SEARCH · Engineering Papers

Results for “distributed machine learning”

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 559 records · Page 31

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Sub-microsecond Transformers for Jet Tagging on FPGAs

We present the first sub-microsecond transformer implementation on an FPGA achieving competitive performance for state-of-the-art high-energy physics benchmarks. Transformers have shown exceptional performance on multiple tasks in modern machine learning applications, including jet tagging at the CERN Large Hadron Collider (LHC). However, their computational complexity prohibits use in real-time applications, such as the hardware trigger system of the collider experiments up until now. In this work, we demonstrate the first application of transformers for jet tagging on FPGAs, achieving $\mathcal{O}(100)$ nanosecond latency with superior performance compared to alternative baseline models. We leverage high-granularity quantization and distributed arithmetic optimization to fit the entire transformer model on a single FPGA, achieving the required throughput and latency. Furthermore, we add multi-head attention and linear attention support to hls4ml, making our work accessible to the broader fast machine learning community. This work advances the next-generation trigger systems for the High Luminosity LHC, enabling the use of transformers for real-time applications in high-energy physics and beyond.

Laatu, Lauri [Imperial Coll., London]↗

DOE OSTI2023

Constraining the Multiscale Structure of Geophysical Fields in Machine Learning: The Case of Precipitation

The use of deep-learning algorithms for estimating the value of geophysical variables from remotely sensed information is rapidly expanding. The typical objective function minimized in such algorithms is the mean square error (MSE), which is known to lead to smooth estimates with compressed dynamical range as compared to the true distribution of the variable of interest. Here, we introduce and evaluate alternative objective functions, focusing on the retrieval of precipitation rates from satellite passive microwave radiometric measurements using a deep convolutional neural network. For this testbed application, the results show that explicitly imposing the preservation of the statistical distribution and spatial wavelet power spectrum of the target variable allows to accurately reproduce extreme values and sharp gradients across multiple scales in this study.

58 GEOSCIENCES↗

DOE OSTI2023

Dynamic Model Agnostic Reliability Evaluation of Machine-Learning Models Integrated in Instrumentation & Control Systems

In recent years, the field of machine learning (ML), specifically neural networks, has grown significantly and has spurred research in its applicability to digital instrumentation and control systems (DI&C). While ML models have shown promise in operational contexts, the trustworthiness of using such algorithms has not been adequately assessed. Failures of ML integrated systems are not well understood, and the lack of comprehensive risk modeling can degrade the trustworthiness in these systems. In recent reports by the National Institute for Standards and Technology (NIST) [1] and the Nuclear Regulatory Commission (NRC) [2], they indicate that trustworthiness in ML is a critical barrier and will play a vital role in the safe, accountable, and secure operation of intelligent systems. Thus, in this work, we demonstrate a dynamic model-agnostic method to quantify the relative reliability of AI/ML predictions by incorporating out-of-distribution (OOD) detection on the training dataset. It is well documented that most ML algorithms excel at interpolation (or near-interpolation) tasks but experience significant performance degradation at extrapolation. The method, referenced as the Laplacian distributed decay for reliability (LADDR), determines the difference between the operational and training datasets which can used to the relative reliability of AI/ML predictions. LADDR is then demonstrated on a feedforward neural network based digital twin used for the prediction of safety significant factors during a loss-of-flow transient. LADDR is used to demonstrate how training data can be used as evidence to support the relative reliability of ML/AI predictions enhancing the overall trustworthiness of the system.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

