Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Distributed training”

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 379 records · Page 21
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

DeepONet-grid-UQ: A trustworthy deep operator framework for predicting the power grid’s post-fault trajectories

This paper proposes a novel data-driven method for the reliable prediction of the power grid’s post-fault trajectories, i.e., the power grid’s dynamic response after a disturbance or fault. Here, the proposed method is based on the recently proposed concept of Deep Operator Networks (DeepONets). Unlike traditional neural networks that learn to approximate functions, DeepONets are designed to approximate nonlinear operators, i.e., mappings between infinite-dimensional spaces. Under this operator framework, we design a novel and efficient DeepONet that (i) takes as inputs the trajectories collected before and during the fault and (ii) outputs the predicted post-fault trajectories. In addition, we endow our method with the much-needed ability to balance efficiency with reliable/trustworthy predictions via uncertainty quantification. To this end, we propose and compare two novel methods that enable quantifying the predictive uncertainty. First, we propose a Bayesian DeepONet (B-DeepONet) that uses stochastic gradient Hamiltonian Monte-Carlo to sample from the posterior distribution of the DeepONet trainable parameters. Then, we design a Probabilistic DeepONet (Prob-DeepONet) that uses a probabilistic training strategy to enable quantifying uncertainty at virtually no extra computational cost. Finally, we validate the proposed methods’ predictive power and uncertainty quantification capability using the New York-New England power grid model.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2023

Stakeholder Training for IEEE 1547-2018 (Abbreviated Final Technical Report)

This project identifies and makes publicly available educational materials to assist stakeholders in adopting the revised IEEE Std 1547-2018, Standard for Interconnection and Interoperability of Distributed Energy Resources with Associated Electric Power Systems Interfaces. The new standard is significantly different than the previous 2003 version, contains new concepts and new technical requirements, and requires educational material on the application of these changes. NREL had a principal role in accelerating the revision process though its chairmanship of the IEEE Std 1547 working group to revise the standard and through technical support provided to the working group. Under this project, NREL and partners compiled and developed publicly available educational materials for IEEE Std 1547-2018, including webinars, white papers, and other resources on related topics.

14 SOLAR ENERGY↗
DOE OSTI2022

Robust and scalable uncertainty estimation with conformal prediction for machine-learned interatomic potentials

Abstract Uncertainty quantification (UQ) is important to machine learning (ML) force fields to assess the level of confidence during prediction, as ML models are not inherently physical and can therefore yield catastrophically incorrect predictions. Established a-posteriori UQ methods, including ensemble methods, the dropout method, the delta method, and various heuristic distance metrics, have limitations such as being computationally challenging for large models due to model re-training. In addition, the uncertainty estimates are often not rigorously calibrated. In this work, we propose combining the distribution-free UQ method, known as conformal prediction (CP), with the distances in the neural network’s latent space to estimate the uncertainty of energies predicted by neural network force fields. We evaluate this method (CP+latent) along with other UQ methods on two essential aspects, calibration, and sharpness, and find this method to be both calibrated and sharp under the assumption of independent and identically-distributed (i.i.d.) data. We show that the method is relatively insensitive to hyperparameters selected, and test the limitations of the method when the i.i.d. assumption is violated. Finally, we demonstrate that this method can be readily applied to trained neural network force fields with traditional and graph neural network architectures to obtain estimates of uncertainty with low computational costs on a training dataset of 1 million images to showcase its scalability and portability. Incorporating the CP method with latent distances offers a calibrated, sharp and efficient strategy to estimate the uncertainty of neural network force fields. In addition, the CP approach can also function as a promising strategy for calibrating uncertainty estimated by other approaches.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2021

Anomaly Detection, Localization and Classification using Drifting Synchrophasor Data Streams

