Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “distributed inference”

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 397 records · Page 22
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Particle control via cryopumping and its impact on the edge plasma profiles of Alcator C-Mod

At the high n e proposed for high-field fusion reactors, it is uncertain whether ionization, as opposed to plasma transport, will be most influential in determining n e at the pedestal and separatrix. A database of Alcator C-Mod discharges is analyzed to evaluate the impact of source modification via cryopumping. The database contains similarly-shaped H-modes at fixed I P = 0.8 MA and B t = 5.4 T, spanning a large range in P net and ionization. Measurements from an edge Thomson scattering system are combined with those from a midplane-viewing Ly α camera to evaluate changes to n e and T e in response to changes in ionization rates, S ion ∙ $n^{sep}_e$ and $T^{ped}_e$ are found to be most sensitive to changes to $S^{sep}_{ion}$, as opposed to $n^{ped}_e$ and $T^{sep}_e$. Dimensionless quantities, namely α MHD and v*, are found to regulate attainable pedestal values. Select discharges at different values of P net and in different pumping configurations are analyzed further using SOLPS-ITER. It is determined that changes to plasma transport coefficients are required to self-consistently model both plasma and neutral edge dynamics. Pumping is found to modify the poloidal distribution of atomic neutral density, n 0 , along the separatrix, increasing n 0 at the active X-point. Opaqueness to neutrals from high n e in the divertor is found to play a role in mediating neutral penetration lengths and hence, the poloidal distribution of neutrals along the separatrix. Pumped discharges thus require a larger particle diffusion coefficient than that inferred purely from 1D experimental profiles at the outer midplane.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2024

Uncertainty quantification in multivariable regression for material property prediction with Bayesian neural networks

With the increased use of data-driven approaches and machine learning-based methods in material science, the importance of reliable uncertainty quantification (UQ) of the predicted variables for informed decision-making cannot be overstated. UQ in material property prediction poses unique challenges, including multi-scale and multi-physics nature of materials, intricate interactions between numerous factors, limited availability of large curated datasets, etc. In this work, we introduce a physics-informed Bayesian Neural Networks (BNNs) approach for UQ, which integrates knowledge from governing laws in materials to guide the models toward physically consistent predictions. To evaluate the approach, we present case studies for predicting the creep rupture life of steel alloys. Experimental validation with three datasets of creep tests demonstrates that this method produces point predictions and uncertainty estimations that are competitive or exceed the performance of conventional UQ methods such as Gaussian Process Regression. Additionally, we evaluate the suitability of employing UQ in an active learning scenario and report competitive performance. The most promising framework for creep life prediction is BNNs based on Markov Chain Monte Carlo approximation of the posterior distribution of network parameters, as it provided more reliable results in comparison to BNNs based on variational inference approximation or related NNs with probabilistic outputs.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Nuclear masses learned from a probabilistic neural network

Machine learning methods and uncertainty quantification have been gaining interest throughout the last several years in low-energy nuclear physics. In particular, Gaussian processes and Bayesian neural networks have increasingly been applied to improve mass model predictions while providing well-quantified uncertainties. In this work, we use the probabilistic Mixture Density Network (MDN) to directly predict the mass excess of the 2016 Atomic Mass Evaluation within the range of measured data, and we extrapolate the inferred models beyond available experimental data. The MDN provides not only mean values but also full posterior distributions both within the training set and extrapolated testing set. We show that the addition of physical information to the feature space increases the accuracy of the match to the training data as well as provides for more physically meaningful extrapolations beyond the the limits of experimental data.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2020

Bayesian inference in band excitation scanning probe microscopy for optimal dynamic model selection in imaging

The universal tendency in scanning probe microscopy (SPM) over the last two decades is to transition from simple 2D imaging to complex detection and spectroscopic imaging modes. The emergence of complex SPM engines brings forth the challenge of reliable data interpretation, i.e., conversion from detected signals to descriptors specific to tip–surface interactions and subsequently to material’s properties. In this work, we implemented a Bayesian inference approach for the analysis of the image formation mechanisms in band excitation SPM. Compared to the point estimates in classical functional fit approaches, Bayesian inference allows for the incorporation of extant knowledge of materials and probe behavior in the form of corresponding prior distribution and return the information on the material functionality in the form of readily interpretable posterior distributions. We explore the nonlinear mechanical behaviors spatially in a classical ferroelectric material, PbTiO 3 . We observe the non-trivial evolution of the Duffing stiffness term and the nonlinearity of the sample surface, determine spatial clustering of the nonlinear response, and perform a Landau analysis on predicting the nonlinear coefficient, which indicates that ferroelectric behavior can be a cause of the observed results. These observations suggest that the spectrum of anomalous behaviors at the ferroelectric domain walls may be broader than previously believed and can extend to non-conventional mechanical properties in addition to static and microwave conductance.

