Search NASA⌕ Search

SEARCH · Search NASA

Results for “linear approximation”

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 1,153 records · Page 64

Scalable Extended Dynamic Mode Decomposition Using Random Kernel Approximation

The Koopman operator is a linear, infinite-dimensional operator that governs the dynamics of system observables; extended dynamic mode decomposition (EDMD) is a data-driven method for approximating the Koopman operator using functions (features) of the system state snapshots. This paper investigates an approach to EDMD in which the features used provide random approximations to a particular kernel function. The objective of this is computational economy for large data sets: EDMD is generally ill-suited for problems with large state dimension, and its dual kernel formulation (KDMD) is well-suited for such problems only if the number of data snapshots is relatively small. We discuss two specific methods for generating features: random Fourier features and the Nÿstrom method. The first method is a data-independent method for translation-invariant kernels only and involves random sampling in feature space; the second method is a data-dependent empirical method that may be used for any kernel and involves random sampling of data. We first discuss how these ideas may be applied in an EDMD context, as well as a means for adaptively adding random Fourier features. As a result, we demonstrate these methods on two example problems and conclude with an analysis of the relative benefits and drawbacks of each method.

59 BASIC BIOLOGICAL SCIENCES↗

Legendre-Tau approximations for functional differential equations

The numerical approximation of solutions to linear functional differential equations are considered using the so called Legendre tau method. The functional differential equation is first reformulated as a partial differential equation with a nonlocal boundary condition involving time differentiation. The approximate solution is then represented as a truncated Legendre series with time varying coefficients which satisfy a certain system of ordinary differential equations. The method is very easy to code and yields very accurate approximations. Convergence is established, various numerical examples are presented, and comparison between the latter and cubic spline approximations is made.

Ito, K.↗

Test particle propagation in magnetostatic turbulence. 3: The approach to equilibrium

The asymptotic behavior, for large time, of the quasi-linear diabatic solutions and their local approximations is considered. A time averaging procedure is introduced which yields the averages of these solutions over time intervals which contain only large time values. A discussion of the quasi-linear diabatic solutions which is limited to those solutions that are bounded from below as functions of time is given. It is shown that as the upper limit of the time averaging interval is allowed to approach infinity the time averaged quasi-linear diabatic solutions must approach isotropy (mu-independence). The first derivative with respect to mu of these solutions is also considered. This discussion is limited to first derivatives which are bounded functions of time. It is shown that as the upper limit of the time averaging interval is allowed to approach infinity, the time averaged first derivative must approach zero everywhere in mu except at mu = 0 where it must approach a large value which is calculated. The impact of this large derivative on the quasi-linear expansion scheme is discussed. An H-theorem for the first local approximation to the quasi-linear diabatic solutions is constructed. Without time averaging, the H-theorem is used to determine sufficient conditions for the first local approximate solutions to asymptote, with increasing time, to exactly the same final state which the time averaged quasi-linear diabatic solutions must approach as discussed above.

Klimas, A. J.↗

Projection-based multifidelity linear regression for data-scarce applications

Surrogate modeling for systems with high-dimensional quantities of interest remains challenging, particularly when training data are costly to acquire. This work develops multifidelity methods for multiple-input multiple-output linear regression targeting data-limited applications with high-dimensional outputs. Multifidelity methods integrate many inexpensive low-fidelity model evaluations with limited, costly high-fidelity evaluations. We introduce two projection-based multifidelity linear regression approaches with linear and nonlinear features that leverage principal component basis vectors for dimensionality reduction and combine multifidelity data through: (i) a direct data augmentation using low-fidelity data, and (ii) a data augmentation incorporating explicit linear corrections between low-fidelity and high-fidelity data. The data augmentation approaches combine high-fidelity and low-fidelity data into a unified training set and train the linear regression model through weighted least squares with fidelity-specific weights. We introduce a proximity-based weighting scheme with automatic weight selection strategy through cross-validation. Here, the proposed multifidelity linear regression methods are demonstrated on approximating the surface pressure field of a hypersonic vehicle in flight and the temperature field on an aircraft disc braking system. In an ultra low-data regime of no more than twelve high-fidelity samples, multifidelity linear regression achieves approximately 2% – 12% improvement in median accuracy and a higher R 2 score relative to single-fidelity methods at comparable computational cost.

