Search NASA⌕ Search

SEARCH · Search NASA

Results for “Approximate 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 1,441 records · Page 80
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Collective neutrino oscillations on a quantum computer with hybrid quantum-classical algorithm

We simulate the time evolution of collective neutrino oscillations in two-flavor settings on a quantum computer. We explore the generalization of Trotter-Suzuki approximation to time-dependent Hamiltonian dynamics. The trotterization steps are further optimized using the Cartan decomposition of two-qubit unitary gates U ϵ SU(4) in the minimum number of controlled-NOT (CNOT) gates making the algorithm more resilient to the hardware noise. As a result, a more efficient hybrid quantum-classical algorithm is also explored to solve the problem on noisy intermediate-scale quantum devices.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2022

CFD-trained ANN Model for Approximating Near-occupant Condition in Real-time Simulations

The main drawback of Computational Fluid Dynamics (CFD) simulations has been the time and resource consuming nature which is not suitable for real-time applications. In this work, we first generated numerous CFD models of a given indoor space to obtain airspeed, temperature, and mean radiant temperature near an occupant as training data. Several artificial neural networks (ANN) models were trained using this CFD simulated data to approximate near real-time environmental conditions for a given occupant. This trained ANN model approach is a part of a real-time simulation of building operations using a combination of software and real hardware (HVAC equipment) approaches. The preliminary results suggest that the CFD- generated training data and the trained ANN model can accurately approximate such conditions in a real-time application, a method that has great potential in building simulation and building digital twin areas of research.

Zhang, Yun↗
DOE OSTI2023

Data assimilation in operator algebras

We develop an algebraic framework for sequential data assimilation of partially observed dynamical systems. In this framework, Bayesian data assimilation is embedded in a nonabelian operator algebra, which provides a representation of observables by multiplication operators and probability densities by density operators (quantum states). In the algebraic approach, the forecast step of data assimilation is represented by a quantum operation induced by the Koopman operator of the dynamical system. Moreover, the analysis step is described by a quantum effect, which generalizes the Bayesian observational update rule. Projecting this formulation to finite-dimensional matrix algebras leads to computational schemes that are i) automatically positivity-preserving and ii) amenable to consistent data-driven approximation using kernel methods for machine learning. Moreover, these methods are natural candidates for implementation on quantum computers. Applications to the Lorenz 96 multiscale system and the El Niño Southern Oscillation in a climate model show promising results in terms of forecast skill and uncertainty quantification.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Rapid and Robust Cross-Correlation-Based Seismic Signal Identification Using an Approximate Nearest Neighbor Method

The matched filtering technique that uses the cross correlation of a waveform of interest with archived signals from a template library has proven to be a powerful tool for detecting events in regions with repeating seismicity. However, waveform correlation is computationally expensive and therefore impractical for large template sets unless dedicated distributed computing hardware and software are used. In this paper, we introduce an approximate nearest neighbor (ANN) approach that enables the use of very large template libraries for waveform correlation. Our method begins with a projection into a reduced dimensionality space, based on correlation with a randomized subset of the full template archive. Searching for a specified number of nearest neighbors for a query waveform is accomplished by iteratively comparing it with the neighbors of its immediate neighbors. We used the approach to search for matches to each of ~2300 analyst-reviewed signal detections reported in May 2010 for the International Monitoring System station MKAR. The template library in this case consists of a data set of more than 200,000 analyst-reviewed signal detections for the same station from February 2002 to July 2016 (excluding May 2010). Of these signal detections, 73% are teleseismic first P and 17% regional phases (Pn, Pg, Sn, and Lg). Finally, the analyses performed on a standard desktop computer show that the proposed ANN approach performs a search of the large template libraries about 25 times faster than the standard full linear search and achieves recall rates greater than 80%, with the recall rate increasing for higher correlation thresholds.