DOE OSTI2020

Nonlinear sparse Bayesian learning for physics-based models

This paper addresses the issue of overfitting while calibrating unknown parameters of over-parameterized physics-based models with noisy and incomplete observations. Here, a semi-analytical Bayesian framework of nonlinear sparse Bayesian learning (NSBL) is proposed to identify sparsity among model parameters during Bayesian inversion. NSBL offers significant advantages over machine learning algorithm of sparse Bayesian learning (SBL) for physics-based models, such as 1) the likelihood function or the posterior parameter distribution is not required to be Gaussian, and 2) prior parameter knowledge is incorporated into sparse learning (i.e. not all parameters are treated as questionable). NSBL employs the concept of automatic relevance determination (ARD) to facilitate sparsity among questionable parameters through parameterized prior distributions. The analytical tractability of NSBL is enabled by employing Gaussian ARD priors and by building a Gaussian mixture-model approximation of the posterior parameter distribution that excludes the contribution of ARD priors. Subsequently, type-II maximum likelihood is executed using Newton's method whereby the evidence and its gradient and Hessian information are computed in a semi-analytical fashion. We show numerically and analytically that SBL is a special case of NSBL for linear regression models. Subsequently, a linear regression example involving multimodality in both parameter posterior pdf and model evidence is considered to demonstrate the performance of NSBL in cases where SBL is inapplicable. Next, NSBL is applied to identify sparsity among the damping coefficients of a mass-spring-damper model of a shear building frame. These numerical studies demonstrate the robustness and efficiency of NSBL in alleviating overfitting during Bayesian inversion of nonlinear physics-based models.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2020

Improving Enzyme Optimum Temperature Prediction with Resampling Strategies and Ensemble Learning

Accurate prediction of the optimal catalytic temperature ( T opt ) of enzymes is vital in biotechnology, as enzymes with high T opt values are desired for enhanced reaction rates. Recently, a machine learning method (temperature optima for microorganisms and enzymes, TOME) for predicting T opt was developed. TOME was trained on a normally distributed data set with a median T opt of 37 °C and less than 5% of T opt values above 85 °C, limiting the method’s predictive capabilities for thermostable enzymes. Due to the distribution of the training data, the mean squared error on T opt values greater than 85 °C is nearly an order of magnitude higher than the error on values between 30 and 50 °C. Here, we apply ensemble learning and resampling strategies that tackle the data imbalance to significantly decrease the error on high T opt values (>85 °C) by 60% and increase the overall R 2 value from 0.527 to 0.632. The revised method, temperature optima for enzymes with resampling (TOMER), and the resampling strategies applied in this work are freely available to other researchers as Python packages on GitHub.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2022

Catlas: an automated framework for catalyst discovery demonstrated for direct syngas conversion

Catalyst discovery is paramount to support access to energy and key chemical feedstocks in a post fossil fuel era. Exhaustive computational searches of large material design spaces using ab initio methods like density functional theory (DFT) are infeasible. We seek to explore large design spaces at relatively low computational cost by leveraging large, generalized, graph-based machine learning (ML) models, which are pretrained and therefore require no upfront data collection or training. We present Catlas, a framework that distributes and automates the generation of adsorbate-surface configurations and ML inference of DFT energies to achieve this goal. Catlas is open source, making ML assisted catalyst screenings easy and available to all. To demonstrate its efficacy, we use Catlas to explore catalyst candidates for the direct conversion of syngas to multi-carbon oxygenates. For this case study, we explore 947 stable/metastable binary, transition metal intermetallics as possible catalyst candidates. On this subset of materials, we are able to predict the adsorption energy of key descriptors, *CO and *OH, with good accuracy (0.16, 0.14 eV MAE, respectively). Using the projected selectivity towards C 2+ oxygenates from an existing microkinetic model, we identified 144 candidate materials. For 10 promising candidates, DFT calculations reveal a good correlation with our assessment using ML. Finally, among the top elemental combinations were Pt–Ti, Pd–V, Ni–Nb, and Ti–Zn, all of which appear unexplored experimentally.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2020

miniGAN: a proxy application for generative adversarial networks

miniGAN is a python-based machine learning proxy application for generative adversarial networks, developed through the Exascale Computing Project's (ECP) ExaLearn project. It will be included in the main ECP proxy application and the machine learning proxy application suite. It is a proxy for ECP cosmological(CosmoFlow, ExaGAN) and wind energy(ExaWind) applications. miniGAN will be distributed to ECP hardware vendors as part of hardware codesign. miniGAN uses the Numpy/PyTorch/TensorFlow/Keras/Horovod frameworks and libraries. It also relies on the Kokkos and Kokkos-Kernels packages developed here at Sandia Labs. SAND2020-2038 M Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-NA0003525.

