SEARCH · Engineering Papers

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,693 records · Page 94

All sources NASA NTRS DOE OSTI

NASA NTRS2006

Comet Mineralogy as Inferred from Infrared Spectra of Comets

For most comets, infrared (IR) spectroscopy (remote sensing) is the method through which we diagnose the mineralogy and size distribution of dust in their comae. The shape and contrast of the IR spectral features depend on the particle size: optically active minerals (absorbing of visible and near-IR solar photons) and submicron solid grains or highly porous (> 90% vacuum) grains primarily contribute to the shapes of the observed resonances. Comet mineralogies typically are determined by fitting thermal emission models of ensembles of discrete mineral grains to observed IR spectral energy distributions. The absorptivities (Q-abs) and scattering efficiencies (Q-scat) of the discrete mineral grains are computed using Mie scattering, Maxwell-Garnet mixing, Discrete Dipole Approximation, and Multi-Layered Sphere codes. These techniques when applied to crystalline minerals, specifically olivine (Mg_x, Fe_1-x)2 Si04, x>0.9, require the use of ellipsoidal shaped particles with elongated axial ratios or hollow spheres to produce the shapes of the resonances observed both from comet comae and laboratory samples. The wavelength positions of the distinct resonances from submicron-radii crystalline silicates, as well as their thermal equilibrium temperatures, constrain the crystalline olivine to have a relatively high Mg-content (x>0.9, or Fo>90). Only resonances computed for submicron Mg-rich crystalline olivine and crystalline orthopyroxene match the observed IR spectral features. However, this has led to the interpretation that micron-radii and larger crystals are absent from comet comae. Furthermore, the mass fraction of silicate crystals is dependent upon whether just the submicron portion of the size distribution is being compared or the submicron crystals compare to the aggregates of porous amorphous silicates that are computationally tractable as porous spheres. We will discuss the Deep Impact results as examples of these challenges to interpreting mid-IR spectra of short-period comets.

Wooden, Diane H.↗

NASA NTRS1988

Finite-dimensional approximation for optimal fixed-order compensation of distributed parameter systems

In controlling distributed parameter systems it is often desirable to obtain low-order, finite-dimensional controllers in order to minimize real-time computational requirements. Standard approaches to this problem employ model/controller reduction techniques in conjunction with LQG theory. In this paper we consider the finite-dimensional approximation of the infinite-dimensional Bernstein/Hyland optimal projection theory. This approach yields fixed-finite-order controllers which are optimal with respect to high-order, approximating, finite-dimensional plant models. The technique is illustrated by computing a sequence of first-order controllers for one-dimensional, single-input/single-output, parabolic (heat/diffusion) and hereditary systems using spline-based, Ritz-Galerkin, finite element approximation. Numerical studies indicate convergence of the feedback gains with less than 2 percent performance degradation over full-order LQG controllers for the parabolic system and 10 percent degradation for the hereditary system.

Bernstein, Dennis S.↗

DOE OSTI2022

An approximate inertial manifold (AIM) based closure for turbulent flows

A closure model for turbulent flows is developed based on a dynamical system theory. An appropriately discretized formulation of the governing equations is considered for this process. The key ingredient is an approximation of the system’s attractor, where all the trajectories in phase space are confined. This approximate inertial manifold based approach provides a path to track trajectories of the system in a lower-dimensional subspace. Unlike conventional coarse-graining approaches, the turbulent field is decomposed into resolved and unresolved dynamics using the properties of the governing equations. The novelty of the approach relies on the reconstruction of the unresolved field constrained by the governing equations. A posteriori tests for homogeneous isotropic turbulence and the Kuramoto–Sivashinsky equation show promising results for considerable dimension reduction with strong convergence properties. The proposed model outperforms the dynamic Smagorinsky model, and the computational overhead is competitive with similar approaches.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2023

Tensor Decompositions for Count Data that Leverage Stochastic and Deterministic Optimization

There is growing interest to extend low-rank matrix decompositions to multi-way arrays, or tensors. One fundamental low-rank tensor decomposition is the canonical polyadic decomposition (CPD). The challenge of fitting a low-rank, nonnegative CPD model to Poisson-distributed count data is of particular interest. Several popular algorithms use local search methods to approximate the global maximum likelihood estimator from local minima. Simultaneously, a recent trend in theoretical computer science and numerical linear algebra leverages randomization to solve very large, hard problems. The typical approach is to use randomization for a fast approximation and determinism for refinement to yield effective algorithms with theoretical guarantees. Two popular algorithms for Poisson CPD reflect that emergent dichotomy: CP Alternating Poisson Regression is a deterministic algorithm and Generalized Canonical Polyadic decomposition makes use of stochastic algorithms in several variants. This work extends recent work to develop two new methods that leverage randomized and deterministic algorithms for improved accuracy and performance.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1990