36 MATERIALS SCIENCE↗
DOE OSTI2022

EdgeAI: Machine learning via direct attached accelerator for streaming data processing at high shot rate x-ray free-electron lasers

We present a case for low batch-size inference with the potential for adaptive training of a lean encoder model. We do so in the context of a paradigmatic example of machine learning as applied in data acquisition at high data velocity scientific user facilities such as the Linac Coherent Light Source-II x-ray Free-Electron Laser. We discuss how a low-latency inference model operating at the data acquisition edge can capitalize on the naturally stochastic nature of such sources. We simulate the method of attosecond angular streaking to produce representative results whereby simulated input data reproduce high-resolution ground truth probability distributions. By minimizing the mean-squared error between the decoded output of the latent representation and the ground truth distributions, we ensure that the encoding layers and resulting latent representation maintains full fidelity for any downstream task, be it classification or regression. We present throughput results for data-parallel inference of various batch sizes, some with throughput exceeding 100 k images per second. We also show in situ training below 10 s per epoch for the full encoder–decoder model as would be relevant for streaming and adaptive real-time data production at our nation’s scientific light sources.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

A practical and efficient approach for Bayesian quantum state estimation

Bayesian inference is a powerful paradigm for quantum state tomography, treating uncertainty in meaningful and informative ways. Yet the numerical challenges associated with sampling from complex probability distributions hampers Bayesian tomography in practical settings. In this article, we introduce an improved, self-contained approach for Bayesian quantum state estimation. Leveraging advances in machine learning and statistics, our formulation relies on highly efficient preconditioned Crank–Nicolson sampling and a pseudo-likelihood. We theoretically analyze the computational cost, and provide explicit examples of inference for both actual and simulated datasets, illustrating improved performance with respect to existing approaches.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2023

DGaaS: GPU as a Service on Distributed Computing System

In the rapidly evolving landscape of scientific computing, Graphics Processing Units (GPUs) have become indispensable for their unparalleled ability to handle parallel tasks in complex calculations, simulations, and data analysis. Their utility is further magnified in machine learning and AI applications, where they significantly accelerate model training and predictive analytics. Within this context, the Triton Inference Server emerges as a pivotal open-source tool, specializing in AI inferencing and optimizing GPU utilization across various platforms and frameworks. This paper presents an in-depth study on distributed High Throughput Computing (HTC), specifically focusing on the HTCondor framework and its resource provisioning tools, GlideinWMS and HEPCloud. These systems enable large-scale scientific experiments like CMS and DUNE to efficiently access and utilize vast computational resources. The paper explores the core architectural components of GlideinWMS, including jobs, user pools, and worker nodes, and discusses their integration with GPUs and the Triton server. The primary aim of this research is to develop a solution that optimizes GPU utilization by leveraging Glideins and containers. This approach allows computational jobs, particularly those involving AI models, to use GPUs only when essential, thereby facilitating efficient sharing of limited GPU resources. To validate this architecture, the study conducted three key tests involving custom scripts, container-based servers, and Triton server deployments. However, the study faces challenges, notably in locating the Triton server and ensuring secure remote access. To address these issues, future work will focus on developing a proxy mechanism and enhancing security protocols. In conclusion, this study offers a comprehensive roadmap for effective and efficient GPU utilization in distributed High Throughput Computing. It aims to contribute significantly to the scientific community by solving pressing problems and implementing robust solutions in collaboration with the GlideinWMS and HEPCloud teams. The research sets the stage for a more efficient, scalable, and cost-effective paradigm in scientific computing.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Resolving Nonequilibrium Shape Variations among Millions of Gold Nanoparticles

