Search NASA⌕ Search

SEARCH · Search NASA

Results for “Ensemble computing.”

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 685 records · Page 38

Computing diffuse reflection from particulate planetary surface with a new function

An equation is derived to compute the amount of diffuse light reflected by a particulate surface, such as on Mars or an asteroid. The method traces the paths of rays within an ensemble of randomly shaped grains and finds the eventual probability of emission. The amount of diffuse, unpolarized emitted light is obtained in terms of the real index of refraction, the imaginary index, and the average diameter of particles making up the surface. The equation is used to compute the empirical rule for obtaining the planetary albedo from the slope of its polarization curve. Accuracy of the equation, estimated at + or - 4%, seems justified because of quantitative agreement with experimental measures of the empirical rule. It is also shown that the equation can be applied to bubble-enclosing surfaces such as volcanic foams. Results for the indices of the moon, Mars, Io, and Europa are obtained and compared with other data.

Wolff, M.↗

Applications of flow models to the generation of correlated lattice QCD ensembles

Machine-learned normalizing flows can be used in the context of lattice quantum field theory to generate statistically correlated ensembles of lattice gauge fields at different action parameters. This work demonstrates how these correlations can be exploited for variance reduction in the computation of observables. Three different proof-of-concept applications are demonstrated using a novel residual flow architecture: continuum limits of gauge theories, the mass dependence of QCD observables, and hadronic matrix elements based on the Feynman–Hellmann approach. In all three cases, it is shown that statistical uncertainties are significantly reduced when machine-learned flows are incorporated as compared with the same calculations performed with uncorrelated ensembles or direct reweighting. Published by the American Physical Society 2024

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Multifidelity Ensemble Kalman Filtering Using Surrogate Models Defined by Theory-Guided Autoencoders

Data assimilation is a Bayesian inference process that obtains an enhanced understanding of a physical system of interest by fusing information from an inexact physics-based model, and from noisy sparse observations of reality. The multifidelity ensemble Kalman filter (MFEnKF) recently developed by the authors combines a full-order physical model and a hierarchy of reduced order surrogate models in order to increase the computational efficiency of data assimilation. The standard MFEnKF uses linear couplings between models, and is statistically optimal in case of Gaussian probability densities. This work extends the MFEnKF into to make use of a broader class of surrogate model such as those based on machine learning methods such as autoencoders non-linear couplings in between the model hierarchies. We identify the right-invertibility property for autoencoders as being a key predictor of success in the forecasting power of autoencoder-based reduced order models. We propose a methodology that allows us to construct reduced order surrogate models that are more accurate than the ones obtained via conventional linear methods. Numerical experiments with the canonical Lorenz'96 model illustrate that nonlinear surrogates perform better than linear projection-based ones in the context of multifidelity ensemble Kalman filtering. We additionality show a large-scale proof-of-concept result with the quasi-geostrophic equations, showing the competitiveness of the method with a traditional reduced order model-based MFEnKF.

97 MATHEMATICS AND COMPUTING↗

Look-up Table for Lidar and Polarimeter Liquid Water Cloud Studies

- Combined lidar and polarimeter retrievals of aerosol, cloud and ocean microphysical properties involve single scattering cloud calculations that are time consuming. - We create a look-up table to speed up calculations for water droplets in the atmosphere. The gained increase in computational efficiency is up to 10^(4). - In our new Lorenz-Mie look-up table we tabulate the light scattering by an ensemble of homogeneous isotropic spheres at wavelengths starting from 0.35 µm. - The look-up table covers liquid water cloud particles with radii in the range of 0.001—500 µm. - The covered complex refractive indices range from 1.25 to 1.36 for the real part and from 0 to 0.001 for the imaginary part. - Using look-up table, we can precisely compute inherent optical properties for the particle size distributions ranging up to 100 µm for the effective radius and up to 0.6 for the effective variance. - We test wavelengths from 0.35 to 2.3 µm and find that the P12 element of the normalized scattering matrix is precise to within 1% for 96.7% of cases; the absorption and backscatter coefficients, and lidar ratio are precise to within 1% for 99% of cases; the P11 , P33 and P34 elements of the normalized scattering matrix, the asymmetry parameter, the extinction and scattering coefficients are precise to within 1% for all cases. - We provide an example of using the look-up table with in situ measurements to determine agreement with remote sensing. - The look-up table along with examples of its use soon will be freely available at https://science.larc.nasa.gov/polarimetry