Ellis, John↗

DOE OSTI2023

An Update on the Geothermal Data Repository's Data Standards and Pipelines: Geospatial Data and Distributed Acoustic Sensing Data

The Department of Energy's (DOE) Geothermal Data Repository (GDR) team has implemented or is currently implementing data standards and automated data pipelines for the following geothermal data types: 1) drilling data, 2) geospatial datasets, and 3) Distributed Acoustic Sensing (DAS) data. These data standards and pipelines are intended to improve the real-world applicability of geothermal machine learning outputs through improving the quality of data. More specifically, through standardizing high-value datasets, the GDR is reducing project-specific data curation requirements, allowing more time to be spent on actual research. By automating this process, the burden of standardization is taken off of the user, overall increasing the availability of standardized data. This paper provides an update on the GDR's transition toward data standardization through automated data pipelines and calls for feedback from the community on how the GDR team can improve this process.

cloud-optimized↗

DOE OSTI2025

I/O in Machine Learning Applications on HPC Systems: A 360-degree Survey

Growing interest in Artificial Intelligence (AI) has resulted in a surge in demand for faster methods of Machine Learning (ML) model training and inference. This demand for speed has prompted the use of high performance computing (HPC) systems that excel in managing distributed workloads. Because data is the main fuel for AI applications, the performance of the storage and I/O subsystem of HPC systems is critical. In the past, HPC applications accessed large portions of data written by simulations or experiments or ingested data for visualizations or analysis tasks. ML workloads perform small reads spread across a large number of random files. This shift of I/O access patterns poses several challenges to modern parallel storage systems. In this paper, we survey I/O in ML applications on HPC systems, and target literature within a 6-year time window from 2019 to 2024. We define the scope of the survey, provide an overview of the common phases of ML, review available profilers and benchmarks, examine the I/O patterns encountered during offline data preparation, training, and inference, and explore I/O optimizations utilized in modern ML frameworks and proposed in recent literature. Lastly, we seek to expose research gaps that could spawn further R&D.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2020

Augmenting machine learning of energy landscapes with local structural information

We present a machine learning approach for accurately predicting formation energies of binary compounds in the context of crystal structure predictions. The success of any machine learning model depends significantly on the choice of representation used to encode the relevant physical information into machine-learnable data. We test different representation schemes based on partial radial and angular distribution functions (RDF+ADF) on Al–Ni and Cd–Te structures generated using our genetic algorithm for structure prediction. We observe a remarkable improvement in predictive accuracy upon transitioning from global to atom-centered representations, resulting in a threefold decrease in prediction errors. We show that a support vector regression model using a combination of atomic radial and angular distribution functions performs best at the formation energy prediction task, providing small root mean squared errors of 3.9 meV/atom and 10.9 meV/atom for Al–Ni and Cd–Te, respectively. We test the performance of our models against common traditional descriptors and find that RDF- and ADF-based representations significantly outperform many of those in the prediction of formation energies. The high accuracy of predictions makes our machine learning models great candidates for the exploration of energy landscapes.

Honrao, Shreyas J. (ORCID:0000000166155257)↗

DOE OSTI2024

Reconstruction of the 4D beam matrix

The widely used transverse parameters characterizing particle beams are the Twiss parameters. These parameters can be measured experimentally but they do not fully characterize the beam since they do not account for possible correlations in particle distribution between two transverse coordinates. These correlations may occur due to uncompensated magnetic field at the cathode or misalignment of focusing quadrupoles in the transport beamline. We test a novel diagnostic for diagnosing full 4D beam matrix which may be used to identify such imperfections. The diagnostic is based on transporting the beam through the beamline which includes a quadrupole and a skew quadrupole magnets and measuring the resulting 2D beam distribution at the screen downstream. Such a measurement can be viewed as measuring a 2D projection of the 4D distribution. Different settings of the quads provide measurements of different slices of the phase space. The reconstruction of the original beam matrix from a number of measurements is done using machine learning algorithm, which provides a fast and reliable way of reconstruction for an arbitrary configuration of the scanning beamline. In August 2024, we set up the diagnostic beamline to perform a quadrupole scan of the beam. The setup includes a skew quadrupole, a regular quadrupole, and a screen. The images on the screen were post-processed to remove experimental artifacts and enhance contrast by eliminating background noise outside the core of the distribution=. The rms parameters of the distribution were then calculated and used as inputs for the reconstruction algorithm. This algorithm attempts to determine the initial beam matrix that produces expected images on the screen closely matching the observed images across all quadrupole settings. The algorithm found a solution in which the expected rms parameters closely align with the observations. Validation of the results is planned for FY25.

