Search NASA⌕ Search

SEARCH · Search NASA

Results for “data driven model”

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 1,495 records · Page 83

A High-level Design for Bidirectional Data Streaming to High-Performance Computing Systems from External Science Facilities

Cutting-edge science is increasingly data-driven due to the emergence of scientific machine learning models that can guide scientists toward fruitful areas of exploration. Experimental science facilities such as light and neutron sources, particle colliders, and radio astronomy telescopes are also producing raw measurement data at rates that exceed available data storage and computing capacity at those facilities. As a result, scientific workflows are being developed that concurrently couple experiments at science facilities with high-performance computing (HPC) facilities to enable analysis of experimental data while the experiment is ongoing, and where analysis results are potentially fed back to the experiment for use in experimental control and/or steering in a time-sensitive manner. Our goal is to design, prototype, and deploy a new capability for the Oak Ridge Leadership Computing Facility (OLCF) that enables such workflows through support for bidirectional, memory-based streaming of data from external experiments into and out of OLCF HPC systems. This high-level design document describes the related work and motivating use cases that inform our understanding of the technical requirements for this capability, and describes a proposed architectural solution that meets these requirements and our plans for demonstrating the capability.

97 MATHEMATICS AND COMPUTING↗

Evaluation of Converter Performance Considering Static and Dynamic Device Part-to-Part Variability

This paper presents a methodology to incorporate and analyze the impact of semiconductor device part-to-part variation on power converter performance. By integrating extensive static and dynamic device characterization data with an automated compact model generation process that reflects manufacturing variability, device models with inherent variability features are utilized in converter simulations for a comprehensive assessment of performance impacts. The traditional converter performance evaluation process typically yields fixed efficiency values, often dismissing the inherent part-to-part variability caused by the manufacturing process of semiconductor devices. To address this limitation, a large population of devices was characterized to capture variations in static parameters-such as transfer, output, and capacitance characteristics-as well as dynamic behaviors, including switching losses. This data-driven approach enables the development of individual compact models, which were then integrated into converter simulations to evaluate efficiency ranges rather than single point estimated values. The converter simulation results show that part-to-part component variation can lead to significant efficiency deviations, exceeding several percentage points in high-power conversion applications. By offering a more accurate representation of converter behavior under real-world manufacturing conditions, this methodology enables designers to anticipate performance variability, improving the robustness of power converter designs.

device characterization↗

A data-driven approach to construct a quantitative relationship between microstructural features of fatigue cracks and contact acoustic nonlinearity

This study demonstrates the feasibility of a data-driven approach to construct a quantitative relationship between nonlinear acoustic parameters and microstructural features of contact interfaces. The near-surface nonlinearity is measured using dynamic acousto-elastic testing (DAET) with a surface wave probe, while the microstructural features are extracted from scanning electron microscopy (SEM) images of fatigue cracks. Four aluminum alloy samples, each having a fatigue crack are prepared. Six local nonlinearity parameters are measured at different locations along the crack propagation direction. A total of 40 local measurements are acquired. A principal component analysis (PCA) reveals that all six nonlinearity parameters are correlated and hence can be replaced by one principal component (PC). Fifteen crack micro-geometrical features at each measurement point were extracted from the SEM images. Regression analysis is used to relate the PC of the nonlinearity parameters to the microstructural features at the crack interface. We compare three regression models that take variable selection into account: stepwise multiple linear regression (MLR), stepwise principal component regression (PCR), and least absolute shrinkage and selection operator (LASSO). Despite having different principles, the three predictive models identify two features as the most significant in predicting the interface nonlinearity: the crack aperture (opening) distribution and the distance to the crack tip. The differences between the three models and the physical interpretation of the data-driven predictions are discussed.