With ongoing automation and digitization of the electric power system, several Phasor Measurement Units(PMUs) have been deployed for monitoring and control. PMU data can have multiple anomalies, and many of the researchers in the past have concentrated on training machine/deep learning algorithms offline for anomaly detection over PMU data (i.e., not in real time). These machine/deep learning algorithms, when trained offline on a sample rather than a population of the dataset, fail to consider the dynamic behavior of the power grid in real-time, resulting in low accuracy. Considering the dynamic behavior of the power grid (e.g., change in load, generation, distributed energy resources (DERs) switching, network, controls), the definition of data anomalies varies in time and requires online training. A fundamental challenge is to enable online (i.e., real-time) training of machine/deep learning algorithms for anomaly detection over streaming PMU data. While machine/deep learning is often desirable to manage data streams, training a deep learning algorithm over streaming PMU data is nontrivial due to changes in data statistics caused by dynamic streaming data. This paper proposes PMUNET: a novel device-level deep learning-based data-driven approach for anomaly detection, localization, and classification over streaming PMU data, using online learning and multivariate data-drift detection algorithm .Two variants of PMUNET, Dynamic data Change Driven Learning (DCDL) and Continuity Driven Learning (CDL), are proposed and compared. DCDL aims to train the deep learning algorithm whenever the definition of anomaly changes due to the power grid dynamics. On the other hand, CDL continuously trains the deep learning algorithm over the PMU data-stream. The experimental results verify that DCDL outperforms CDL and other efficient anomaly detection methods over multiple events such as faults and load/ generator/capacitor/DERs variations/switching for IEEE 14 and 39 Bus test system as well as real PMU industrial data. The result verifies that DCDL variant of PMUNET improves over existing approach with a gain of 2% - 10% in terms of accuracy, false-positive rate, and false-negative rate.

adversarial deep learning↗
DOE OSTI2024

Training custom light curve models of SN Ia subpopulations selected according to host galaxy properties

ABSTRACT Type Ia supernova (SN Ia) cosmology analyses include a luminosity step function in their distance standardization process to account for an observed yet unexplained difference in the post-standardization luminosities of SNe Ia originating from different host galaxy populations [e.g. high-mass ($M \gtrsim 10^{10} \, {\rm M}_{\odot }$) versus low-mass galaxies]. We present a novel method for including host-mass correlations in the SALT3 (Spectral Adaptive Light curve Template 3) light curve model used for standardizing SN Ia distances. We split the SALT3 training sample according to host-mass, training independent models for the low- and high-host-mass samples. Our models indicate that there are different average Si ii spectral feature strengths between the two populations, and that the average spectral energy distribution of SNe from low-mass galaxies is bluer than the high-mass counterpart. We then use our trained models to perform an SN cosmology analysis on the 3-yr spectroscopically confirmed Dark Energy Survey SN sample, treating SNe from low- and high-mass host galaxies as separate populations throughout. We find that our mass-split models reduce the Hubble residual scatter in the sample, albeit at a low statistical significance. We do find a reduction in the mass-correlated luminosity step but conclude that this arises from the model-dependent re-definition of the fiducial SN absolute magnitude rather than the models themselves. Our results stress the importance of adopting a standard definition of the SN parameters (x0, x1, c) in order to extract the most value out of the light curve modelling tools that are currently available and to correctly interpret results that are fit with different models.

Taylor, G. (ORCID:0000000157563259)↗
DOE OSTI2025

Evaluation of data driven low-rank matrix factorization for accelerated solutions of the Vlasov equation

Low-rank methods have shown success in accelerating simulations of a collisionless plasma described by the Vlasov equation, but still rely on computationally costly linear algebra every time step. We propose a data-driven factorization method using artificial neural networks, specifically with convolutional layer architecture, that trains on existing simulation data. At inference time, the model outputs a low-rank decomposition of the distribution field of the charged particles, and we demonstrate that this step is faster than the standard linear algebra technique. Numerical experiments show that the method achieves comparable reconstruction accuracy for interpolation tasks, generalizing to unseen test data in a manner beyond just memorizing training data; patterns in factorization also inherently followed the same numerical trend as those within algebraic methods (e.g., truncated singular-value decomposition). However, when training on the first 70% of a time-series data and testing on the remaining 30%, the method fails to meaningfully extrapolate. Despite this limiting result, the technique may have benefits for simulations in a statistical steady-state or otherwise showing temporal stability. These results suggest that while the model offers a computationally efficient alternative for datasets with temporal stability, its current formulation is best suited for interpolation rather than for predicting future states in time-evolving systems. This study thus lays the groundwork for further refinement of neural network-based approaches to low-rank matrix factorization in high-dimensional plasma simulations.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