Eduard Chemyakin↗

Data Assimilation - Advances and Applications

This presentation provides an overview of data assimilation (model calibration) for complex computer experiments. Calibration refers to the process of probabilistically constraining uncertain physics/engineering model inputs to be consistent with observed experimental data. An initial probability distribution for these parameters is updated using the experimental information. Utilization of surrogate models and empirical adjustment for model form error in code calibration form the basis for the statistical methodology considered. The role of probabilistic code calibration in supporting code validation is discussed. Incorporation of model form uncertainty in rigorous uncertainty quantification (UQ) analyses is also addressed. Design criteria used within a batch sequential design algorithm are introduced for efficiently achieving predictive maturity and improved code calibration. Predictive maturity refers to obtaining stable predictive inference with calibrated computer codes. These approaches allow for augmentation of initial experiment designs for collecting new physical data. A standard framework for data assimilation is presented and techniques for updating the posterior distribution of the state variables based on particle filtering and the ensemble Kalman filter are introduced.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Model Calibration for Cancer Risk Projections According to Uncertain Data

This paper presents forward and inverse formulations for the calibration of computational models according to uncertain data. Uncertainty in the data might be caused by a poor metrology system, measurement noise, missing or uncontrollable input variables, or by the inability to directly measure the inputs and/or outputs of interest. The forward approach performs the calibration in the space of the model’s output thereby requiring repeated model simulations. Conversely, the inverse approach leverages an ensemble of solutions to an inverse problem in order to perform the calibration in the space of the model’s parameters. As such, the computational demands of the inverse approach are considerably lower. These strategies are applied to the calibration of a radiation model that in-forms cancer risk projections for future deep space missions.

uncertainty quantification↗

From deep TLS validation to ensembles of atomic models built from elemental motions

Procedures are described for extracting the vibration and libration parameters corresponding to a given set of TLS matrices and their simultaneous validation. Knowledge of these parameters allows the generation of structural ensembles corresponding to these matrices. The translation–libration–screw model first introduced by Cruickshank, Schomaker and Trueblood describes the concerted motions of atomic groups. Using TLS models can improve the agreement between calculated and experimental diffraction data. Because the T, L and S matrices describe a combination of atomic vibrations and librations, TLS models can also potentially shed light on molecular mechanisms involving correlated motions. However, this use of TLS models in mechanistic studies is hampered by the difficulties in translating the results of refinement into molecular movement or a structural ensemble. To convert the matrices into a constituent molecular movement, the matrix elements must satisfy several conditions. Refining the T, L and S matrix elements as independent parameters without taking these conditions into account may result in matrices that do not represent concerted molecular movements. Here, a mathematical framework and the computational tools to analyze TLS matrices, resulting in either explicit decomposition into descriptions of the underlying motions or a report of broken conditions, are described. The description of valid underlying motions can then be output as a structural ensemble. All methods are implemented as part of the PHENIX project.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

HydraGNN_Predictive_GFM_2024 - Ensemble of predictive graph foundation models for ground state atomistic materials modeling