58 GEOSCIENCES↗
DOE OSTI2014

Towards a unified picture of the water self-ions at the air-water interface: a density functional theory perspective

The propensities of the water self ions, H3O+ and OH- , for the air-water interface has implications for interfacial acid-base chemistry. Despite numerous experimental and computational studies, no consensus has been reached on the question of whether or not H3O+ and/or OH- prefer to be at the water surface or in the bulk. Here we report a molecular dynamics simulation study of the bulk vs. interfacial behavior of H3O+ and OH- that employs forces derived from density functional theory with a generalized gradient approximation exchangecorrelation functional (specifically, BLYP) and empirical dispersion corrections. We computed the potential of mean force (PMF) for H3O+ as a function of the position of the ion in a 215-molecule water slab. The PMF is flat, suggesting that H3O+ has equal propensity for the air-water interface and the bulk. We compare the PMF for H3O+ to our previously computed PMF for OH- adsorption, which contains a shallow minimum at the interface, and we explore how differences in solvation of each ion at the interface vs. the bulk are connected with interfacial propensity. We find that the solvation shell of H3O+ is only slightly dependent on its position in the water slab, while OH- partially desolvates as it approaches the interface, and we examine how this difference in solvation behavior is manifested in the electronic structure and chemistry of the two ions. DJT was supported by National Science Foundation grant CHE-0909227. CJM was supported by the U.S. Department of Energy‘s (DOE) Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences and Biosciences. Pacific Northwest National Laboratory (PNNL) is operated for the Department of Energy by Battelle. The potential of mean force required resources of the Oak Ridge Leadership Computing Facility at the Oak Ridge National Laboratory, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DEAC05-00OR22725. The remaining simulations and analysis used resources of the National Energy Research Scientific Computing Center, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231. at at Lawrence Berkeley National Laboratory. MDB is grateful for the support of the Linus Pauling Distinguished Postdoctoral Fellowship Program at PNNL.

Baer, Marcel D.↗
DOE OSTI2018

Equilibration and analysis of first-principles molecular dynamics simulations of water

First-principles molecular dynamics (FPMD) simulations based on density functional theory are becoming increasingly popular for the description of liquids. In view of the high computational cost of these simulations, the choice of an appropriate equilibration protocol is critical. In this study, we assess two methods of estimation of equilibration times using a large dataset of first-principles molecular dynamics simulations of water. The Gelman-Rubin potential scale reduction factor [A. Gelman and D. B. Rubin, Stat. Sci. 7, 457 (1992)] and the marginal standard error rule heuristic proposed by White [Simulation 69, 323 (1997)] are evaluated on a set of 32 independent 64-molecule simulations of 58 ps each, amounting to a combined cumulative time of 1.85 ns. The availability of multiple independent simulations also allows for an estimation of the variance of averaged quantities, both within MD runs and between runs. We analyze atomic trajectories, focusing on correlations of the Kohn-Sham energy, pair correlation functions, number of hydrogen bonds, and diffusion coefficient. The observed variability across samples provides a measure of the uncertainty associated with these quantities, thus facilitating meaningful comparisons of different approximations used in the simulations. We find that the computed diffusion coefficient and average number of hydrogen bonds are affected by a significant uncertainty in spite of the large size of the dataset used. Lastly, a comparison with classical simulations using the TIP4P/2005 model confirms that the variability of the diffusivity is also observed after long equilibration times. Complete atomic trajectories and simulation output files are available online for further analysis.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2014

Dynamic load balance scheme for the DSMC algorithm

The direct simulation Monte Carlo (DSMC) algorithm, devised by Bird, has been used over a wide range of various rarified flow problems in the past 40 years. While the DSMC is suitable for the parallel implementation on powerful multi-processor architecture, it also introduces a large load imbalance across the processor array, even for small examples. The load imposed on a processor by a DSMC calculation is determined to a large extent by the total of simulator particles upon it. Since most flows are impulsively started with initial distribution of particles which is surely quite different from the steady state, the total of simulator particles will change dramatically. The load balance based upon an initial distribution of particles will break down as the steady state of flow is reached. The load imbalance and huge computational cost of DSMC has limited its application to rarefied or simple transitional flows. In this paper, by taking advantage of METIS, a software for partitioning unstructured graphs, and taking the total of simulator particles in each cell as a weight information, the repartitioning based upon the principle that each processor handles approximately the equal total of simulator particles has been achieved. The computation must pause several times to renew the total of simulator particles in each processor and repartition the whole domain again. Thus the load balance across the processors array holds in the duration of computation. The parallel efficiency can be improved effectively. The benchmark solution of a cylinder submerged in hypersonic flow has been simulated numerically. Besides, hypersonic flow past around a complex wing-body configuration has also been simulated. The results have displayed that, for both of cases, the computational time can be reduced by about 50%.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2019