BioADAPT-MRC: adversarial learning-based domain adaptation improves biomedical machine reading comprehension task

ABSTRACT Motivation Biomedical machine reading comprehension (biomedical-MRC) aims to comprehend complex biomedical narratives and assist healthcare professionals in retrieving information from them. The high performance of modern neural network-based MRC systems depends on high-quality, large-scale, human-annotated training datasets. In the biomedical domain, a crucial challenge in creating such datasets is the requirement for domain knowledge, inducing the scarcity of labeled data and the need for transfer learning from the labeled general-purpose (source) domain to the biomedical (target) domain. However, there is a discrepancy in marginal distributions between the general-purpose and biomedical domains due to the variances in topics. Therefore, direct-transferring of learned representations from a model trained on a general-purpose domain to the biomedical domain can hurt the model’s performance. Results We present an adversarial learning-based domain adaptation framework for the biomedical machine reading comprehension task (BioADAPT-MRC), a neural network-based method to address the discrepancies in the marginal distributions between the general and biomedical domain datasets. BioADAPT-MRC relaxes the need for generating pseudo labels for training a well-performing biomedical-MRC model. We extensively evaluate the performance of BioADAPT-MRC by comparing it with the best existing methods on three widely used benchmark biomedical-MRC datasets—BioASQ-7b, BioASQ-8b and BioASQ-9b. Our results suggest that without using any synthetic or human-annotated data from the biomedical domain, BioADAPT-MRC can achieve state-of-the-art performance on these datasets. Availability and implementation BioADAPT-MRC is freely available as an open-source project at https://github.com/mmahbub/BioADAPT-MRC. Supplementary information Supplementary data are available at Bioinformatics online.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2022

Network Reconfiguration for Enhanced Operational Resilience Using Reinforcement Learning

This paper proposes a reinforcement learning-based approach for distribution network reconfiguration(DNR) to enhance the resilience of the electric power supply. Resilience enhancements usually require solving large-scale stochastic optimization problems that are computationally expensive and sometimes infeasible. The exceptional performance of reinforcement learning techniques has encouraged their adoption in various power system control studies, specifically resilience-based real-time applications. In this paper, a single agent framework is developed using an Actor-Critic algorithm (ACA) to determine statuses of tie-switches in a distribution feeder impacted by an extreme weather event. The proposed approach provides a fast-acting control algorithm that reconfigures the feeder topology to reduce or even avoid load shedding. The problem is formulated as a discrete Markov decision process in such a way that a system state captures the system topology and its operational characteristics. An action is made to open or close a specific set of tie-switches after which a reward is calculated to evaluate the practicality and advantage of that action. The iterative Markov process is used to train the proposed ACA under diverse failure scenarios and is demonstrated on the 33-node distribution feeder system. Results show the capability of the proposed ACA to determine proper switching action of tie-switches with accuracy exceeding 93%.

actor critic↗
DOE OSTI2023

Using deep transfer learning and satellite imagery to estimate urban air quality in data-poor regions

Urban air pollution is a critical public health challenge in low-and-middle-income countries (LMICs). At the same time, LMICs tend to be data-poor, lacking adequate infrastructure to monitor air quality (AQ). As LMICs undergo rapid urbanization, the socio-economic burden of poor AQ will be immense. Here we present a globally scalable two-step deep learning (DL) based approach for AQ estimation in LMIC cities that mitigates the need for extensive AQ infrastructure on the ground. We train a DL model that can map satellite imagery to AQ in high-income countries (HICs) with sufficient ground data, and then adapt the model to learn meaningful AQ estimates in LMIC cities using transfer learning. The trained model can explain up to 54% of the variation in the AQ distribution of the target LMIC city without the need for target labels. The approach is demonstrated for Accra in Ghana, Africa, with AQ patterns learned and adapted from two HIC cities, specifically Los Angeles and New York.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