We provide the ensemble of fifteen pre-trained graph foundation models (GFMs) for atomistic materials modeling applications. Each one of the fifteen GFMs has been trained on five open-source datasets that (once aggregated) amount to over 154 million atomistic structures, which cover over two-thirds of the natural elements of the periodic table and that comprises a broad set of organic and inorganic compounds. This vast set of atomistic structures comprises ground state configurations that are dynamically stable (i.e., equilibrated structures with atomic forces approximately close to zero values) as well as dynamically unstable structures (i.e., non-equilibrium structures with non-negligible non-zero values of atomic forces). The ensemble of datasets aggregated does NOT include excited states. The datasets have been curated to remove atomistic structures with spectral norm of the force tensor above 100 eV/angstrom. Moreover, a linear term of the energy was computed for each dataset using a linear regression model that uses the chemical concentration of each natural element as regressor. The linear term predicted by the linear regression model has been subtracted from each original energy value to perform a re-alignment of the energy values across different electronic structures approximation theories performed to generate the diverse multi-source, multi-fidelity datasets. The folder "ADIOS_files" contains the set of pre-processed datasets in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used for the development and training of GFMs in this work. The "ADIOS_files" directory contains 6 sub-directories named as follows: - ANI1x-v3.bp - MPTrj-v3.bp - OC2020-20M-v3.bp - OC2020-v3.bp - OC2022-v3.bp - qm7x-v3.bp Each sub-directory contains the pre-processed datasets converted in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used to the development, training, and performance testing of the ensemble go predictive graph foundation models. Each GFM was developed using HydraGNN (https://github.com/ORNL/HydraGNN) as underlying graph neural network (GNN) architecture. The multi-task learning (MTL) capability of HydraGNN was used to simultaneously train the GFMs on labeled values for direct predictions of energy (a total system property of an atomistic structure that measures the chemical stability) and atomic forces (an atomic level property of an atomistic structure that measures the dynamical stability). The hyper parameters of the GFM have been tuned using scalable hyperparameter optimization (HPO) algorithms implemented in the software DeepHyper (https://github.com/deephyper/deephyper). The pre-training of each HPO trial was performed using distributed data parallelism (DDP) to scale the training across 128 compute nodes of the exascale OLCF supercomputer Frontier. Each HPO trial was trained only for 10 epochs and an early stopping was performed to avoid wasting significant computational resources on GNN architectures that were clearly underperforming. For each HPO trial, the 'omnistat' tool developed by (AMD Research - Advanced Micro Device) was used to measure the total energy consumption in kWh. The ensemble of GFMs was obtained by selecting the fifteen best performing HPO trials. Four models have been selected for their clear advantage in accuracy, and these are the GFMs with IDs 229, 156, 147, 260. Additional eleven models have been selected based on judicious balance between accuracy and energy consumption needed for training, and these are the GFMs with IDs 165, 78, 137, 1, 175, 171, 181, 67, 179, 167, 351. Each selected GFM of the ensemble was continued to cumulate a total of at most 30 epochs. In some cases, the total number of epochs actually performed was les than 30 due to two combined factors: (1) the size of the GFM (i.e., the number of model parameters to train) and (2) the total wall-clock time for which the computational resources could be allocated on OLCF-Frontier. The "Ensemble_of_models" directory contains 15 sub-directories named as follows: - gfm_0.229 - gfm_0.156 - gfm_0.147 - gfm_0.260 - gfm_0.165 - gfm_0.78 - gfm_0.137 - gfm_0.1 - gfm_0.175 - gfm_0.171 - gfm_0.181 - gfm_0.67 - gfm_0.179 - gfm_0.167 - gfm_0.351 Each one of these sub-directories refers to one of the fifteen HPO trials that have been selected to continue the pre-training with at most 30 epochs. With each sub-directory associated with a specific HPO trial, the following files can be found: - config.json: file for argument parsing to develop and train an HydraGNN architecture - gfm_0.ID_epoch_N.pk: file with model parameters for HPO ID trial after N epochs of training The ensemble of fifteen GFM architectures was used for (1) ensemble averaging to stabilize the predictions of energy and atomic forces after pre-training for post-processing analysis and (2) ensemble uncertainty quantification (UQ). The code used to develop, pre-train, and load the pre-trained models for post-processing analysis is available on the ORNL-GitHub at the following link: https://github.com/ORNL/HydraGNN/tree/Predictive_GFM_2024

36 MATERIALS SCIENCE↗

Hybrid nested sampling algorithm for Bayesian model selection applied to inverse subsurface flow problems

A Hybrid Nested Sampling (HNS) algorithm is proposed for efficient Bayesian model calibration and prior model selection. The proposed algorithm combines, Nested Sampling (NS) algorithm, Hybrid Monte Carlo (HMC) sampling and gradient estimation using Stochastic Ensemble Method (SEM). NS is an efficient sampling algorithm that can be used for Bayesian calibration and estimating the Bayesian evidence for prior model selection. Nested sampling has the advantage of computational feasibility. Within the nested sampling algorithm, a constrained sampling step is performed. For this step, we utilize HMC to reduce the correlation between successive sampled states. HMC relies on the gradient of the logarithm of the posterior distribution, which we estimate using a stochastic ensemble method based on an ensemble of directional derivatives. SEM only requires forward model runs and the simulator is then used as a black box and no adjoint code is needed. The developed HNS algorithm is successfully applied for Bayesian calibration and prior model selection of several nonlinear subsurface flow problems.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A simple grand canonical approach to compute the vapor pressure of bulk and finite size systems