Surrogate Model Selection for Design Space Approximation And Surrogatebased Optimization

Surrogate models are used to map input data to output data when the actual relationship between the two is unknown or computationally expensive to evaluate for sensitivity analysis, uncertainty propagation and surrogate based optimization. This work evaluates the performance of eight surrogate modeling techniques for design space approximation and surrogate based optimization applications over a set of generated datasets with known characteristics. With this work, we aim to provide general rules for selecting an appropriate surrogate model form solely based on the characteristics of the data being modeled. Furthermore, the computational experiments revealed that, in general, multivariate adaptive regression spline models (MARS) and single hidden layer feed forward neural networks (ANN) yielded the most accurate predictions over the design space while Random Forest (RF) models most reliably identified the locations of the optimums when used for surrogate-based optimization.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Multi-fidelity Uncertainty Quantification for Homogenization Problems in Structure-Property Relationships from Crystal Plasticity Finite Elements

Crystal plasticity finite element method (CPFEM) has been an integrated computational materials engineering (ICME) workhorse to study materials behaviors and structure-property relationships for the last few decades. These relations are mappings from the microstructure space to the materials properties space. Due to the stochastic and random nature of microstructures, there is always some uncertainty associated with materials properties, for example, in homogenized stress-strain curves. For critical applications with strong reliability needs, it is often desirable to quantify the microstructure-induced uncertainty in the context of structure-property relationships. However, this uncertainty quantification (UQ) problem often incurs a large computational cost because many statistically equivalent representative volume elements (SERVEs) are needed. In this article, we apply a multi-level Monte Carlo (MLMC) method to CPFEM to study the uncertainty in stress-strain curves, given an ensemble of SERVEs at multiple mesh resolutions. By using the information at coarse meshes, we show that it is possible to approximate the response at fine meshes with a much reduced computational cost. We focus on problems where the model output is multi-dimensional, which requires us to track multiple quantities of interest (QoIs) at the same time. In conclusion, our numerical results show that MLMC can accelerate UQ tasks around 2.23x, compared to the classical Monte Carlo (MC) method, which is widely known as ensemble average in the CPFEM literature.

36 MATERIALS SCIENCE↗
DOE OSTI2015

On the electronic, structural, and thermodynamic properties of Au supported on α-Fe{sub 2}O{sub 3} surfaces and their interaction with CO

Extensive first principles calculations are carried out to investigate Au monomers and dimers supported on α-Fe{sub 2}O{sub 3}(0001) surfaces in terms of structure optimizations, electronic structure analyses, and ab initio thermodynamics calculations of surface phase diagrams. All computations rely on density functional theory in the generalized gradient approximation (Perdew-Burke-Ernzerhof (PBE)) and account for on-site Coulomb interactions via inclusion of a Hubbard correction (PBE+U). The relative stability of Au monomers/dimers on the stoichiometric termination of α-Fe{sub 2}O{sub 3}(0001) decorated with various vacancies (multiple oxygen vacancies, iron vacancy, and mixed iron-oxygen vacancies) has been computed as a function of the oxygen chemical potential. The charge rearrangement induced by Au at the oxide contact is analyzed in detail and discussed. On one hand, ab initio thermodynamics predicts that under O-rich conditions, structures obtained by replacing a surface Fe atom with a Au atom are thermodynamically stable over a wide range of temperatures. On the other hand, the complex of a CO molecule on a Au atom substituting surface Fe atoms is thermodynamically stable only in a much more narrow range of values of the O chemical potential under O-rich conditions. In the case of a Au dimer, under O-rich conditions, supported Au atoms at an O-Fe di-vacancy are more stable. However, upon CO adsorption, the complex of a CO molecule and 2 Au atoms located at a single Fe vacancy is more favorable.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

Offline Maximizing Minimally Invasive Proper Orthogonal Decomposition for Reduced-Order Modeling of S n Radiation Transport