Superstructure Optimization of Waste Plastic Pyrolysis, Integrating Thermal, Catalytic, and Plasma Technologies with Machine Learning

Global plastic waste generation exceeds 430 million tonnes per year, yet fewer than 9% are recycled in the United States. Pyrolysis offers a chemical recycling route at scale, but existing techno-economic and life cycle assessments fix product yields to single pure polymers, producing economic and environmental outputs that break down when the feed composition changes. Here, we present a superstructure optimization framework that addresses this by embedding a composition-aware random forest yield predictor, trained on 566 pyrolysis experiments, within a full-scale process simulation. Product distributions update automatically as feed allocation shifts across four reactor chemistries: conventional thermal, catalytic (HZSM-5), thermal oxo-degradation, and nonequilibrium CO2 plasma. The optimal superstructure achieves minimum selling prices of −0.56 to −0.76/kg feed and global warming potentials of −0.276 to −0.322 kg CO2-eq/kg feed across four commodity price scenarios, confirming profitable, carbon-negative operation without tipping fees. Carbon abatement costs of $\$$0.46 to $\$$1.25/kg CO2-eq are competitive with direct air capture. Sensitivity analysis shows that the catalytic-plasma split fraction is the single largest driver of both economic and climate performance, while hydrocracking allocation in the wax upgrading stage is emission-neutral across the full variable range. Mixed plastic waste streams, evaluated as composition-variable feedstocks rather than pure resins, are profitable and carbon-negative across realistic market conditions. These results give a quantitative basis for reactor selection, circular economy investment, and policy design targeting chemical recycling on a large scale.

Life cycle assessment↗
DOE OSTI2025

A Probabilistic Model for Global EMIC Wave Activity Using Van Allen Probes Observations

Electromagnetic ion cyclotron (EMIC) waves play a key role in radiation belt dynamics through resonant interactions. However, their low occurrence probability, high variability, and spatial intermittency pose challenges for accurate modeling. In this study, we present a machine learning (ML)-based global EMIC wave model built on the entire data set from the Van Allen Probes mission. To capture the distinct statistical characteristics of wave occurrence and amplitude, the model is separated into two modules: an occurrence model trained using ML techniques, and a wave amplitude model sampled from observed probability distributions. The input parameters are limited to real-time or predictable variables to ensure practical applicability. Our model shows strong performance across the entire test set and demonstrates improved predictive capability over a baseline random occurrence model, particularly during quiet geomagnetic conditions. Evaluation during both quiet and active periods confirms the model's ability to represent the clustered and intermittent nature of EMIC wave activity. Furthermore, the model provides global estimates of wave power, enabling integration with radiation belt electron data and showing signatures consistent with wave-induced scattering. We found a good correlation between the global wave activity from the model and relativistic electron observation by Van Allen Probes, regardless of the availability of in situ wave observations. The modular structure of the model also allows for straightforward expansion for additional wave properties, such as wave frequency, which can be modeled independently. This flexible, event-sensitive approach offers a promising framework for data-driven radiation belt simulations and space weather applications.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

The miniJPAS survey quasar selection – II. Machine learning classification with photometric measurements and uncertainties