In this article we introduce a simple grand canonical screening (GCS) approach to accurately compute vapor pressures from molecular dynamics or Monte Carlo simulations. This procedure entails a screening of chemical potentials using a conventional grand canonical scheme, and therefore it is straightforward to implement for any kind of interface. The scheme is validated against data obtained from Gibbs ensemble simulations for water and argon. Then, it is applied to obtain the vapor pressure of the coarse-grained mW water model, and it is shown that the computed value is in excellent accord with the one formally deduced using statistical thermodynamics arguments. Finally, this methodology is used to calculate the vapor pressure of a water nanodroplet of 94 molecules. Interestingly, the result is in perfect agreement with the one predicted by the Kelvin equation for a homogeneous droplet of that size.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Optimal Power Management of Battery Energy Storage Systems via Ensemble Kalman Inversion

Optimal power management of battery energy storage systems (BESS) is crucial for their safe and efficient operation. Numerical optimization techniques are frequently utilized to solve the optimal power management problems. However, these techniques often fall short of delivering real-time solutions for large-scale BESS due to their computational complexity. To address this issue, this paper proposes a computationally efficient approach. We introduce a new set of decision variables called power-sharing ratios corresponding to each cell, indicating their allocated power share from the output power demand. We then formulate an optimal power management problem to minimize the system-wide power losses while ensuring compliance with safety, balancing, and power supply-demand match constraints. To efficiently solve this problem, a parametrized control policy is designed and leveraged to transform the optimal power management problem into a parameter estimation problem. We then implement the ensemble Kalman inversion to estimate the optimal parameter set. The proposed approach significantly reduces computational requirements due to 1) the much lower dimensionality of the decision parameters and 2) the estimation treatment of the optimal power management problem. Finally, we conduct extensive simulations to validate the effectiveness of the proposed approach. The results show promise in accuracy and computation time compared with explored numerical optimization techniques.

Farakhor, Amir↗

Scalable Metropolis Monte Carlo for simulation of hard shapes

We design and implement a scalable hard particle Monte Carlo simulation toolkit (HPMC), and release it open source as part of HOOMD-blue. HPMC runs in parallel on many CPUs and many GPUs using domain decomposition. We employ BVH trees instead of cell lists on the CPU for fast performance, especially with large particle size disparity, and optimize inner loops with SIMD vector intrinsics on the CPU. Our GPU kernel proposes many trial moves in parallel on a checkerboard and uses a block-level queue to redistribute work among threads and avoid divergence. HPMC supports a wide variety of shape classes, including spheres/disks, unions of spheres, convex polygons, convex spheropolygons, concave polygons, ellipsoids/ellipses, convex polyhedra, convex spheropolyhedra, spheres cut by planes, and concave polyhedra. NVT and NPT ensembles can be run in 2D or 3D triclinic boxes. Additional integration schemes permit Frenkel–Ladd free energy computations and implicit depletant simulations. In a benchmark system of a fluid of 4096 pentagons, HPMC performs 10 million sweeps in 10 min on 96 CPU cores on XSEDE Comet. The same simulation would take 7.6 h in serial. HPMC also scales to large system sizes, and the same benchmark with 16.8 million particles runs in 1.4 h on 2048 GPUs on OLCF Titan.

Computer Science↗

Unraveling Adsorbate-Induced Structural Evolution of Iron Carbide Nanoparticles

Iron carbide (Fe x C y ) nanoparticles (NPs) are promising candidates for replacing platinum group metals in industrial applications, such as high-temperature Fischer–Tropsch synthesis. However, due to their amorphous nature, characterization of the active sites has been challenging experimentally and computationally. Here, using a combined density functional theory (DFT), neural network interatomic potential-assisted global optimization, and ensemble learning study, we evaluate dynamic surface changes associated with syngas (H and CO) interactions. For this purpose, we have developed a general procedure that we use to model an experimentally relevant 270-atom Fe 182 C 88 NP using the neural network-assisted stochastic surface walk global optimization algorithm (SSW-NN). Once generated, the Fe 182 C 88 NP active sites and particle morphology are thoroughly characterized before the effects of syngas adsorbate interactions are explored by using DFT and molecular dynamics simulations. Lastly, we explore correlations between geometric and electronic features of the active sites and the adsorption of H (H ads ), using a regularized random forest machine learning algorithm. In doing so, we identified the Fe–C coordination number and p orbital occupancy as the most important descriptors affecting H ads . Furthermore, using a combined ML and quantum chemistry approach, our work demonstrates a general and efficient procedure for generating and probing complex surface phenomena on binary nanoparticles.