43 PARTICLE ACCELERATORS↗

DOE OSTI2024

Multi-Edge Graph Convolutional Networks for Power Systems

The exponential electrification of transportation has contributed to highly intermittent load variations in the distribution grid. This uncertainty has raised challenges for distribution system operation and control. Accurate nodal voltage estimation is highly essential for the safe and reliable operation of the grid. Graph convolutional networks have been used in machine-learning-based models for power grid applications like voltage estimation for their ability to capture the network topology of the grid. This paper presents a novel multi-edge graph convolutional layer that considers resistance and reactance as edge attributes. This layer is created by modifying the message-passing function within the graph convolutional network. The novel layer is then used to create a multi-edge graph convolutional network-based surrogate model for estimating voltage in the distribution network with highly uncertain electric vehicle loads. Results indicate improved performance of the multi-edge graph convolutional network model when compared to a standard graph convolutional network model.

Ravi, Abhijith↗

DOE OSTI2021

Real-Time Edge AI for Distributed Systems (READS): Progress on Beam Loss De-Blending for the Fermilab Main Injector and Recycler

The Fermilab Main Injector enclosure houses two accelerators, the Main Injector and Recycler. During normal operation, high intensity proton beams exist simultaneously in both. The two accelerators share the same beam loss monitors (BLM) and monitoring system. Beam losses in the Main Injector enclosure are monitored for tuning the accelerators and machine protection. Losses are currently attributed to a specific machine based on timing. However, this method alone is insufficient and often inaccurate, resulting in more difficult machine tuning and unnecessary machine downtime. Machine experts can often distinguish the correct source of beam loss. This suggests a machine learning (ML) model may be producible to help de-blend losses between machines. Work is underway as part of the Fermilab Real-time Edge AI for Distributed Systems Project (READS) to develop a ML empowered system that collects streamed BLM data and additional machine readings to infer in real-time, which machine generated beam loss.

43 PARTICLE ACCELERATORS↗

DOE OSTI2022

Star–Galaxy Image Separation with Computationally Efficient Gaussian Process Classification

Abstract We introduce a novel method for discerning optical telescope images of stars from those of galaxies using Gaussian processes (GPs). Although applications of GPs often struggle in high-dimensional data modalities such as optical image classification, we show that a low-dimensional embedding of images into a metric space defined by the principal components of the data suffices to produce high-quality predictions from real large-scale survey data. We develop a novel method of GP classification hyperparameter training that scales approximately linearly in the number of image observations, which allows for application of GP models to large-size Hyper Suprime-Cam Subaru Strategic Program data. In our experiments, we evaluate the performance of a principal component analysis embedded GP predictive model against other machine-learning algorithms, including a convolutional neural network and an image photometric morphology discriminator. Our analysis shows that our methods compare favorably with current methods in optical image classification while producing posterior distributions from the GP regression that can be used to quantify object classification uncertainty. We further describe how classification uncertainty can be used to efficiently parse large-scale survey imaging data to produce high-confidence object catalogs.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2022

Transient Data Library of Solar Grid Integrated Distributed System

This submission contains an open-source library of transient events in distributed system with high solar PV. The library includes the collected data, related documents and scripts for loading the data. The data library is built for transient event detection and machine learning based analysis algorithm development. The data was collected via both field test and software simulation. The units for the data are included in the data file headers for each data series. A text editor or spreadsheet software, such as Excel, and Matlab is required to view the data.

algorithms↗

DOE OSTI2026

Reconstruction of Six-Dimensional Phase Space

A phase space is a mathematical representation of all possible physical states of a system. Particle beams at Fermilab exist within a six-dimensional (6D) phase space defined by three positional components, (x, y, z) and three momentum components, (px, py, pz). To reconstruct this space implies taking measurement data from detectors and mapping out particle behavior using computational methods. The beam detectors, however, are only able to detect spatial distribution among the events of the beam, therefore being limited to positional data. Also, due to the vast number of events in a particle beam, it is extremely difficult to analyze and differentiate every single one’s behavior. However, with Machine Learning (ML), which can distinguish between patterns and map out particle behavior more efficiently. We first used the particle beam software, G4beamline, to simulate a 10,000-event muon beam, adjusting parameters such as initial momentum magnitude (p¬0) and virtual detector position. Using ten virtual detectors, we analyzed p0 values such that minimum 9,990 events were analyzed by every detector. We then input the data from these beam simulations to a C++ program, that randomly selects 100 events, and creates a 2D histogram based on spatial distribution, detector position, and event intensity. This process is repeated 100 times to create 100 histograms per p0 value. These images were then input to a modified ResNet18 Convolutional Neural Network (CNN) for training, and to predict p0 from some unseen set of histograms. The model was accurate when trained on momentum increments of 5 MeV/c and provided with denser training samples around highly variable test values. These results displayed machine learning being able to accurately predict p0 from being trained on different particle behaviors.

