Search NASA⌕ Search

SEARCH · Search NASA

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 1,405 records · Page 78

Do Machine Learning Approaches Offer Skill Improvement for Short-Term Forecasting of Wind Gust Occurrence and Magnitude?

Abstract Wind gusts, and in particular intense gusts, are societally relevant but extremely challenging to forecast. This study systematically assesses the skill enhancement that can be achieved using artificial neural networks (ANNs) for forecasting of wind gust occurrence and magnitude. Geophysical predictors from the ERA5 reanalysis are used in conjunction with an autoregressive term in regression and ANN models with different predictors, and varying model complexity. Models are derived and assessed for the warm (April–September) and cold (October–March) seasons for three high passenger volume airports in the United States. Model uncertainty is assessed by deriving models for 1000 different randomly selected training (70%) and testing (30%) subsets. Gust prediction fidelity in independent test samples is critically dependent on inclusion of an autoregressive term. Gust occurrence probabilities derived using five-layer ANNs exhibit consistently higher fidelity than those from regression models and shallower ANNs. Inclusion of the autoregressive term and increasing the number of hidden layers in ANNs from 1 to 5 also improve the model performance for gust magnitudes (lower RMSE, increased correlation, and model standard deviations that more closely approximate observed values). Deeper ANNs (e.g., 20 hidden layers) exhibit higher skill in forecasting strong (17–25.7 m s −1 ) and damaging (≥25.7 m s −1 ) wind gusts. However, such deep networks exhibit evidence of overfitting and still substantially underestimate (by 50%) the frequency of strong and damaging wind gusts at the three airports considered herein. Significance Statement Improved short-term forecasting of wind gusts will enhance aviation safety and logistics and may offer other societal benefits. Here we present a rigorous investigation of the relative skill of models of wind gust occurrence and magnitude that employ different statistical methods. It is shown that artificial neural networks (ANNs) offer considerable skill enhancement over regression methods, particularly for strong and damaging wind gusts. For wind gust magnitudes in particular, application of deeper learning networks (e.g., five or more hidden layers) offers tangible improvements in forecast accuracy. However, deeper networks are vulnerable to overfitting and exhibit substantial variability with the specific training and testing data subset used. Also, even deep ANNs reproduce only half of strong and damaging wind gusts. These results indicate the need for future work to elucidate the dynamical mechanisms of intense wind gusts and advance solutions to their prediction.

54 ENVIRONMENTAL SCIENCES↗

Addressing GPU memory limitations for Graph Neural Networks in High-Energy Physics applications

Introduction Reconstructing low-level particle tracks in neutrino physics can address some of the most fundamental questions about the universe. However, processing petabytes of raw data using deep learning techniques poses a challenging problem in the field of High Energy Physics (HEP). In the Exa.TrkX Project, an illustrative HEP application, preprocessed simulation data is fed into a state-of-art Graph Neural Network (GNN) model, accelerated by GPUs. However, limited GPU memory often leads to Out-of-Memory (OOM) exceptions during training, due to the large size of models and datasets. This problem is exacerbated when deploying models on High-Performance Computing (HPC) systems designed for large-scale applications. Methods We observe a high workload imbalance issue during GNN model training caused by the irregular sizes of input graph samples in HEP datasets, contributing to OOM exceptions. We aim to scale GNNs on HPC systems, by prioritizing workload balance in graph inputs while maintaining model accuracy. Our paper introduces diverse balancing strategies aimed at decreasing the maximum GPU memory footprint and avoiding the OOM exception, across various datasets. Results Our experiments showcase memory reduction of up to 32.14% compared to the baseline. We also demonstrate the proposed strategies can avoid OOM in application. Additionally, we create a distributed multi-GPU implementation using these samplers to demonstrate the scalability of these techniques on the HEP dataset. Discussion By assessing the performance of these strategies as data loading samplers across multiple datasets, we can gauge their effectiveness in both single-GPU and distributed environments. Our experiments, conducted on datasets of varying sizes and across multiple GPUs, broaden the applicability of our work to various GNN applications that handle input datasets with irregular graph sizes.

