf development and function. In addition to sharing all of the few-shot learning code, we are releasing all images, manual segmentations, model predictions, 68 extracted leaf phenotypes, and a new set of SNPs called against the v4 P. trichocarpa genome for 1,419 genotypes.

Lagergren, John [Oak Ridge National Lab. (ORNL), O↗
DOE OSTI2023

Hybrid Cyber-attack Detection in Photovoltaic Farms

Here, to address the cyber-physical security in PV farms, a hybrid cyber-attack detection is proposed in this manuscript. To secure PV farms, the proposed method integrates model-based and data-driven methods by fusing the detection score at the device and system levels. First, a model-based cyber-attack detection method is developed for each PV inverter. A residual between the estimation of the Kalman filter and measurement is calculated. By leveraging the calculated residual from all inverters, a squared Mahalanobis distance is developed for device detection score generation. At the system level, a convolutional neural network (CNN) is proposed to detect cyber-attack using the waveform data at the point of common coupling (PCC) in PV farms. To improve the CNN detection accuracy, a set of well-designed features are extracted from the raw waveform data. Finally, a weighted detection score fusion method is proposed to combine device and system detection scores by using their complementary strength. The feasibility and robustness of the proposed method are validated by testing cases and a comparative experiment.

14 SOLAR ENERGY↗
DOE OSTI2015

SU-F-BRB-05: Collision Avoidance Mapping Using Consumer 3D Camera

Purpose: To develop a fast and economical method of scanning a patient’s full body contour for use in collision avoidance mapping without the use of ionizing radiation. Methods: Two consumer level 3D cameras used in electronic gaming were placed in a CT simulator room to scan a phantom patient set up in a high collision probability position. A registration pattern and computer vision algorithms were used to transform the scan into the appropriate coordinate systems. The cameras were then used to scan the surface of a gantry in the treatment vault. Each scan was converted into a polygon mesh for collision testing in a general purpose polygon interference algorithm. All clinically relevant transforms were applied to the gantry and patient support to create a map of all possible collisions. The map was then tested for accuracy by physically testing the collisions with the phantom in the vault. Results: The scanning fidelity of both the gantry and patient was sufficient to produce a collision prediction accuracy of 97.1% with 64620 geometry states tested in 11.5 s. The total scanning time including computation, transformation, and generation was 22.3 seconds. Conclusion: Our results demonstrate an economical system to generate collision avoidance maps. Future work includes testing the speed of the framework in real-time collision avoidance scenarios. Research partially supported by a grant from Varian Medical Systems.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

Time delay lens modelling challenge

ABSTRACT In recent years, breakthroughs in methods and data have enabled gravitational time delays to emerge as a very powerful tool to measure the Hubble constant H0. However, published state-of-the-art analyses require of order 1 yr of expert investigator time and up to a million hours of computing time per system. Furthermore, as precision improves, it is crucial to identify and mitigate systematic uncertainties. With this time delay lens modelling challenge, we aim to assess the level of precision and accuracy of the modelling techniques that are currently fast enough to handle of order 50 lenses, via the blind analysis of simulated data sets. The results in Rungs 1 and 2 show that methods that use only the point source positions tend to have lower precision ($10\!-\!20{{\ \rm per\ cent}}$) while remaining accurate. In Rung 2, the methods that exploit the full information of the imaging and kinematic data sets can recover H0 within the target accuracy (|A| < 2 per cent) and precision (<6 per cent per system), even in the presence of a poorly known point spread function and complex source morphology. A post-unblinding analysis of Rung 3 showed the numerical precision of the ray-traced cosmological simulations to be insufficient to test lens modelling methodology at the percent level, making the results difficult to interpret. A new challenge with improved simulations is needed to make further progress in the investigation of systematic uncertainties. For completeness, we present the Rung 3 results in an appendix and use them to discuss various approaches to mitigating against similar subtle data generation effects in future blind challenges.