Nanoparticles, exhibiting functionally relevant structural heterogeneity, are at the forefront of cutting-edge research. Now, high-throughput single-particle imaging (SPI) with X-ray free-electron lasers (XFELs) creates opportunities for recovering the shape distributions of millions of particles that exhibit functionally relevant structural heterogeneity. To realize this potential, three challenges have to be overcome: (1) simultaneous parametrization of structural variability in real and reciprocal spaces; (2) efficiently inferring the latent parameters of each SPI measurement; (3) scaling up comparisons between 10 5 structural models and 10 6 XFEL-SPI measurements. Here, we describe how we overcame these three challenges to resolve the nonequilibrium shape distributions within millions of gold nanoparticles imaged at the European XFEL. These shape distributions allowed us to quantify the degree of asymmetry in these particles, discover a relatively stable “shape envelope” among nanoparticles, discern finite-size effects related to shape-controlling surfactants, and extrapolate nanoparticles’ shapes to their idealized thermodynamic limit. Ultimately, these demonstrations show that XFEL SPI can help transform nanoparticle shape characterization from anecdotally interesting to statistically meaningful.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Aerosol Activation in Radiation Fog at the Atmospheric Radiation Program Southern Great Plains Site

Environmental supersaturation is a key parameter in the formation of fog and clouds, yet it cannot be measured directly and must be inferred. Calculating the ambient supersaturation in fog requires knowledge of the aerosol hygroscopicity as well as particle size distribution, and relatively few values have been reported in the literature. Here we use κ-Köhler theory to derive aerosol activation properties based on particle hygroscopicity and dry particle size distributions, and then estimate the effective peak supersaturation during eight cases of radiative fog at the Southern Great Plains site in rural north-central Oklahoma, USA. The mean hygroscopicity parameter κ of particles likely to act as cloud condensation nuclei varied from 0.14 to 0.43. Ambient effective peak supersaturation during the fog episodes was between 0.01%–0.07%, with most values below 0.04%. Furthermore, the minimum 50% dry activation diameter generally ranged between 300–400 nm with little activation of particles with diameters below 300 nm.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Mass calibration of distant SPT galaxy clusters through expanded weak-lensing follow-up observations with HST, VLT, & Gemini-South

Expanding from previous work, we present weak-lensing (WL) measurements for a total sample of 30 distant (z median = 0.93) massive galaxy clusters from the South Pole Telescope Sunyaev–Zel’dovich (SPT-SZ) Survey, measuring galaxy shapes in Hubble Space Telescope (HST) Advanced Camera for Surveys images. We remove cluster members and preferentially select z ≳ 1.4 background galaxies via V - I colour, employing deep photometry from VLT/FORS2 and Gemini-South/GMOS. We apply revised calibrations for the WL shape measurements and the source redshift distribution to estimate the cluster masses. In combination with earlier Magellan/Megacam results for lower-redshifts clusters, we infer refined constraints on the scaling relation between the SZ detection significance and the cluster mass, in particular regarding its redshift evolution. The mass scale inferred from the WL data is lower by a factor $0.76^{+0.10}_{-0.14}$ (at our pivot redshift z = 0.6) compared to what would be needed to reconcile a flat Planck νΛCDM cosmology (in which the sum of the neutrino masses is a free parameter) with the observed SPT-SZ cluster counts. In order to sensitively test the level of (dis-)agreement between SPT clusters and Planck, further expanded WL follow-up samples are needed.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

Modeling regional precipitation over the Indus River basin of Pakistan using statistical downscaling