Lee, Claire Songhyun↗

Selective Amnesia using Contrastive Subnet Erasure for Class Level Unlearning in Vision Models

We study concept-level forgetting in pretrained vision models: removing an entire semantic category so the system no longer recognizes that object in unseen images and contexts, rather than merely forgetting specific training examples. Prior work either applies blunt global projections or fine-tunes parameters, which can introduce collateral damage to unrelated features, add compute, and become unstable as forgetting strength increases. We introduce Contrastive Subnet Erasure (CSE), a training-free, encoder-centric edit that targets a compact set of channels most responsible for the class and attenuates them in a calibrated manner. The modification is algebraically folded into the subsequent layer, yielding no inference-time overhead and leaving task heads unchanged. To evaluate whether forgetting generalizes beyond the data used to specify the class, we introduce a cross dataset protocol in which the class is defined on a source dataset and performance is measured on a disjoint target dataset drawn from a different distribution with no shared images. This setup tests whether the model still fails to recognize the object when it looks different or appears in new scenes, and it helps avoid overfitting to patterns in the source dataset. Across CIFAR 10, CIFAR 100, and ImageNet under this protocol, CSE achieves stronger forgetting of the target class while better preserving non target utility than existing baselines in both single class and multi class settings. Overall, CSE provides a simple, stable, and deployment-ready mechanism for class-level unlearning in vision.

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

Predicting U 3 O 8 powder processing conditions: An AI/ML approach analyzing deep learning embeddings of SEM micrographs

High-resolution SEM images of uranium-oxide powders encode micro- and nanoscale clues to their synthesis route and calcination temperature. We trained a ResNet-50 model on 11 commercial-scale U₃O₈ classes, ammonium diuranate (ADU) or uranyl peroxide (H₂O₂) precursors calcined at temperatures ranging from 400 to 750 °C and added a 256-D projection head before the classifier to analyze the learned representation. The best of eight seeds reached 92.4 % accuracy on reserved testing data, but our focus is the structure of the embedding space rather than the accuracy and labels. We quantify class relatedness in the original 256-D space using centroid similarity and distributional distances, and we use Uniform Manifold Approximation Projection (UMAP) for visualization. ‘Unknown’ images from different preparation methods, SEM operators, and from the literature localized near the expected classes under a nearest-centroid analysis without retraining, as well as clustered in similar UMAP space. In conclusion, this embedding-centered workflow complements black-box classification by providing quantitative, similarity-based comparisons of U₃O₈ morphologies and reduces storage space by up to 98 % for image data used in millisecond vector search comparisons.

36 MATERIALS SCIENCE↗

Vehicle energy consumption estimation using large scale simulations and machine learning methods

Vehicle energy consumption and powertrain operations for future vehicle powertrain technologies are predicted through full-vehicle simulations using Autonomie, a simulation tool developed at Argonne National Laboratory. Over 1 million vehicles, with different powertrain types (conventional and numerous electrified powertrains) and component technologies (advanced engines, batteries, electrical machines, light weighting) are simulated for outputs including vehicle fuel economy and cost. Despite the ability of Autonomic to predict fuel economy given various combination of vehicle inputs, the full-vehicle simulation is very time-consuming. It takes approximately 84 h to simulate for 33,060 vehicles even using distributed computing on a cluster of 128 worker nodes. Moreover, Autonomic is unable to populate a continuous output space through simulation since it only yields outputs with discrete values instead of an input-output relation. This paper proposes a novel large-scale learning and prediction process (LSLPP) via machine learning approaches to answer the need for a continuous space of vehicle fuel economy and a more efficient simulation process. By learning in a supervised fashion to discover structures in data, the machine learning approaches seek to train a model that can explain the data with which we can perform prediction over unseen data. As such, we are able to accomplish analytic continuation on the current discrete space of simulation results. In addition, we present a random sampling -based numerosity reduction algorithm incorporated in LSLPP that reduces simulation runs and improves efficiency of data procurement. We analytically derive a theoretical bound in order to guarantee the effectiveness of our algorithm. We also compare random sampling to stratified sampling and suggest sampling strategies. A user-friendly tool is developed to support the process. Our experimental results confirm that the proposed LSLPP is able to greatly accelerate prediction and analysis of vehicle fuel economy. The output of this study is used by the US government to evaluate the impact of its R&D funding on energy security and CO 2 emission as well as for the setting of Corporate Average Fuel Economy (CAFE) standards.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

WE-G-207-05: Relationship Between CT Image Quality, Segmentation Performance, and Quantitative Image Feature Analysis