Ding, X.↗
DOE OSTI2023

Supporting information for Few-Shot Learning Enables Population-Scale Analysis of Leaf Traits in Populus trichocarpa

In this work, we use few-shot learning to segment the body and vein architecture of P. trichocarpa leaves from high-resolution scans obtained in the UC Davis common garden. Leaf and vein segmentation are formulated as separate tasks, in which convolutional neural networks (CNNs) are used to iteratively expand partial segmentations until reaching stopping criteria. Our leaf and vein segmentation approaches use just 50 and 8 manually traced images for training, respectively, and are applied to a set of 2,634 top and bottom leaf scans. We show that both methods achieve high segmentation accuracy, in some cases exceeding even human-level segmentation. The leaf and vein segmentations are subsequently used to extract 68 morphological traits using traditional open-source image processing tools, which are validated using real-world physical measurements. For a biological perspective, we perform a genome-wide association study using the vein density trait to discover novel genetic architectures associated with multiple physiological processes relating to leaf development and function. In addition to sharing all of the few-shot learning code (see https://github.com/jlager/few-shot-leaf-segmentation), we are releasing all images, manual segmentations, model predictions, 68 extracted leaf phenotypes, and a new set of SNPs called against the v4 P. trichocarpa genome for 1,419 genotypes. The data folder includes all images, ground truth segmentations, predicted segmentations, and extracted leaf traits. All images encode the sample ID in the file name by indicating the treatment, block, row, position, and leaf side, respectively. For example, the file, C_1_1_2_bot.jpeg, indicates the control treatment, block 1, row 1, position 2, and the bottom side of the leaf. Tabulated results include position IDs as well as the corresponding genotype IDs. The images folder includes the 2,906 high-resolution leaf scans taken in the field. The leaf_masks folder includes 50 ground truth segmentations used for training the leaf tracing algorithm. The leaf_preds folder includes the 2,906 predicted segmentations from the leaf tracing algorithm. The vein_masks folder includes 8 ground truth segmentations used for training the vein growing algorithm. The vein_preds folder includes the 1,453 predicted segmentations from the vein growing algorithm. The vein_probs folder includes the 1,453 predicted probability maps from the vein growing algorithm before thresholding. The genomes folder includes the set of SNPs called against the v4 P. trichocarpa genome for 1,419 genotypes with a README file detailing the steps taken. The results folder includes: (i) raw values of the 68 predicted leaf traits in digital_traits.tsv, (ii) manually measured values of petiole length and width in manual_traits.tsv, (iii) thin plate spline (TPS) adjusted values of the vein density trait in vein_density_tps_adj.tsv, (iv) best linear unbiased prediction (BLUP) adjusted values of the vein density trait in vein_density_blups.tsv, and (v) GWAS results for the vein density trait, including chromosome positions and corresponding P values, in gwas_results.csv.

09 BIOMASS FUELS↗
DOE OSTI2026

DP-TwoLevel: two-stage gradient subspace learning for differentially private federated learning

Federated learning (FL) enables collaborative model training across distributed data sources without sharing raw data, but faces fundamental challenges in communication efficiency and privacy. Differentially private (DP) training mitigates information leakage but introduces noise that degrades model performance, especially in high-dimensional settings. We propose DP-TwoLevel, a hierarchical gradient projection method that improves utility under fixed DP constraints by exploiting low-dimensional structure in model updates. Our approach learns a two-level PCA-based representation of gradients and applies DP noise in a reduced-dimensional subspace, thereby lowering the effective noise magnitude while preserving dominant signal components. We evaluate the method across three datasets (MNIST, Fashion-MNIST, CIFAR-10) and three privacy regimes (ϵ∈0.5, 1.0, 2.0). Across nine experimental settings, DP-TwoLevel consistently outperforms DP-FedAvg, achieving an average accuracy improvement of 9.44%, with larger gains observed in lower ϵ(higher-noise) regimes (up to +22.31%). We further analyze scalability across models ranging from 100K to 1.49M parameters and identify a variance-based success criterion: performance remains strong when the projection preserves more than 75% of gradient variance, degrades in a marginal regime (65–75%), and fails below this threshold. Our results demonstrate that structure-aware dimensionality reduction can significantly improve the privacy–utility tradeoff in FL without modifying formal privacy guarantees. We also provide empirical evidence of scaling limitations for global projections and motivate per-layer extensions for larger models.