Complex processes govern spatiotemporal distribution of precipitation within the high-mountainous headwater regions (commonly known as the upper Indus basin (UIB)), of the Indus River basin of Pakistan. Reliable precipitation simulations particularly over the UIB present a major scientific challenge due to regional complexity and inadequate observational coverage. Here, we present a statistical downscaling approach to model observed precipitation of the entire Indus basin, with a focus on UIB within available data constraints. Taking advantage of recent high altitude (HA) observatories, we perform precipitation regionalization using K-means cluster analysis to demonstrate effectiveness of low-altitude stations to provide useful precipitation inferences over more uncertain and hydrologically important HA of the UIB. We further employ generalized linear models (GLM) with gamma and Tweedie distributions to identify major dynamic and thermodynamic drivers from a reanalysis dataset within a robust cross-validation framework that explain observed spatiotemporal precipitation patterns across the Indus basin. Final statistical models demonstrate higher predictability to resolve precipitation variability over wetter southern Himalayans and different lower Indus regions, by mainly using different dynamic predictors. The modeling framework also shows an adequate performance over more complex and uncertain trans-Himalayans and the northwestern regions of the UIB, particularly during the seasons dominated by the westerly circulations. However, the cryosphere-dominated trans-Himalayan regions, which largely govern the basin hydrology, require relatively complex models that contain dynamic and thermodynamic circulations. Furthermore, we also analyzed relevant atmospheric circulations during precipitation anomalies over the UIB, to evaluate physical consistency of the statistical models, as an additional measure of reliability. Overall, our results suggest that such circulation-based statistical downscaling has the potential to improve our understanding towards distinct features of the regional-scale precipitation across the upper and lower Indus basin. Additionally, such understanding should help to assess the response of this complex, data-scarce, and climate-sensitive river basin amid future climatic changes, to serve communal and scientific interests.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Robust inference of ecosystem soil water stress from eddy covariance data

Eddy covariance data are invaluable for determining ecosystem water use strategies under soil water stress. However, existing stress inference methods require numerous subjective data processing and model specification assumptions whose effect on the inferred soil water stress signal is rarely quantified. These uncertainties may confound the stress inference and the generalization of ecosystem water use strategies across multiple sites and studies. In this research, we quantify the sensitivity of soil water stress signals inferred from eddy covariance data to the prevailing data and modeling assumptions (i.e., their robustness) to compile a comprehensive list of sites with robust soil water stress signals and assess the performance of current stress inference methods. To accomplish this, we identify the most prevalent assumptions from the literature and perform a digital factorial experiment to extract probability distributions of plausible soil water stress signals and model performance at 151 FLUXNET2015 and AmeriFlux-FLUXNET sites. Here, we develop a new framework that summarizes these probability distributions to classify and rank the robustness of each site’s soil water stress signal, which we display with a user-friendly heat map. We estimate that only 5%–36% of sites exhibit a robust soil water stress signal due to deficient model performance and poorly constrained ecosystem water use parameters. We also find that the lack of robustness is site-specific, which undermines grouping stress signals by broad ecosystem categories or comparing results across studies with differing assumptions. Lastly, existing stress inference methods appear better suited for eddy covariance sites with grass/annual vegetation. Our findings call for more careful and consistent inference of ecosystem water stress from eddy covariance data.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Learning Latent Interactions for Event Identification via Graph Neural Networks and PMU Data

Phasor measurement units (PMUs) are being widely installed on power systems, providing a unique opportunity to enhance wide-area situational awareness. One essential application is the use of PMU data for real-time event identification. However, how to take full advantage of all PMU data in event identification is still an open problem. Thus, we propose a novel method that performs event identification by mining interaction graphs among different PMUs. The proposed interaction graph inference method follows an entirely data-driven manner without knowing the physical topology. Moreover, unlike previous works that treat interactive learning and event identification as two different stages, our method learns interactions jointly with the identification task, thereby improving the accuracy of graph learning and ensuring seamless integration between the two stages. Moreover, to capture multi-scale event patterns, a dilated inception-based method is investigated to perform feature extraction of PMU data. To test the proposed data-driven approach, a large real-world dataset from tens of PMU sources and the corresponding event logs have been utilized in this work. We report numerical results validate that our method has higher classification accuracy compared to previous methods.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2022

An Accurate, Error-Tolerant, and Energy-Efficient Neural Network Inference Engine Based on SONOS Analog Memory

In this work, we demonstrate SONOS (silicon-oxide-nitrideoxide- silicon) analog memory arrays that are optimized for neural network inference. The devices are fabricated in a 40nm process and operated in the subthreshold regime for in-memory matrix multiplication. Subthreshold operation enables low conductances to be implemented with low error, which matches the typical weight distribution of neural networks, which is heavily skewed toward near-zero values. This leads to high accuracy in the presence of programming errors and process variations. We simulate the end-to-end neural network inference accuracy, accounting for the measured programming error, read noise, and retention loss in a fabricated SONOS array. Evaluated on the ImageNet dataset using ResNet50, the accuracy using a SONOS system is within 2.16% of floating-point accuracy without any retraining. The unique error properties and high On/Off ratio of the SONOS device allow scaling to large arrays without bit slicing, and enable an inference architecture that achieves 20 TOPS/W on ResNet50, a >10× gain in energy efficiency over state-of-the-art digital and analog inference accelerators.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Early retirement of power plants in climate mitigation scenarios