Deterministic solutions to the Sn radiation transport equation can be computationally expensive to calculate. Reduced-order modeling enables efficient approximation of the full-order model (FOM) solution. We propose a novel method for constructing reduced-order models (ROMs) of the S n radiation transport equation, offline maximizing minimally invasive (OMMI) proper orthogonal decomposition (POD). POD uses the method of snapshots to create a reduced-order basis for constructing an ROM. Minimally invasive POD leverages the sweep infrastructure existing in deterministic transport codes to create a POD-based ROM, even when infeasible by traditional methods. Offline maximizing minimally invasive proper orthogonal decomposition (OMMI-POD) extends minimally invasive POD by performing sweeps offline, therefore maximizing the potential speedup. OMMI-POD does so by creating a library of reduced systems from a training set. This library of reduced systems is then interpolated to provide a rapid approximate solution of the S n radiation transport equation. The model is evaluated on a set of test problems, achieving a low error with a 466 times speedup over the FOM. Also presented is a study of the effect of sampling method on the performance of OMMI-POD, specifically comparing naive uniform sampling to the more accurate and computationally expensive greedy sampling.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Many-Body Basis Set Amelioration Method for Incremental Full Configuration Interaction

Incremental full configuration interaction (iFCI) is a polynomial-cost electronic structure method that systematically approaches the FCI limit by employing the method of increments to solve the Schrödinger equation through a many-body expansion. This article introduces the many-body basis set amelioration (MBBSA) method, which is designed to allow iFCI to be applicable to larger atomic orbital basis sets. MBBSA uses a series of inexpensive iFCI calculations to approximate the correlation energy that would be found using a more expensive, highly accurate iFCI calculation. Here, when compared to standard iFCI computations on smaller molecules in triple-zeta and larger basis sets, MBBSA provides approximations to the total and relative energies within chemical accuracy. MBBSA exhibits a reduced cost of between 60-92% when compared to standard iFCI calculations, with larger systems experiencing the largest benefit. Tests of MBBSA on two reactions that involve highly correlated systems, the automerization of cyclobutadiene and a Criegee intermediate reaction, show that MBBSA has practical utility for studying realistic chemistries.

Basis sets↗
DOE OSTI2015

The Initial State of a Primordial Anisotropic Stage of Inflation

We investigate the possibility that the inflationary period in the early universe was preceded by a primordial stage of strong anisotropy. In particular we focus on the simplest model of this kind, where the spacetime is described by a non-singular Kasner solution that quickly evolves into an isotropic de Sitter space, the so-called Kasner-de Sitter solution. The initial Big Bang singularity is replaced, in this case, by a horizon. We show that the extension of this metric to the region behind the horizon contains a timelike singularity which will be visible by cosmological observers. This makes it impossible to have a reliable prediction of the quantum state of the cosmological perturbations in the region of interest. In this paper we consider the possibility that this Kasner-de Sitter universe is obtained as a result of a quantum tunneling process effectively substituting the region behind the horizon by an anisotropic parent vacuum state, namely a 1+1 dimensional spacetime compactified over an internal flat torus, T{sub 2}, which we take it to be of the form de Sitter{sub 2}×T{sub 2} or Minkowski{sub 2}×T{sub 2}. As a first approximation to understand the effects of this anisotropic initial state, we compute the power spectrum of a massless scalar field in these backgrounds. In both cases, the spectrum converges at small scales to the isotropic scale invariant form and only present important deviations from it at the largest possible scales. We find that the decompactification scenario from M{sub 2}×T{sub 2} leads to a suppressed and slightly anisotropic power spectrum at large scales which could be related to some of the anomalies present in the current CMB data. On the other hand, the spectrum of the universe with a dS{sub 2}×T{sub 2} parent vacuum presents an enhancement in power at large scales not consistent with observations.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2016

An adaptive Gaussian process-based method for efficient Bayesian experimental design in groundwater contaminant source identification problems: ADAPTIVE GAUSSIAN PROCESS-BASED INVERSION

Surrogate models are commonly used in Bayesian approaches such as Markov Chain Monte Carlo (MCMC) to avoid repetitive CPU-demanding model evaluations. However, the approximation error of a surrogate may lead to biased estimations of the posterior distribution. This bias can be corrected by constructing a very accurate surrogate or implementing MCMC in a two-stage manner. Since the two-stage MCMC requires extra original model evaluations, the computational cost is still high. If the information of measurement is incorporated, a locally accurate approximation of the original model can be adaptively constructed with low computational cost. Based on this idea, we propose a Gaussian process (GP) surrogate-based Bayesian experimental design and parameter estimation approach for groundwater contaminant source identification problems. A major advantage of the GP surrogate is that it provides a convenient estimation of the approximation error, which can be incorporated in the Bayesian formula to avoid over-confident estimation of the posterior distribution. The proposed approach is tested with a numerical case study. Without sacrificing the estimation accuracy, the new approach achieves about 200 times of speed-up compared to our previous work using two-stage MCMC.