Purpose: Segmentation quality can affect quantitative image feature analysis. The objective of this study is to examine the relationship between computed tomography (CT) image quality, segmentation performance, and quantitative image feature analysis. Methods: A total of 90 pathology proven breast lesions in 87 dedicated breast CT images were considered. An iterative image reconstruction (IIR) algorithm was used to obtain CT images with different quality. With different combinations of 4 variables in the algorithm, this study obtained a total of 28 different qualities of CT images. Two imaging tasks/objectives were considered: 1) segmentation and 2) classification of the lesion as benign or malignant. Twenty-three image features were extracted after segmentation using a semi-automated algorithm and 5 of them were selected via a feature selection technique. Logistic regression was trained and tested using leave-one-out-cross-validation and its area under the ROC curve (AUC) was recorded. The standard deviation of a homogeneous portion and the gradient of a parenchymal portion of an example breast were used as an estimate of image noise and sharpness. The DICE coefficient was computed using a radiologist’s drawing on the lesion. Mean DICE and AUC were used as performance metrics for each of the 28 reconstructions. The relationship between segmentation and classification performance under different reconstructions were compared. Distributions (median, 95% confidence interval) of DICE and AUC for each reconstruction were also compared. Results: Moderate correlation (Pearson’s rho = 0.43, p-value = 0.02) between DICE and AUC values was found. However, the variation between DICE and AUC values for each reconstruction increased as the image sharpness increased. There was a combination of IIR parameters that resulted in the best segmentation with the worst classification performance. Conclusion: There are certain images that yield better segmentation or classification performance. The best segmentation Result does not necessarily lead to the best classification Result. This work has been supported in part by grants from the NIH R21-EB015053. R Nishikawa is receives royalties form Hologic, Inc.

60 APPLIED LIFE SCIENCES↗

Development of a Full-Scale Connected U-Net for Reflectivity Inpainting in Spaceborne Radar Blind Zones

CloudSat’s Cloud Profiling Radar is a valuable tool for remotely monitoring high-latitude snowfall, but its ability to observe hydrometeor activity near the Earth’s surface is limited by a radar blind zone caused by ground clutter contamination. This study presents the development of a deeply supervised U-Net-style convolutional neural network to predict cold season reflectivity profiles within the blind zone at two Arctic locations. The network learns to predict the presence and intensity of near-surface hydrometeors by coupling latent features encoded in blind zone-aloft clouds with additional context from collocated atmospheric state variables (i.e., temperature, specific humidity, and wind speed). Results show that the U-Net predictions outperform traditional linear extrapolation methods, with low mean absolute error, a 38% higher Sørensen–Dice coefficient, and vertical reflectivity distributions 60% closer to observed values. The U-Net is also able to detect the presence of near-surface cloud with a critical success index (CSI) of 72% and cases of shallow cumuliform snowfall and virga with 18% higher CSI values compared to linear methods. An explainability analysis shows that reflectivity information throughout the scene, especially at cloud edges and at the 1.2-km blind zone threshold, along with atmospheric state variables near the tropopause, are the most significant contributors to model skill. This surface-trained generative inpainting technique has the potential to enhance current and future remote sensing precipitation missions by providing a better understanding of the nonlinear relationship between blind zone reflectivity values and the surrounding atmospheric state.

54 ENVIRONMENTAL SCIENCES↗

From Images to Dark Matter: End-to-end Inference of Substructure from Hundreds of Strong Gravitational Lenses

Abstract Constraining the distribution of small-scale structure in our universe allows us to probe alternatives to the cold dark matter paradigm. Strong gravitational lensing offers a unique window into small dark matter halos (<10 10 M ⊙ ) because these halos impart a gravitational lensing signal even if they do not host luminous galaxies. We create large data sets of strong lensing images with realistic low-mass halos, Hubble Space Telescope (HST) observational effects, and galaxy light from HST’s COSMOS field. Using a simulation-based inference pipeline, we train a neural posterior estimator of the subhalo mass function (SHMF) and place constraints on populations of lenses generated using a separate set of galaxy sources. We find that by combining our network with a hierarchical inference framework, we can both reliably infer the SHMF across a variety of configurations and scale efficiently to populations with hundreds of lenses. By conducting precise inference on large and complex simulated data sets, our method lays a foundation for extracting dark matter constraints from the next generation of wide-field optical imaging surveys.

79 ASTRONOMY AND ASTROPHYSICS↗

Transfer Learning Trained LSTM Models for Household Load Profile Forecasting

Grid edge renewable energy resources, such as rooftop solar photovoltaics, closely interact with consumer load profiles. Therefore, forecasting future electricity demand, ideally at the individual household level, is indispensable. In this paper, we present a transfer learning enhanced household load profile forecasting method. First, we tune a long short-term memory forecasting model to perform day-ahead prediction of household electricity load profiles. Then we improve these individualized models using transfer learning, and we use k-means clustering to create optimal source data sets. We find average improvements of 4.38% (largest improvement of 10.71%) when the entire data set was used to train the source model and 2.45% (largest improvement of 11.57%) in the mean absolute error when households were first clustered and used to train separate source models for each cluster. We find that transfer learning with clustered data can effectively boost the forecasting performance of the LSTM models. We use realistic household power measurements for 148 real residential households in Austin, Texas.