International efforts to avoid dangerous climate change aim for large and rapid reductions of fossil fuel CO 2 emissions worldwide, including nearly complete decarbonization of the electric power sector. However, achieving such rapid reductions may depend on early retirement of coal- and natural gas-fired power plants. Here, we analyze future fossil fuel electricity demand in 171 energy-emissions scenarios from Integrated Assessment Models (IAMs), evaluating the implicit retirements and/or reduced operation of generating infrastructure. Although IAMs calculate retirements endogenously, the structure and methods of each model differ; we use a standard approach to infer retirements in outputs from all six major IAMs and—unlike the IAMs themselves—we begin with the age distribution and region-specific operating capacities of the existing power fleet. We find that coal-fired power plants in scenarios consistent with international climate targets (i.e. keeping global warming well-below 2 °C or 1.5 °C) retire one to three decades earlier than historically has been the case. If plants are built to meet projected fossil electricity demand and instead allowed to operate at the level and over the lifetimes they have historically, the roughly 200 Gt CO 2 of additional emissions this century would be incompatible with keeping global warming well-below 2 °C. Thus, ambitious climate mitigation scenarios entail drastic, and perhaps un-appreciated, changes in the operating and/or retirement schedules of power infrastructure.

20 FOSSIL-FUELED POWER PLANTS↗
DOE OSTI2021

Nonequilibrium information entropy approach to ternary fission of actinides

Ternary fission of actinides probes the state of the nucleus at scission. Light clusters are produced in space and time very close to the scission point. Within the nonequilibrium statistical operator method, a generalized Gibbs distribution is constructed from the information given by the observed yields of isotopes. Using this relevant statistical operator, yields are calculated taking excited states and continuum correlations into account, in accordance with the virial expansion of the equation of state. Furthermore, clusters with mass number A≤10 are well described using the nonequilibrium generalizations of temperature and chemical potentials. Improving the virial expansion, in-medium effects may become of importance in determining the contribution of weakly bound states and continuum correlations to the intrinsic partition function. Yields of larger clusters, which fail to reach this quasiequilibrium form of the relevant distribution, are described by nucleation kinetics, and a saddle-to-scission relaxation time of about 7000 fm/c is inferred. Light-charged particle emission, described by reaction kinetics and virial expansions, may therefore be regarded as a very important tool to probe the nonequilibrium time evolution of actinide nuclei during fission.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2020

Quantification of Ebola virus replication kinetics in vitro

Mathematical modelling has successfully been used to provide quantitative descriptions of many viral infections, but for the Ebola virus, which requires biosafety level 4 facilities for experimentation, modelling can play a crucial role. Ebola virus modelling efforts have primarily focused on in vivo virus kinetics, e.g., in animal models, to aid the development of antivirals and vaccines. But, thus far, these studies have not yielded a detailed specification of the infection cycle, which could provide a foundational description of the virus kinetics and thus a deeper understanding of their clinical manifestation. Here, we obtain a diverse experimental data set of the Ebola virus infection in vitro, and then make use of Bayesian inference methods to fully identify parameters in a mathematical model of the infection. Our results provide insights into the distribution of time an infected cell spends in the eclipse phase (the period between infection and the start of virus production), as well as the rate at which infectious virions lose infectivity. We suggest how these results can be used in future models to describe co-infection with defective interfering particles, which are an emerging alternative therapeutic.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Multi-fidelity is the new annealing: Gradient-free learning of posterior densities via transport maps

To tackle concentrated, multi-modal Bayesian inference problems, we propose using an annealed importance sampling procedure. To do this, we form a sequence of annealed distributions and employ transport maps to act as a surrogate of each distribution. This process is demonstrated on a few examples, favorably showing its potential for efficiently parallelizing the process of PDE evaluations and allowing for surrogates that we can sample from exactly.

van Bloemen Waanders, Bart G [Sandia National Labo↗