36 MATERIALS SCIENCE↗

Towards a Semantically-Enabled Control Strategy for Building Simulations: Integration of Semantic Technologies and Model Predictive Control

State-of-the-art building simulation control methods incorporate physical constraints into their mathematical models, but omit implicit constraints associated with policies of operation and dependency relationships among rules representing those constraints. To overcome these shortcomings, there is a recent trend in enabling the control strategies with inference-based rule checking capabilities. One solution is to exploit semantic web technologies in building simulation control. Such approaches provide the tools for semantic modeling of domains, and the ability to deduce new information based on the models through use of Description Logic (DL). In a step toward enabling this capability, this paper presents a cross-disciplinary data-driven control strategy for building energy management simulation that integrates semantic modeling and formal rule checking mechanisms into a Model Predictive Control (MPC) formulation. The results show that MPC provides superior levels of performance when initial conditions and inputs are derived from inference-based rules.

Delgoshaei, Parastoo↗

CMLM (Co-Optimized Machine-Learned Manifolds) [SWR-23-41]

Co-optimized Machine-Learned Manifolds (CMLM) is a data-driven approach for developing reduced-order manifold models for high-dimensional chemically reacting systems. It involves a specially designed neural network, the training of which simultaneously optimizes linear combinations of species that define the manifold, nonlinear mapping to outputs of interest such as reaction rates, and (optionally) subfilter closure for large eddy simulation. This software package provides an implementation of the CMLM approach in Python using the PyTorch machine learning library. A few example cases are included, showing how the tool can be applied to different types of data from 0D and 1D reacting simulations performed using Cantera. The neural networks can be saved in a format that is readable by the Pele suite of combustion solvers for use in reacting computational fluid dynamics simulations. This software repository contains several python scripts to perform various tasks associated with the Co-optimized Machine Learned Manifolds (CMLM) model, which is described in Perry, Henry de Frahan, and Yellapantula, CNF, 2022 (https://doi.org/10.1016/j.combustflame.2022.112286). This includes not only the code that defines the CMLM model, but also scripts to generate suitable training data, scripts to pre-process the data, scripts to train the CMLM model, and scripts to plot the output, as well as various other helper files. The scripts depend on several commonly used python libraries for data analysis and chemical reaction computations. The trained models that result from this tool are designed to work with the an interface being implemented in the Pele suite of reacting flow solvers (https://github.com/AMReX-Combustion).

Perry, Bruce↗

Machine learning accelerated turbulence modeling of transient flashing jets

We report modeling the sudden depressurization of superheated liquids through nozzles is a challenge because the pressure drop causes rapid flash boiling of the liquid. The resulting jet usually demonstrates a wide range of structures, including ligaments and droplets, due to both mechanical and thermodynamic effects. As the simulation comprises increasingly numerous phenomena, the computational cost begins to increase. One way to moderate the additional cost is to use machine learning surrogacy for specific elements of the calculation. This study presents a machine learning-assisted computational fluid dynamics approach for simulating the atomization of flashing liquids accounting for distinct stages, from primary atomization to secondary breakup to small droplets using the Σ - Y model coupled with the homogeneous relaxation model. Notably, the models for thermodynamic non-equilibrium (HRM) and Σ - Y are coupled, for the first time, with a deep neural network that simulates the turbulence quantities, which are then used in the prediction of superheated liquid jet atomization. The data-driven component of this method is used for turbulence modeling, avoiding the solution of the two-equation turbulence model typically used for Reynolds-averaged Navier-Stokes simulations for these problems. Both the accuracy and speed of the hybrid approach are evaluated, demonstrating adequate accuracy and at least 25% faster computational fluid dynamics simulations than the traditional approach. This acceleration suggests that perhaps additional components of the calculation could be replaced for even further benefit. Published under an exclusive license by AIP Publishing.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