Efficient eigensolution reanalysis of nonclassically damped structures

Effective methods of approximate eigensolution reanalysis of modified nonclassically damped structures are developed. For structures with passive or active discrete damping devices or with damping treatment, the system becomes nonproportionally damped and the computation of its dynamic responses may require the use of complex modes. For larger systems, the computation of complex modes is very expensive. Thus it is desirable to have approximate reanalysis techniques for the efficient evaluation of the effect of design changes. In recent years, the assumed mode reanalysis method was successfully applied to minimum weight design of undamped structures with natural frequency constraints. The accuracy of the assumed mode reanalysis method can be improved dramatically if the global approximation function includes the normal modes of the original system and their derivatives. This approach was demonstrated to be effective even for a system with shape changes. The approach used by Noor et al. for eigensolution reanalysis of undamped structures is extended to treat a nonclassically damped system.

Wang, Bo Ping↗

NASA NTRS1992

A statistical model for evaluating GOPEX uplink performance

This article describes a statistical model to analyze the signal intensity received at the solid-state imaging (SSI) camera of the Galileo optical communications system from an Earth-based transmitter (GOPEX) demonstration. The analytical model assumes that the optical beam possesses a Gaussian profile and the communication channel has a log-normal scattering characteristic. The atmospheric-induced jitter is modelled as two independent zero mean Gaussian random variables. By modelling the system parameters as a set of independent and identically distributed (iid) random variables, the combined impact of uncertainties due to system parameters and the turbulent atmosphere can be approximated by a log-normal distributed signal intensity at the spacecraft. A Monte-Carlo software simulation package was also developed to compute the confidence interval probabilities for general optical beam profiles. Numerical results show that the approximation is valid for a wide range of operation scenarios.

Kiesaleh, K.↗

DOE OSTI2022

Phoebe: a high-performance framework for solving phonon and electron Boltzmann transport equations

Understanding the electrical and thermal transport properties of materials is critical to the design of electronics, sensors, and energy conversion devices. Computational modeling can accurately predict material properties but, in order to be reliable, requires accurate descriptions of electron and phonon states and their interactions. While first-principles methods are capable of describing the energy spectrum of each carrier, using them to compute transport properties is still a formidable task, both computationally demanding and memory intensive, requiring integration of fine microscopic scattering details for estimation of macroscopic transport properties. To address this challenge, we present Phoebe—a newly developed software package that includes the effects of electron–phonon, phonon–phonon, boundary, and isotope scattering in computations of electrical and thermal transport properties of materials with a variety of available methods and approximations. This open source C++ code combines MPI-OpenMP hybrid parallelization with GPU acceleration and distributed memory structures to manage computational cost, allowing Phoebe to effectively take advantage of contemporary computing infrastructures. We demonstrate that Phoebe accurately and efficiently predicts a wide range of transport properties, opening avenues for accelerated computational analysis of complex crystals.

36 MATERIALS SCIENCE↗

DOE OSTI2024

Parallel Randomized Tucker Decomposition Algorithms

The Tucker tensor decomposition is a natural extension of the singular value decomposition (SVD) to multiway data. Here, we propose to accelerate Tucker tensor decomposition algorithms by using randomization and parallelization. We present two algorithms that scale to large data and many processors, significantly reduce both computation and communication cost compared to previous deterministic and randomized approaches, and obtain nearly the same approximation errors. The key idea in our algorithms is to perform randomized sketches with Kronecker-structured random matrices, which reduces computation compared to unstructured matrices and can be implemented using a fundamental tensor computational kernel. We provide probabilistic error analysis of our algorithms and implement a new parallel algorithm for the structured randomized sketch. Our experimental results demonstrate that our combination of randomization and parallelization achieves accurate Tucker decompositions much faster than alternative approaches. We observe up to a 16X speedup over the fastest deterministic parallel implementation on 3D simulation data.

Tucker decompositions↗

NASA NTRS1989

Unsteady Navier-Stokes computations over airfoils using both fixed and dynamic meshes

A finite volume implicit approximate factorization method which solves the thin layer Navier-Stokes equations was used to predict unsteady turbulent flow airfoil behavior. At a constant angle of attack of 16 deg, the NACA 0012 airfoil exhibits an unsteady periodic flow field with the lift coefficient oscillating between 0.89 and 1.60. The Strouhal number is 0.028. Results are similar at 18 deg, with a Strouhal number of 0.033. A leading edge vortex is shed periodically near maximum lift. Dynamic mesh solutions for unstalled airfoil flows show general agreement with experimental pressure coefficients. However, moment coefficients and the maximum lift value are underpredicted. The deep stall case shows some agreement with experiment for increasing angle of attack, but is only qualitatively comparable past stall and for decreasing angle of attack.