Adsorption↗

Mapping Composition Evolution through Synthesis, Purification, and Depolymerization of Random Heteropolymers

Random heteropolymers (RHPs) consisting of three or more comonomers have been routinely used to synthesize functional materials. While increasing the monomer variety diversifies the side-chain chemistry, this substantially expands the sequence space and leads to ensemble-level sequence heterogeneity. Most studies have relied on monomer composition and simulated sequences to design RHPs, but the questions remain unanswered regarding heterogeneities within each RHP ensemble and how closely these simulated sequences reflect the experimental outcomes. Here, we quantitatively mapped out the evolution of monomer compositions in four-monomer-based RHPs throughout a design-synthesis-purification-depolymerization process. By adopting a Jaacks method, we first determined 12 reactivity ratios directly from quaternary methacrylate RAFT copolymerization experiments to account for the influences of competitive monomer addition and the reversible activation/deactivation equilibria. The reliability of in silico analysis was affirmed by a quantitative agreement (<4% difference) between the simulated RHP compositions and the experimental results. Furthermore, we mapped out the conformation distribution within each ensemble in different solvents as a function of monomer chemistry, composition, and segmental characteristics via high-throughput computation based on self-consistent field theory (SCFT). These comprehensive studies confirmed monomer composition as a viable design parameter to engineer RHP-based functional materials as long as the reactivity ratios are accurately determined and the livingness of RHP synthesis is ensured.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Optimizing a Thomson scattering diagnostic for fast dynamics and high background

The Madison Symmetric Torus (MST) presents challenging conditions for Thomson scattering (TS) measurements. The MST plasmas are reversed-field pinches (RFPs) with electron density n{sub e}<3x10{sup 13} cm{sup -3}, typically 1x10{sup 13} cm{sup -3}. The TS system was designed to measure from 10 eV to 2 keV; however, six polychromators were upgraded from four to eight spectral channels to resolve to 10 keV. There is no diverter or vertical field, so wall interaction results in high background light both from ion and neutral bremsstrahlungs and from line radiation. Also during standard plasmas, the RFP exhibits regular reconnection sawteeth events during which the plasma current, density, and temperature profiles are flattened. These events are of interest both due to the reconnection physics and to their contribution to the MST equilibrium and confinement. These events occur over 100 {mu}S and exhibit large changes in background light and fast changes in temperature. During improved confinement plasmas, there are no sawteeth; the background is low but the temperature can be over an order of magnitude higher. Data analysis of the system has been developed to accommodate both the large dynamic range of the temperature, the fast dynamics, and the fast changing, high amplitude background. Special attention has been paid to the sources of error, in particular, the contribution of the background. A response-function method reduces the measured uncertainty by a factor of 2. Numerical techniques have been developed which are extremely robust. Two methods are used, a conventional {chi}{sup 2} minimization using a Levenberg-Marquardt algorithm coupled with Monte Carlo modeling for the error bar and a Bayesian statistics method. The Bayesian method computes the probability distribution for the measured photons and electron temperature and this information can be used to ensemble data and will allow future integrated data analysis efforts.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

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

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

74 ATOMIC AND MOLECULAR PHYSICS↗

Hierarchy of N-point functions in the {Lambda}CDM and ReBEL cosmologies

In this work we investigate higher-order statistics for the {Lambda}CDM and ReBEL scalar-interacting dark matter models by analyzing 180h{sup -1} Mpc dark matter N-body simulation ensembles. The N-point correlation functions and the related hierarchical amplitudes, such as skewness and kurtosis, are computed using the counts-in-cells method. Our studies demonstrate that the hierarchical amplitudes S{sub n} of the scalar-interacting dark matter model significantly deviate from the values in the {Lambda}CDM cosmology on scales comparable and smaller than the screening length r{sub s} of a given scalar-interacting model. The corresponding additional forces that enhance the total attractive force exerted on dark matter particles at galaxy scales lower the values of the hierarchical amplitudes S{sub n}. We conclude that hypothetical additional exotic interactions in the dark matter sector should leave detectable markers in the higher-order correlation statistics of the density field. We focused in detail on the redshift evolution of the dark matter field's skewness and kurtosis. From this investigation we find that the deviations from the canonical {Lambda}CDM model introduced by the presence of the 'fifth' force attain a maximum value at redshifts 0.5<z<2. We therefore conclude that moderate redshift data are better suited for setting observational constraints on the investigated ReBEL models.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