Astrophysical surveys rely heavily on the classification of sources as stars, galaxies, or quasars from multiband photometry. Surveys in narrow-band filters allow for greater discriminatory power, but the variety of different types and redshifts of the objects present a challenge to standard template-based methods. In this work, which is part of a larger effort that aims at building a catalogue of quasars from the miniJPAS survey, we present a machine learning-based method that employs convolutional neural networks (CNNs) to classify point-like sources including the information in the measurement errors. We validate our methods using data from the miniJPAS survey, a proof-of-concept project of the Javalambre Physics of the Accelerating Universe Astrophysical Survey (J-PAS) collaboration covering ∼1 deg 2 of the northern sky using the 56 narrow-band filters of the J-PAS survey. Due to the scarcity of real data, we trained our algorithms using mocks that were purpose-built to reproduce the distributions of different types of objects that we expect to find in the miniJPAS survey, as well as the properties of the real observations in terms of signal and noise. We compare the performance of the CNNs with other well-established machine learning classification methods based on decision trees, finding that the CNNs improve the classification when the measurement errors are provided as inputs. The predicted distribution of objects in miniJPAS is consistent with the putative luminosity functions of stars, quasars, and unresolved galaxies. Our results are a proof of concept for the idea that the J-PAS survey will be able to detect unprecedented numbers of quasars with high confidence.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

Robust Online Sequential RVFLNs for Data Modeling of Dynamic Time-Varying Systems with Application of an Ironmaking Blast Furnace

In a world where the increasing complexity of modern industrial processes brings difficulties for accurate mathematical modeling, taking advantage of data has become an efficient solution to complex dynamic process modeling issue. In this paper, we develop a novel robust online sequential version of random vector functional-link networks (RVFLNs) for data-driven modeling of dynamic time-varying system and applied it in a blast furnace (BF) ironmaking process. First, to overcome the time-varying dynamics of process and to enable the RVFLNs to learn online with avoiding data saturation, an improved online sequential version of RVFLNs (OS-RFVLNs) is first presented by online sequential learning with forgetting factor. This improved OS-RVFLNs algorithm is not only suitable for the real-time and large data transfer situation, but also can adjust the sensitivity of the algorithm to different samples with the help of the introduced forgetting factor. Second, since the output weights of the improved OS-RVFLNs as well as other RVFLNs algorithms are obtained by the least squares approach, a robustness problem may occur when the training dataset is contaminated with various outliers. To solve this problem, a Cauchy distribution weighted M-estimator is introduced to improve the robustness of the improved OS- RVFLNs. For this proposed robust OS-RVFLNs (R-OS- RVFLNs), since the weights of different outlier data are properly determined by the Cauchy distribution function, their corresponding contribution on modeling can be properly distinguished. Thus robust and better modeling results can be achieved. Experiments using actual industrial data of BF ironmaking process and comparative studies have demonstrated that the proposed method produces a better estimation accuracy and stronger robustness than other methods.

Blast furnace (BF), Modelling, Dynamic systems↗
DOE OSTI2024

pnnl/emp-gnn

Efficient Graph Neural Network for Predicting Molecular Properties software can compare the prediction quality with ab initio DFT results reported by the high-performance state-of-theart NWChem quantum chemistry package [1] through Mean Absolute Error (MAE) obtained by the fitting between DFT and model predictions i.e, MPNN [2]. We also demonstrated the performance benefits by grouping large molecules by atom sizes, and executing GNN models on different types of resources. Since the training times depend on the number of atoms, we demonstrate the impact of distributing the workloads on two GPUs with varying capabilities (e.g., NVIDIA A100 vs. GeForce RTX 2080 Ti) to optimize the efficiency. We are at the precipice of broad adoption of GNNs for molecular property prediction tasks; hence, our work is timely by comparing model prediction against classical approaches with the intent of providing initial screening for specific classes of molecules.

Lee, Hyungro↗
DOE OSTI2024

MENT-Flow: maximum-entropy phase space tomography using normalizing flows

Generative models can be trained to reproduce low-dimensional projections of high-dimensional phase space distributions. Normalizing flows are generative models that parameterize invertible transformations, allowing exact probability density evaluation and sampling. Consequently, flows are unbiased entropy estimators and could be used to solve the high-dimensional maximum-entropy tomography (MENT) problem. In this work, we evaluate a flow-based MENT solver (MENT-Flow) against exact maximum-entropy solutions and Minerbo's iterative MENT algorithm in two dimensions.