The Cannon: A data-driven approach to stellar label determination

New spectroscopic surveys offer the promise of stellar parameters and abundances (“stellar labels”) for hundreds of thousands of stars; this poses a formidable spectral modeling challenge. In many cases, there is a subset of reference objects for which the stellar labels are known with high(er) fidelity. We take advantage of this with The Cannon, a new data-driven approach for determining stellar labels from spectroscopic data. The Cannon learns from the “known” labels of reference stars how the continuum-normalized spectra depend on these labels by fitting a flexible model at each wavelength; then, The Cannon uses this model to derive labels for the remaining survey stars. We illustrate The Cannon by training the model on only 542 stars in 19 clusters as reference objects, with T{sub eff}, log g, and [Fe/H] as the labels, and then applying it to the spectra of 55,000 stars from APOGEE DR10. The Cannon is very accurate. Its stellar labels compare well to the stars for which APOGEE pipeline (ASPCAP) labels are provided in DR10, with rms differences that are basically identical to the stated ASPCAP uncertainties. Beyond the reference labels, The Cannon makes no use of stellar models nor any line-list, but needs a set of reference objects that span label-space. The Cannon performs well at lower signal-to-noise, as it delivers comparably good labels even at one-ninth the APOGEE observing time. We discuss the limitations of The Cannon and its future potential, particularly, to bring different spectroscopic surveys onto a consistent scale of stellar labels.

79 ASTRONOMY AND ASTROPHYSICS↗

Dark matter interpretations of ATLAS searches for the electroweak production of supersymmetric particles in $ \sqrt{s}=8 $ TeV proton-proton collisions

A selection of searches by the ATLAS experiment at the LHC for the electroweak production of SUSY particles are used to study their impact on the constraints on dark matter candidates. The searches use 20 fb -1 of proton-proton collision data at √s=8 TeV. A likelihood-driven scan of a five-dimensional effective model focusing on the gaugino-higgsino and Higgs sector of the phenomenological minimal supersymmetric Standard Model is performed. This scan uses data from direct dark matter detection experiments, the relic dark matter density and precision flavour physics results. Further constraints from the ATLAS Higgs mass measurement and SUSY searches at LEP are also applied. A subset of models selected from this scan are used to assess the impact of the selected ATLAS searches in this five-dimensional parameter space. These ATLAS searches substantially impact those models for which the mass m($\tilde{χ}$$0\atop{1}$) of the lightest neutralino is less than 65 GeV, excluding 86% of such models. The searches have limited impact on models with larger m($\tilde{χ}$$0\atop{1}$) due to either heavy electroweakinos or compressed mass spectra where the mass splittings between the produced particles and the lightest supersymmetric particle is small.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Modeling of Coastal Effluent Transport: an Approach to Linking of Far-field and Near-field Models

One of the challenges in effluent transport modeling in coastal tidal environments is the proper calculation of initial dilution in connection with the far-field transport model. In this study, an approach of external linkage of far-field and near-field effluent transport models is presented, and applied to simulate the effluent transport in the Port Angeles Harbor, Washington in the Strait of Juan de Fuca. A near-field plume model was used to calculate the effluent initial dilution and a three-dimensional (3-D) hydrodynamic model was developed to simulate the tidal circulation and far-field effluent transport in the Port Angeles Harbor. In the present study, the hydrodynamic model was driven by tides and surface winds. Observed water surface elevation and velocity data were used to calibrate the model over a period covering the neap-spring tidal cycle. The model was also validated with observed surface drogue trajectory data. The model successfully reproduced the tidal dynamics in the study area and good agreements between model results and observed data were obtained. This study demonstrated that the linkage between the near-field and far-field models in effluent transport modeling can be achieved through iteratively adjusting the model grid sizes such that the far-field modeled dilution ratio and effluent concentration in the effluent discharge model grid cell match the concentration calculated by the near-field plume model.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DeepCare: Improving Patient Care using Deep Learning on Electronic Health Records

Coordinating patient care using electronic health records (EHR) data presents an exciting but formidable opportunity in data extraction, analysis and modeling. Traditional methods use a manual feature driven approach to model patients with age, family history and symptoms to predict disease outcomes. We propose a novel approach to model patients based on their streaming electronic health records data combined with information from medical knowledge bases, which has been gained over years of medical research. Using a combination of representation learning and long short term memory (LSTM) networks we plan to model patient evolution over time, leading to more accurate and individualized predictive models for patient’s diseases. Our approach will be transformative in providing critical decision support for patient care, enabling accurate understanding and evolution of diseases in patients.