deep learning↗

Automated Integration of Continental-Scale Observations in Near-Real Time for Simulation and Analysis of Biosphere–Atmosphere Interactions

The National Ecological Observatory Network (NEON) is a continental-scale observatory with sites across the US collecting standardized ecological observations that will operate for multiple decades. To maximize the utility of NEON data, we envision edge computing systems that gather, calibrate, aggregate, and ingest measurements in an integrated fashion. Edge systems will employ machine learning methods to cross-calibrate, gap-fill and provision data in near-real time to the NEON Data Portal and to High Performance Computing (HPC) systems, running ensembles of Earth system models (ESMs) that assimilate the data. For the first time gridded EC data products and response functions promise to offset pervasive observational biases through evaluating, benchmarking, optimizing parameters, and training new machine learning parameterizations within ESMs all at the same model-grid scale. Leveraging open-source software for EC data analysis, we are already building software infrastructure for integration of near-real time data streams into the International Land Model Benchmarking (ILAMB) package for use by the wider research community. We will present a perspective on the design and integration of end-to-end infrastructure for data acquisition, edge computing, HPC simulation, analysis, and validation, where Artificial Intelligence (AI) approaches are used throughout the distributed workflow to improve accuracy and computational performance.

Durden, David J.↗

A large-scale evaluation of automated metadata inference approaches on sensors from air handling units

Building automation systems provide abundant sensor data to enable the potential of using data analytics to, among other things, improve the energy efficiency of the building. However, deployment of these applications for buildings, such as, fault detection and diagnosis (FDD) on multiple buildings remains a challenge due to the non-trivial efforts of organizing, managing and extracting metadata associated with sensors (e.g., information about their location, function, etc.), which is required by applications. One of the reasons leading to the problem is that varying conventions, acronyms, and standards are used to define this metadata. To better understand the nature of the problem, as well as the performance and scalability of existing solutions, in this report we implement and test 6 different time-series based metadata inference approaches on sensors from 614 air handling units (AHU) instrumented in 35 building sites accounting for more than 400 buildings distributed across United States of America. We infer 12 types of sensors and actuators in AHUs required by a rule-based FDD application: AHU performance and assessment rules (APAR). Our results show that: (1) the average performance of these approaches in terms of accuracy is similar across building sites, though there is significant variance; (2) the expected accuracy of classifying the type of points required by APAR for a new unseen building is, on average, 75%; (3) the performance of the model does not decrease as long as training data and testing data are extracted from adjacent months.

42 ENGINEERING↗

Automated defect identification in electroluminescence images of solar modules

Solar photovoltaic (PV) modules are susceptible to manufacturing defects, mishandling problems or extreme weather events that can limit energy production or cause early device failure. Trained professionals use electroluminescence (EL) images to identify defects in modules, however, field surveys or inline image acquisition can generate millions of EL images, which are infeasible to analyze by rote inspection. Here, we develop a rapid automatic computer vision pipeline (~0.5 seconds/module) to analyze EL images and identify defects including cracks, intra-cell defects, oxygen-induced defects, and solder disconnections. Defect identification is achieved with a machine learning model (Random Forest, ResNet models and YOLO) trained on 762 manually-labeled EL images of PV modules. We compare model performance on an imbalanced real-world validation set containing 134 EL images and determine that ResNet18 and YOLO are the optimal models; we next evaluated these models on a dedicated testing set (129 module images) with resulting macro F1 scores of 0.83 (ResNet18) and 0.78 (YOLO). Using a field EL survey of a PV power plant damaged in a vegetation fire, we analyze 18,954 EL images (2.4 million cells) and inspect the spatial distribution of defects on the solar modules. The results find increased frequency of ‘crack’, ‘solder’ and ‘intra-cell’ defects on the edges of the solar module closest to the ground after fire. We also find an abnormal increase of striation rings on cells which were assumed to be caused mainly in fabrication process. Our methods are published as open-source software. It can also be used to identify other kinds of defects or process different types of solar cells with minor modification on models by transfer learning.