Zhang, Jiangjiang↗
DOE OSTI2014

Sphere-plate Casimir interaction in (D + 1)-dimensional spacetime

In this paper, we derive the formula for the Casimir interaction energy between a sphere and a plate in (D + 1)-dimensional Minkowski spacetime. It is assumed that the scalar field satisfies the Dirichlet or Neumann boundary conditions on the sphere and the plate. As in the D = 3 case, the formula is of TGTG type. One of our main contributions is deriving the translation matrices which express the change of bases between plane waves and spherical waves for general D. Using orthogonality of Gegenbauer polynomials, it turns out that the final TGTG formula for the Casimir interaction energy can be simplified to one that is similar to the D = 3 case. To illustrate the application of the formula, both large separation and small separation asymptotic behaviors of the Casimir interaction energy are computed. The large separation leading term is proportional to L{sup −D+1} if the sphere is imposed with Dirichlet boundary condition, and to L{sup −D−1} if the sphere is imposed with Neumann boundary condition, where L is distance from the center of the sphere to the plane. For the small separation asymptotic behavior, it is shown that the leading term is equal to the one obtained using proximity force approximation. The next-to-leading order term is also computed using perturbation method. It is shown that when the space dimension D is larger than 5, the next-to-leading order has sign opposite to the leading order term. Moreover, the ratio of the next-to-leading order term to the leading order term is linear in D, indicating a larger correction at higher dimensions.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2017

Analytic derivative couplings and first-principles exciton/phonon coupling constants for an ab initio Frenkel-Davydov exciton model: Theory, implementation, and application to compute triplet exciton mobility parameters for crystalline tetracene

Recently, we introduced an ab initio version of the Frenkel-Davydov exciton model for computing excited-state properties of molecular crystals and aggregates. Within this model, supersystem excited states are approximated as linear combinations of excitations localized on molecular sites, and the electronic Hamiltonian is constructed and diagonalized in a direct-product basis of non-orthogonal configuration state functions computed for isolated fragments. In this work, we derive and implement analytic derivative couplings for this model, including nuclear derivatives of the natural transition orbital and symmetric orthogonalization transformations that are part of the approximation. Nuclear derivatives of the exciton Hamiltonian’s matrix elements, required in order to compute the nonadiabatic couplings, are equivalent to the “Holstein” and “Peierls” exciton/phonon couplings that are widely discussed in the context of model Hamiltonians for energy and charge transport in organic photovoltaics. As an example, we compute the couplings that modulate triplet exciton transport in crystalline tetracene, which is relevant in the context of carrier diffusion following singlet exciton fission.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2018

Stimulus Sensitivity of a Spiking Neural Network Model

Some recent papers relate the criticality of complex systems to their maximal capacity of information processing. In the present paper, we consider high dimensional point processes, known as age-dependent Hawkes processes, which have been used to model spiking neural networks. Using mean-field approximation, the response of the network to a stimulus is computed and we provide a notion of stimulus sensitivity. It appears that the maximal sensitivity is achieved in the sub-critical regime, yet almost critical for a range of biologically relevant parameters.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2026

Deep Neural Networks are Adaptive to Function Regularity and Data Distribution in Approximation and Estimation

Deep learning has exhibited remarkable results across diverse areas. To understand its success, substantial research has been directed towards its theoretical foundations. Nev- ertheless, the majority of these studies examine how well deep neural networks can model functions with uniform regularities. In this paper, we explore a different angle: how deep neural networks can adapt to varying degrees of smoothness in functions and nonuni- form data distributions across different locations and scales. More precisely, we focus on a broad class of functions defined by nonlinear tree-based approximation methods. This class encompasses a range of function types, such as functions with uniform regularities and discontinuous functions. We develop nonparametric approximation and estimation theories for this class using deep ReLU networks. Our results show that deep neural networks are adaptive to the nonuniform smoothness of functions and nonuniform data distributions at different locations and scales. We apply our results to several function classes, and derive the corresponding approximation and generalization errors. The validity of our results is demonstrated through numerical experiments.

97 MATHEMATICS AND COMPUTING↗