60 APPLIED LIFE SCIENCES↗

Multifidelity deep operator networks for data-driven and physics-informed problems

Operator learning for complex nonlinear systems is increasingly common in modeling multi-physics and multi-scale systems. However, training such high-dimensional operators requires a large amount of expensive, high-fidelity data, either from experiments or simulations. In this work, we present a composite Deep Operator Network (DeepONet) for learning using two datasets with different levels of fidelity to accurately learn complex operators when sufficient high-fidelity data is not available. Additionally, we demonstrate that the presence of low-fidelity data can improve the predictions of physics-informed learning with DeepONets. We demonstrate the new multi-fidelity training in diverse examples, including modeling of the ice-sheet dynamics of the Humboldt glacier, Greenland, using two different fidelity models and also using the same physical model at two different resolutions.

97 MATHEMATICS AND COMPUTING↗

Global soil consumption of atmospheric carbon monoxide: an analysis using a process-based biogeochemistry model

Carbon monoxide (CO) plays an important role in controlling the oxidizing capacity of the atmosphere by reacting with OH radicals that affect atmospheric methane (CH 4 ) dynamics. We develop a process-based biogeochemistry model to quantify the CO exchange between soils and the atmosphere with a 5 min internal time step at the global scale. The model is parameterized using the CO flux data from the field and laboratory experiments for 11 representative ecosystem types. The model is then extrapolated to global terrestrial ecosystems using monthly climate forcing data. Global soil gross consumption, gross production, and net flux of the atmospheric CO are estimated to be from -197 to -180, 34 to 36, and -163 to -145 Tg CO yr -1 (1 Tg = 10 12 g), respectively, when the model is driven with satellite-based atmospheric CO concentration data during 2000–2013. Tropical evergreen forest, savanna and deciduous forest areas are the largest sinks at 123 Tg CO yr -1 . The soil CO gross consumption is sensitive to air temperature and atmospheric CO concentration, while the gross production is sensitive to soil organic carbon (SOC) stock and air temperature. By assuming that the spatially distributed atmospheric CO concentrations (~ 128 ppbv) are not changing over time, the global mean CO net deposition velocity is estimated to be 0.16–0.19 mm s -1 during the 20th century. Under the future climate scenarios, the CO deposition velocity will increase at a rate of 0.0002–0.0013 mm s -1 yr -1 during 2014–2100, reaching 0.20–0.30 mm s -1 by the end of the 21st century, primarily due to the increasing temperature. Areas near the Equator, the eastern US, Europe and eastern Asia will be the largest sinks due to optimum soil moisture and high temperature. The annual global soil net flux of atmospheric CO is primarily controlled by air temperature, soil temperature, SOC and atmospheric CO concentrations, while its monthly variation is mainly determined by air temperature, precipitation, soil temperature and soil moisture.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Factorization Machine‐Based Active Learning for Functional Materials Design with Optimal Initial Data

The optimization of functional materials is important to enhance their properties, but their complex geometries pose great challenges to optimization. Data-driven algorithms efficiently navigate such complex design spaces by learning relationships between material structures and performance metrics to discover high-performance functional materials. Surrogate-based active learning, continually improving its surrogate model by iteratively including high-quality data points, has emerged as a cost-effective data-driven approach. Furthermore, it can be coupled with quantum computing to enhance optimization processes, especially when paired with a special form of surrogate model (i.e., quadratic unconstrained binary optimization), formulated by factorization machine (FM). However, current practices often overlook the variability in design space sizes when determining the initial data size for optimization. In this work, we investigate the optimal initial data sizes required for efficient convergence across various design space sizes. By employing averaged piecewise linear regression, we identify initiation points where convergence begins, highlighting the crucial role of employing adequate initial data in achieving efficient optimization. These results contribute to the efficient optimization of functional materials by ensuring faster convergence and reducing computational costs in FM-based active learning.

active learning↗

Machine learning of consistent thermodynamic models using automatic differentiation