Kotevska, Olivera [ORNL] (ORCID:0000000316772243)↗
DOE OSTI2016

Eulerian frequency analysis of structural vibrations from high-speed video

An approach for the analysis of the frequency content of structural vibrations from high-speed video recordings is proposed. The techniques and tools proposed rely on an Eulerian approach, that is, using the time history of pixels independently to analyse structural motion, as opposed to Lagrangian approaches, where the motion of the structure is tracked in time. The starting point is an existing Eulerian motion magnification method, which consists in decomposing the video frames into a set of spatial scales through a so-called Laplacian pyramid [1]. Each scale — or level — can be amplified independently to reconstruct a magnified motion of the observed structure. The approach proposed here provides two analysis tools or pre-amplification steps. The first tool provides a representation of the global frequency content of a video per pyramid level. This may be further enhanced by applying an angular filter in the spatial frequency domain to each frame of the video before the Laplacian pyramid decomposition, which allows for the identification of the frequency content of the structural vibrations in a particular direction of space. This proposed tool complements the existing Eulerian magnification method by amplifying selectively the levels containing relevant motion information with respect to their frequency content. This magnifies the displacement while limiting the noise contribution. The second tool is a holographic representation of the frequency content of a vibrating structure, yielding a map of the predominant frequency components across the structure. In contrast to the global frequency content representation of the video, this tool provides a local analysis of the periodic gray scale intensity changes of the frame in order to identify the vibrating parts of the structure and their main frequencies. Validation cases are provided and the advantages and limits of the approaches are discussed. The first validation case consists of the frequency content retrieval of the tip of a shaker, excited at selected fixed frequencies. The goal of this setup is to retrieve the frequencies at which the tip is excited. The second validation case consists of two thin metal beams connected to a randomly excited bar. It is shown that the holographic representation visually highlights the predominant frequency content of each pixel and locates the global frequencies of the motion, thus retrieving the natural frequencies for each beam.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2017

Understanding the many-body expansion for large systems. III. Critical role of four-body terms, counterpoise corrections, and cutoffs

Papers I and II in this series [R. M. Richard et al., J. Chem. Phys. 141, 014108 (2014); K. U. Lao et al., ibid. 144, 164105 (2016)] have attempted to shed light on precision and accuracy issues affecting the many-body expansion (MBE), which only manifest in larger systems and thus have received scant attention in the literature. Many-body counterpoise (CP) corrections are shown to accelerate convergence of the MBE, which otherwise suffers from a mismatch between how basis-set superposition error affects subsystem versus supersystem calculations. In water clusters ranging in size up to (H2O)37, four-body terms prove necessary to achieve accurate results for both total interaction energies and relative isomer energies, but the sheer number of tetramers makes the use of cutoff schemes essential. To predict relative energies of (H2O)20 isomers, two approximations based on a lower level of theory are introduced and an ONIOM-type procedure is found to be very well converged with respect to the appropriate MBE benchmark, namely, a CP-corrected supersystem calculation at the same level of theory. Results using an energy-based cutoff scheme suggest that if reasonable approximations to the subsystem energies are available (based on classical multipoles, say), then the number of requisite subsystem calculations can be reduced even more dramatically than when distance-based thresholds are employed. The end result is several accurate four-body methods that do not require charge embedding, and which are stable in large basis sets such as aug-cc-pVTZ that have sometimes proven problematic for fragment-based quantum chemistry methods. Even with aggressive thresholding, however, the four-body approach at the self-consistent field level still requires roughly ten times more processors to outmatch the performance of the corresponding supersystem calculation, in test cases involving 1500–1800 basis functions.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2015

A Parallel Computing Framework for Finding the Supported Solutions to a Biobjective Network Optimization Problem