Rumsey, Christopher L.↗

NASA NTRS2008

Improved Linear Algebra Methods for Redshift Computation from Limited Spectrum Data - II

Given photometric broadband measurements of a galaxy, Gaussian processes may be used with a training set to solve the regression problem of approximating the redshift of this galaxy. However, in practice solving the traditional Gaussian processes equation is too slow and requires too much memory. We employed several methods to avoid this difficulty using algebraic manipulation and low-rank approximation, and were able to quickly approximate the redshifts in our testing data within 17 percent of the known true values using limited computational resources. The accuracy of one method, the V Formulation, is comparable to the accuracy of the best methods currently used for this problem.

Foster, Leslie↗

NASA NTRS1995

Numerical solution of differential equations by artificial neural networks

Conventionally programmed digital computers can process numbers with great speed and precision, but do not easily recognize patterns or imprecise or contradictory data. Instead of being programmed in the conventional sense, artificial neural networks (ANN's) are capable of self-learning through exposure to repeated examples. However, the training of an ANN can be a time consuming and unpredictable process. A general method is being developed by the author to mate the adaptability of the ANN with the speed and precision of the digital computer. This method has been successful in building feedforward networks that can approximate functions and their partial derivatives from examples in a single iteration. The general method also allows the formation of feedforward networks that can approximate the solution to nonlinear ordinary and partial differential equations to desired accuracy without the need of examples. It is believed that continued research will produce artificial neural networks that can be used with confidence in practical scientific computing and engineering applications.

Meade, Andrew J., Jr.↗

DOE OSTI2021

Solving inverse problems in stochastic models using deep neural networks and adversarial training

Inverse problems associated with stochastic models constitute a significant portion of scientific and engineering applications. In such cases the unknown quantities are distributions. The applicability of traditional methods is limited because of their demanding assumptions or prohibitive computational consumption; for example, maximum likelihood methods require closed-form density functions, and Markov Chain Monte Carlo needs a large number of simulations. We propose a new method that estimates the unknown distribution by matching the statistical properties between observed and simulated random processes. We leverage the expressive power of neural networks to approximate the unknown distribution and use a discriminative neural network for computing the statistical discrepancies between the observed and simulated random processes. Here we demonstrated numerically that the proposed methods can estimate both the model parameters and learn complicated unknown distributions.

42 ENGINEERING↗

NASA NTRS1974

The passage of an infinite swept airfoil through an oblique gust

An approximate solution is reported for the unsteady aerodynamic response of an infinite swept wing encountering a vertical oblique gust in a compressible stream. The approximate expressions are of closed form and do not require excessive computer storage or computation time, and further, they are in good agreement with the results of exact theory. This analysis is used to predict the unsteady aerodynamic response of a helicopter rotor blade encountering the trailing vortex from a previous blade. Significant effects of three dimensionality and compressibility are evident in the results obtained. In addition, an approximate solution for the unsteady aerodynamic forces associated with the pitching or plunging motion of a two dimensional airfoil in a subsonic stream is presented. The mathematical form of this solution approaches the incompressible solution as the Mach number vanishes, the linear transonic solution as the Mach number approaches one, and the solution predicted by piston theory as the reduced frequency becomes large.

Adamczyk, J. L.↗

NASA NTRS1997

Parameterizing Grid-Averaged Longwave Fluxes for Inhomogeneous Marine Boundary Layer Clouds

This paper examines the relative impacts on grid-averaged longwave flux transmittance (emittance) for Marine Boundary Layer (MBL) cloud fields arising from horizontal variability of optical depth tau and cloud sides, First, using fields of Landsat-inferred tau and a Monte Carlo photon transport algorithm, it is demonstrated that mean all-sky transmittances for 3D variable MBL clouds can be computed accurately by the conventional method of linearly weighting clear and cloudy transmittances by their respective sky fractions. Then, the approximations of decoupling cloud and radiative properties and assuming independent columns are shown to be adequate for computation of mean flux transmittance. Since real clouds have nonzero geometric thicknesses, cloud fractions A'(sub c) presented to isotropic beams usually exceed the more familiar vertically projected cloud fractions A(sub c). It is shown, however, that when A(sub c)less than or equal to 0.9, biases for all-sky transmittance stemming from use of A(sub c) as opposed to A'(sub c) are roughly 2-5 times smaller than, and opposite in sign to, biases due to neglect of horizontal variability of tau. By neglecting variable tau, all-sky transmittances are underestimated often by more than 0.1 for A(sub c) near 0.75 and this translates into relative errors that can exceed 40% (corresponding errors for all-sky emittance are about 20% for most values of A(sub c). Thus, priority should be given to development of General Circulation Model (GCM) parameterizations that account for the effects of horizontal variations in unresolved tau, effects of cloud sides are of secondary importance. On this note, an efficient stochastic model for computing grid-averaged cloudy-sky flux transmittances is furnished that assumes that distributions of tau, for regions comparable in size to GCM grid cells, can be described adequately by gamma distribution functions. While the plane-parallel, homogeneous model underestimates cloud transmittance by about an order of magnitude when 3D variable cloud transmittances are less than or equal to 0.2 and by approx. 20% to 100% otherwise, the stochastic model reduces these biases often by more than 80%.