In this study, we propose a data-driven method to describe consistent equations of state (EOS) for arbitrary systems. Complex EOS are traditionally obtained by fitting suitable analytical expressions to thermophysical data. A key aspect of EOS is that the relationships between state variables are given by derivatives of the system free energy. In this work, we model the free energy with an artificial neural network and utilize automatic differentiation to directly learn the derivatives of the free energy. We demonstrate this approach on two different systems, the analytic van der Waals EOS and published data for the Lennard-Jones fluid, and we show that it is advantageous over direct learning of thermodynamic properties (i.e., not as derivatives of the free energy but as independent properties), in terms of both accuracy and the exact preservation of the Maxwell relations. Furthermore, the method implicitly provides the free energy of a system without explicit integration.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

GCAM-Demeter-LU

This GCAM-Demeter-LU is made available under the Open Data Commons Attribution License:http://opendatacommons.org/licenses/by/1.0/. The dataset includes the projected global gridded land cover (excluding the Antarctic) for the period of 2015-2100 at 0.05-degree resolution and 5-year time step under fifteen SSP-RCP scenarios driven by five GCMs (i.e., gfdl, hadgem, ipsl, miroc, and noresm), using the Global Change Analysis Model (GCAM) and a geospatial downscaling model (Demeter). The data has been stored in self-describing NetCDF format. The dataset also includes the mean and standard deviation of the results driven by the five GCMs. More specifically, the data in each year includes grid-explicit fraction (in percent) of each of the 32 plant functional types that are widely used in current Earth system models. The NetCDF files are named as “GCAM_Demeter_LU_SSP_RCP_Model_Year.nc”, where “SSP” and “RCP” denote the SSP and RCP scenarios, including 'ssp1_rcp26', 'ssp1_rcp45', 'ssp1_rcp60', 'ssp2_rcp26', 'ssp2_rcp45', 'ssp2_rcp60', 'ssp3_rcp45', 'ssp3_rcp60', 'ssp4_rcp26', 'ssp4_rcp45', 'ssp4_rcp60', 'ssp5_rcp26', 'ssp5_rcp45', 'ssp5_rcp60', and 'ssp5_rcp85'. “Year” denotes the year of the land use data, and “GCM” denotes the source driving forcing data from five global climate models (gfdl, hadgem, ipsl, miroc, and noresm), or the mean (“modelmean”) and the standard deviation (“modelstd”) of the results from the five GCMs.See https://github.com/JGCRI/chen_et_al_2020a for details on how to reproduce this data. Rights and permissions Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver http://creativecommons.org/publicdomain/zero/1.0/ applies to the metadata files associated with this article

99 GENERAL AND MISCELLANEOUS↗

Non-intrusive reduced order modeling of natural convection in porous media using convolutional autoencoders: Comparison with linear subspace techniques

Natural convection in porous media is a highly nonlinear multiphysical problem relevant to many engineering applications (e.g., the process of CO 2 sequestration). Here, we extend and present a non-intrusive reduced order model of natural convection in porous media employing deep convolutional autoencoders for the compression and reconstruction and either radial basis function (RBF) interpolation or artificial neural networks (ANNs) for mapping parameters of partial differential equations (PDEs) on the corresponding nonlinear manifolds. To benchmark our approach, we also describe linear compression and reconstruction processes relying on proper orthogonal decomposition (POD) and ANNs. Further, we present comprehensive comparisons among different models through three benchmark problems. The reduced order models, linear and nonlinear approaches, are much faster than the finite element model, obtaining a maximum speed-up of 7 × 10 6 because our framework is not bound by the Courant–Friedrichs–Lewy condition; hence, it could deliver quantities of interest at any given time contrary to the finite element model. Our model’s accuracy still lies within a relative error of 7% in the worst-case scenario. We illustrate that, in specific settings, the nonlinear approach outperforms its linear counterpart and vice versa. We hypothesize that a visual comparison between principal component analysis (PCA) and t-Distributed Stochastic Neighbor Embedding (t-SNE) could indicate which method will perform better prior to employing any specific compression strategy.

97 MATHEMATICS AND COMPUTING↗