Abstract Solving multi‐objective combinatorial optimization problems can quickly become computationally challenging when applied to large networks generated from big data. Present‐day first‐rate Mixed Integer Programming (MIP) solvers have parallelism built‐in to take advantage of multicore architectures; but specialized network optimization algorithms that can often solve graph problems more efficiently than a general MIP solver are typically programmed serially. Thus, these network algorithm implementations do not take full advantage of modern multicore computer capabilities. Many parallel computing languages include a fork/join framework into their concurrency packages that allow for simple conversion of recursive serial algorithms to operate in parallel without having to use complicated low‐level message‐passing interfaces. Fork/join is particularly well suited to divide and conquer algorithms such as Non‐Inferior Set Estimation (NISE) method for computing the supported Pareto front of a multi‐objective optimization problem. This work develops a simple general parallel NISE (pNISE) implementation using the Java 7 concurrency tools, and then offers results from a specific implementation of solving a bi‐objective shortest path problem. The results indicate that this method is capable of sizeable speed‐ups on both small‐scale personal computers as well as large‐scale shared memory supercomputers. Finally, other network problems suitable to this method are discussed, including multi‐objective variants of the minimum spanning tree problem, transportation problem, assignment problem, maximum flow problem, and the minimum‐cost flow problem. Copyright © 2015 John Wiley & Sons, Ltd.

Medrano, Fernando Antonio↗
DOE OSTI2022

E(3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials

Abstract This work presents Neural Equivariant Interatomic Potentials (NequIP), an E(3)-equivariant neural network approach for learning interatomic potentials from ab-initio calculations for molecular dynamics simulations. While most contemporary symmetry-aware models use invariant convolutions and only act on scalars, NequIP employs E(3)-equivariant convolutions for interactions of geometric tensors, resulting in a more information-rich and faithful representation of atomic environments. The method achieves state-of-the-art accuracy on a challenging and diverse set of molecules and materials while exhibiting remarkable data efficiency. NequIP outperforms existing models with up to three orders of magnitude fewer training data, challenging the widely held belief that deep neural networks require massive training sets. The high data efficiency of the method allows for the construction of accurate potentials using high-order quantum chemical level of theory as reference and enables high-fidelity molecular dynamics simulations over long time scales.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2013

Flexible nuclear screening approximation to the two-electron spin–orbit coupling based on ab initio parameterization

The derivation, implementation, and validation of a new approximation to the two-electron spin–orbit coupling (SOC) terms is reported. The approximation, referred to as flexible nuclear screening spin–orbit, is based on the effective one-electron spin–orbit operator and accounts for two-electron SOC effects by screening nuclear charges. A highly flexible scheme for the nuclear screening is developed, mainly using parameterization based on ab initio atomic SOC calculations. Tabulated screening parameters are provided for contracted and primitive Gaussian-type basis functions of the ANO-RCC basis set for elements from H to Cm. The strategy for their adaptation to any other Gaussian basis set is presented and validated. A model to correct for the effect of splitting of transition metal d orbitals on their SOC matrix elements is introduced. The method is applied to a representative set of molecules, and compared to exact treatment and other approximative approaches at the same level of relativistic theory. The calculated SOC matrix elements are in very good agreement with their “exact” values; deviation below 1% is observed on average. The presented approximation is considered to be generally applicable, simple to implement, highly efficient, and accurate.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2020

Microscopic study of the coupled-wire construction and plausible realization in spin-dependent optical lattices

Coupled-wire constructions offer particularly simple and powerful models to capture the essence of strongly correlated topological phases of matter. They often rely on effective theories valid in the low-energy and strong-coupling limits, which impose severe constraints on the physical systems where they could be realized. In this work, we investigate the microscopic relevance of a class of coupled-wire models and their possible experimental realization in cold-atom experiments. We connect with earlier results and prove the emergence of fractional quantum Hall states in the limit of strong interwire tunneling. Contrary to previous studies relying on renormalization group arguments, our microscopic approach exposes the connection between coupled-wire constructions and model wave functions in continuum Landau levels. Then, we use exact-diagonalization methods to investigate the appearance of these fractional quantum Hall states in more realistic settings. We examine the parameter regimes where these strongly correlated phases arise, and provide a way to detect their appearance in cold-atom experiments through standard time-of-flight measurements. Motivated by this experimental probe, we finally propose a realization of our model with cold atoms in spin-dependent optical lattices. Our estimates show that the previous fractional quantum Hall phases lie within experimentally accessible parameter regimes, giving a viable route toward their experimental study.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