Barker, Howard W.↗

DOE OSTI2020

Quantum approximate optimization of the long-range Ising model with a trapped-ion quantum simulator

Quantum computers and simulators may offer significant advantages over their classical counterparts, providing insights into quantum many-body systems and possibly improving performance for solving exponentially hard problems, such as optimization and satisfiability. Here, we report the implementation of a low-depth Quantum Approximate Optimization Algorithm (QAOA) using an analog quantum simulator. We estimate the ground-state energy of the Transverse Field Ising Model with long-range interactions with tunable range, and we optimize the corresponding combinatorial classical problem by sampling the QAOA output with high-fidelity, single-shot, individual qubit measurements. We execute the algorithm with both an exhaustive search and closed-loop optimization of the variational parameters, approximating the ground-state energy with up to 40 trapped-ion qubits. We benchmark the experiment with bootstrapping heuristic methods scaling polynomially with the system size. We observe, in agreement with numerics, that the QAOA performance does not degrade significantly as we scale up the system size and that the runtime is approximately independent from the number of qubits. We finally give a comprehensive analysis of the errors occurring in our system, a crucial step in the path forward toward the application of the QAOA to more general problem instances.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

NASA NTRS2012

Best Merge Region Growing Segmentation with Integrated Non-Adjacent Region Object Aggregation

Best merge region growing normally produces segmentations with closed connected region objects. Recognizing that spectrally similar objects often appear in spatially separate locations, we present an approach for tightly integrating best merge region growing with non-adjacent region object aggregation, which we call Hierarchical Segmentation or HSeg. However, the original implementation of non-adjacent region object aggregation in HSeg required excessive computing time even for moderately sized images because of the required intercomparison of each region with all other regions. This problem was previously addressed by a recursive approximation of HSeg, called RHSeg. In this paper we introduce a refined implementation of non-adjacent region object aggregation in HSeg that reduces the computational requirements of HSeg without resorting to the recursive approximation. In this refinement, HSeg s region inter-comparisons among non-adjacent regions are limited to regions of a dynamically determined minimum size. We show that this refined version of HSeg can process moderately sized images in about the same amount of time as RHSeg incorporating the original HSeg. Nonetheless, RHSeg is still required for processing very large images due to its lower computer memory requirements and amenability to parallel processing. We then note a limitation of RHSeg with the original HSeg for high spatial resolution images, and show how incorporating the refined HSeg into RHSeg overcomes this limitation. The quality of the image segmentations produced by the refined HSeg is then compared with other available best merge segmentation approaches. Finally, we comment on the unique nature of the hierarchical segmentations produced by HSeg.

Tilton, James C.↗

NASA NTRS1999

The IRAS Galaxy Atlas (IGA)

In 1993 we proposed a project to NASA having the goal of producing a new infrared map of our Galaxy. In particular, we proposed to reprocess the IRAS data taken in the early 1980's using modern image processing algorithms and the large Intel parallel computers of the Center for Advanced Computing Research, (at that time called the Caltech Concurrent Supercomputing Facilities - CCSF). The rationale was simple: what took approximately 100 days on a typical workstation would take less than a day on the multi-processor parallel computers, thus making a high-resolution infrared atlas of the Galaxy feasible.

Prince, Thomas A.↗

NASA NTRS2015

Modeling of Non-Isothermal Cryogenic Fluid Sloshing

A computational fluid dynamic model was used to simulate the thermal destratification in an upright self-pressurized cryostat approximately half-filled with liquid nitrogen and subjected to forced sinusoidal lateral shaking. A full three-dimensional computational grid was used to model the tank dynamics, fluid flow and thermodynamics using the ANSYS Fluent code. A non-inertial grid was used which required the addition of momentum and energy source terms to account for the inertial forces, energy transfer and wall reaction forces produced by the shaken tank. The kinetics-based Schrage mass transfer model provided the interfacial mass transfer due to evaporation and condensation at the sloshing interface. The dynamic behavior of the sloshing interface, its amplitude and transition to different wave modes, provided insight into the fluid process at the interface. The tank pressure evolution and temperature profiles compared relatively well with the shaken cryostat experimental test data provided by the Centre National D'Etudes Spatiales.

Cryogenics↗