Shirlee, Jermain [Fermilab]↗

DOE OSTI2025

Scalability analysis of heavy-duty gas turbines using data-driven machine learning

With the increasing integration of variable renewable energy sources into power systems, the role of flexible power generation technologies like gas turbines (GT) in rapid grid balancing remains crucial. This sustained importance underscores the need for scaled and precise modeling of GT to ensure effective integration within evolving energy frameworks. While physics-driven GT models integrate thermodynamics, fluid dynamics, and combustion principles, they often rely on approximate mathematical representations to accommodate scaling that may not capture the actual complex dynamics for GTs and inertial effects associated to GTs with different ratings. In this study, a data-driven model is proposed using machine learning (ML) techniques to conduct GT scalability analysis and performance evaluation with high accuracy. The ML model, trained on data from various operating conditions and performance parameters, aims to uncover intricate relationships and patterns, resembling GT characteristics at different scales (ratings). The model is developed to capture complex system interaction and to adapt to changing operational scenarios at different capacities, providing valuable insights of power system dynamics. In this study, the real-time digital simulator platform was employed to generate training data for the ML model and assess its dynamic characteristics. The ultimate objective was to develop a detailed modeling framework based on governing equations and data-driven ML capable of predicting key performance indicators, in thermal systems such as GTs, including power output, speed, fuel consumption, and exhaust temperature under diverse operating conditions at different scales. The developed ML framework demonstrated high accuracy, with mean relative errors for GT power prediction, reference speed, exhaust temperature, and compressor pressure ratio (CPR) parameters consistently below 0.1% across typical load fluctuation scenarios. Maximum deviations were limited to approximately 0.5 K for exhaust temperature and 0.009 for CPR, underscoring the model’s ability to replicating dynamic GT behavior with high precision. The adaptability of the ML model enables its application across diverse operational conditions and its extension to other thermal systems. By leveraging advanced ML techniques, this study presents a robust and scalable modeling framework that enhances GT simulation precision, facilitating improved integration into evolving power systems.

24 POWER TRANSMISSION AND DISTRIBUTION↗

DOE OSTI2025

Harnessing the Power of Machine Learning and Omics to Identify Environmental Regulation on Microbial Functional Composition for Soil C, N, and P Cycling

Microbial enzyme-mediated soil organic matter (SOM) decomposition regulates many key ecosystem functions, such as elemental cycling, soil carbon sequestration, and soil fertility. However, representing microbial processes in Earth system models (ESMs) remains challenging due to a limited understanding of the spatial patterns of diverse microbial functions responsible for soil carbon (C), nitrogen (N), and phosphorus (P) cycling as well as the underlying mechanisms regulating their relative abundances across various environments. We collected published metagenomics data across the continental US (CONUS) to identify hundreds of microbial genes involved in soil C, N, and P cycling and grouped them into eight enzyme functional classes (EFCs). Each EFC represented a group of gene-encoded potential enzymes that decompose similar soil compounds. By integrating the abundances of omics-informed EFCs with the corresponding environmental information, we trained a machine learning (ML) model to identify key edaphic, climate, and vegetation factors regulating the abundances of each EFC. Quantitative analysis of effects of these factors revealed that the spatial distribution of eight EFCs for soil C, N, and P cycling across CONUS reflected potential resource optimization strategies of microbial communities under nutrient limitation, preferential organic-mineral associations, and climatological stresses. This insight, together with the interpreted ML tool and the CONUS-level benchmark for EFCs abundances, paves the way for parameterizing environmental-regulated microbial functional dynamics in biogeochemical models.

machine learning↗