14 SOLAR ENERGY↗

A Machine Learning Framework for Predicting Microphysical Properties of Ice Crystals From Cloud Particle Imagery

The microphysical properties of ice crystals are important because they significantly alter the radiative properties and spatiotemporal distributions of clouds, which in turn strongly affect Earth's climate. However, it is challenging to measure key properties of ice crystals, such as mass or morphological features. Here, we present a proof-of-concept framework for predicting three-dimensional (3D) microphysical properties of ice crystals from in situ two-dimensional (2D) imagery. First, we computationally generated synthetic ice crystals using 3D modeling software along with geometric parameters estimated from the 2021 Ice Cryo-Encapsulation Balloon (ICEBall) field campaign. Then, we used synthetic crystals to train machine learning (ML) models to predict effective density ($ρ_e$), effective surface area ($A_e$), and number of bullets ($N_b$) from synthetic rosette imagery. On unseen synthetic images, our ML models accurately predicted ice crystal properties. ResNet-18 performed best, achieving $R^2$ values of 0.99 and 0.98 for $ρ_e$ and $A_e$, respectively, and MAE of 0.10 for mathematical equation in single view tasks. Stereo view ResNet-18 further reduced RMSE by 40% for $ρ_e$ and $A_e$ and reduced MAE by 0.08 for $N_b$. This work provides a novel ML-driven framework for estimating ice microphysical properties from in situ imagery, which will allow for downstream constraints on microphysical parameterizations, such as the mass-size relationship.

Ko, J. [Columbia Univ., New York, NY (United State↗

Large-Scale Solar Panel Mapping from Aerial Images Using Deep Convolutional Networks

Up-to-date maps of installed solar photovoltaic panels are a critical input for policy and financial assessment of solar distributed generation. However, such maps for large areas are not available. With high coverage and low cost, aerial images enable large-scale mapping, bit it is highly difficult to automatically identify solar panels from images, which are small objects with varying appearances dispersed in complex scenes. We introduce a new approach based on deep convolutional networks, which effectively learns to delineate solar panels in aerial scenes. The approach has successfully mapped solar panels in imagery covering 200 square kilometers in two cities, using only 12 square kilometers of training data that are manually labeled.

Yuan, Jiangye↗

Simulation-Based Parameter Optimization Framework for Large-Scale Hybrid Smart Grid Communications Systems Design

The design of reliable, dynamic, fault-tolerant hybrid smart grid communication networks is a challenge to achieve for autonomous power grids. Hybrid networks use different communications technologies for different area networks. A simulation-based parameter optimization framework is proposed to tune parameters of hybrid communication technologies to achieve the optimal network performance. It consists of three main components: a parallel executor used to speedup a list of simulations; a sampler running simulations using the parallel executor at each generation; and a hybrid stochastic optimization algorithm for tuning configurable parameters of hybrid designs and applications. The proposed hybrid metaheuristic optimization algorithm combines an evolutionary algorithm with a gradient method to quickly achieve an approximately global optimum solution. Three optimization test functions are employed to train the adjustable parameters of the hybrid algorithm. Results show the proposed parameter optimization framework can help the designer choose the right hybrid architecture with an optimal parameter set for a large-scale broadband PLC-WiMAX hybrid smart grid communication network.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Cybersecurity for Energy Systems: Foundational Concepts for Bangladesh Electric Utilities [Slides]

This slide deck was developed for a training for the Northern Electricity Supply Company (NESCO) in Rajshahi, Bangladesh. The topics include: understanding the cybersecurity landscape in power utilities, fundamentals of cybersecurity for power utilities, regulatory compliance and standards, and best practices and strategies. The concepts in this slide deck are relevant for electric utilities throughout Bangladesh.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Oakland University Cybersecurity Center (Final Scientific/Technical Report)

This report summarizes the outcomes of Award DE-CR0000023, “Oakland University Cybersecurity Center,” a 31-month project funded by the U.S. Department of Energy Office of Cybersecurity, Energy Security, and Emergency Response (CESER). The project addressed cybersecurity risks facing small and medium-sized manufacturers (SMMs) transitioning to Industry 4.0. The project integrated customer discovery, applied research, and cybersecurity training development. A total of 51 cybersecurity assessments identified significant gaps in baseline practices, incident response, and workforce capability. Research efforts produced a scalable mitigation framework tailored to SMM environments, and workforce analysis identified persistent talent gaps. Eight cybersecurity training modules were developed and deployed via Oakland University’s Professional and Continuing Education (PACE) platform. All objectives were completed, with 98.93% federal budget utilization and cost share exceeding requirements. The project establishes a scalable model for strengthening cybersecurity resilience and workforce capacity across U.S. manufacturing supply chains.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Autonomous Utility Pole Identification

The implementation of small unmanned aerial systems (sUAS) for the purpose of powerline inspection is an emerging concept among utility companies. Such operations can produce a significant amount of useful data for the purpose of training different machine learning models or keep track of the integrity of electrical infrastructure. As such, this system is developed for the purpose of properly cataloging and leveraging this collected data.

24 POWER TRANSMISSION AND DISTRIBUTION↗