Hoover, Austin↗
DOE OSTI2025

PUFFIn Software Modeling for Quality Management

PUFFIn (PENELOPE User Friendly Fast Interface) was designed as a fast and simple Monte Carlo simulation tool for the transport of photons and electrons, with a primary purpose as a learning and education tool for a broad range of static configurations in the radiation processing industry. Development of the PUFFIn software is funded by the Office of Radiological Security (ORS) within the United States National Nuclear Security Administration (NNSA). PUFFIn helps fill the education and knowledge gaps in the industry, as identified in reports by Fermilab (2017) and the IAEA (2020). PUFFIn uses the PENELOPE (NEA-2023) physics engine to perform simulations on static configurations. PUFFIn has support for multiple geometry types from simple, single material simulations to full 3D configurations created from CAD input files or images from X-Ray Tomography scans. PUFFin was designed to be easy for the novice user, it will generate the input and geometry files required by PENLOPE and will display the output plots within the PUFFIn interface. PUFFin is distributed for free but requires a free workshop so users can be adequately trained in its use. Workshops have been presented in the past at Texas A&M university, the Aerial-CRT facility in Strasbourg France and Jakarta Indonesia. PUFFIn simulations have been validated by 10 MeV ebeam experiments done at Aerial-CRT in France (Radiation Physics and Chemistry 222 (2024) 111774). Further user experimental comparisons were made at the medical product hands on workshop at Texas A&M in October 2024.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2021

Simulation of Electron-Proton Scattering Events by a Feature-Augmented and Transformed Generative Adversarial Network (FAT-GAN)

We apply generative adversarial network (GAN) technology to build an event generator that simulates particle production in electron-proton scattering that is free of theoretical assumptions about underlying particle dynamics. The difficulty of efficiently training a GAN event simulator lies in learning the complicated patterns of the distributions of the particles physical properties. We develop a GAN that selects a set of transformed features from particle momenta that can be generated easily by the generator, and uses these to produce a set of augmented features that improve the sensitivity of the discriminator. The new Feature-Augmented and Transformed GAN (FAT-GAN) is able to faithfully reproduce the distribution of final state electron momenta in inclusive electron scattering, without the need for input derived from domain-based theoretical assumptions. The developed technology can play a significant role in boosting the science of existing and future accelerator facilities, such as the Electron-Ion Collider.

Alanazi, Yasir↗
DOE OSTI2021

Machine Learning Predictions of Transition Probabilities in Atomic Spectra

Forward modeling of optical spectra with absolute radiometric intensities requires knowledge of the individual transition probabilities for every transition in the spectrum. In many cases, these transition probabilities, or Einstein A-coefficients, quickly become practically impossible to obtain through either theoretical or experimental methods. Complicated electronic orbitals with higher order effects will reduce the accuracy of theoretical models. Experimental measurements can be prohibitively expensive and are rarely comprehensive due to physical constraints and sheer volume of required measurements. Due to these limitations, spectral predictions for many element transitions are not attainable. In this work, we investigate the efficacy of using machine learning models, specifically fully connected neural networks (FCNN), to predict Einstein A-coefficients using data from the NIST Atomic Spectra Database. For simple elements where closed form quantum calculations are possible, the data-driven modeling workflow performs well but can still have lower precision than theoretical calculations. For more complicated nuclei, deep learning emerged more comparable to theoretical predictions, such as Hartree–Fock. Unlike experiment or theory, the deep learning approach scales favorably with the number of transitions in a spectrum, especially if the transition probabilities are distributed across a wide range of values. It is also capable of being trained on both theoretical and experimental values simultaneously. In addition, the model performance improves when training on multiple elements prior to testing. The scalability of the machine learning approach makes it a potentially promising technique for estimating transition probabilities in previously inaccessible regions of the spectral and thermal domains on a significantly reduced timeline.

74 ATOMIC AND MOLECULAR PHYSICS↗