data augmentation↗

Digital control of high performance aircraft using adaptive estimation techniques

In this paper, an adaptive signal processing algorithm is joined with gain-scheduling for controlling the dynamics of high performance aircraft. A technique is presented for a reduced-order model (the longitudinal dynamics) of a high performance STOL aircraft. The actual controller views the nonlinear behavior of the aircraft as equivalent to a randomly switching sequence of linear models taken from a preliminary piecewise-linear fit of the system nonlinearities. The adaptive nature of the estimator is necessary to select the proper sequence of linear models along the flight trajectory. Nonlinear behavior is approximated by effective switching of the linear models at random times, with durations reflecting aircraft motion in response to pilot commands.

Van Landingham, H. F.↗

Legendre-tau approximations for functional differential equations

The numerical approximation of solutions to linear retarded functional differential equations are considered using the so-called Legendre-tau method. The functional differential equation is first reformulated as a partial differential equation with a nonlocal boundary condition involving time-differentiation. The approximate solution is then represented as a truncated Legendre series with time-varying coefficients which satisfy a certain system of ordinary differential equations. The method is very easy to code and yields very accurate approximations. Convergence is established, various numerical examples are presented, and comparison between the latter and cubic spline approximation is made.

Ito, K.↗

High Angular Resolution Imaging of Solar Radio Bursts from the Lunar Surface

Locating low frequency radio observatories on the lunar surface has a number of advantages, including positional stability and a very low ionospheric radio cutoff. Here, we describe the Radio Observatory on the lunar Surface for Solar studies (ROLSS), a concept for a low frequency, radio imaging interferometric array designed to study particle acceleration in the corona and inner heliosphere. ROLSS would be deployed during an early lunar sortie or by a robotic rover as part of an unmanned landing. The preferred site is on the lunar near side to simplify the data downlink to Earth. The prime science mission is to image type II and type III solar radio bursts with the aim of determining the sites at and mechanisms by which the radiating particles are accelerated. Secondary science goals include constraining the density of the lunar ionosphere by measuring the low radio frequency cutoff of the solar radio emissions or background galactic radio emission, measuring the flux, particle mass, and arrival direction of interplanetary and interstellar dust, and constraining the low energy electron population in astrophysical sources. Furthermore, ROLSS serves a pathfinder function for larger lunar radio arrays. Key design requirements on ROLSS include the operational frequency and angular resolution. The electron densities in the solar corona and inner heliosphere are such that the relevant emission occurs below 10 M Hz, essentially unobservable from Earth's surface due to the terrestrial ionospheric cutoff. Resolving the potential sites of particle acceleration requires an instrument with an angular resolution of at least 2 deg at 10 MHz, equivalent to a linear array size of approximately one kilometer. The major components of the ROLSS array are 3 antenna arms, each of 500 m length, arranged in a Y formation, with a central electronics package (CEP) at their intersection. Each antenna arm is a linear strip of polyimide film (e.g., Kapton(TradeMark)) on which 16 single polarization dipole antennas are located by depositing a conductor (e.g., silver). The arms also contain transmission lines for carrying the radio signals from the science antennas to the CEP. Operations would consist of data acquisition during the lunar day, with data downlinks to Earth one or more times every 24 hours.

MacDowall, Robert J.↗

Rotation of linear polarization plane and circular polarization from cosmological pseudoscalar fields

We discuss the rotation of the linear polarization plane and the production of circular polarization generated by a cosmological pseudoscalar field. We compute analytically and numerically the propagation of the Stokes parameters from the last scattering surface for an oscillating and a monotonic decreasing pseudoscalar field. For the models studied in this paper, we show the comparison between the widely used approximation in which the linear polarization rotation angle is constant in time and the exact result.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Delay Line Detectors for the UVCS and Sumer Instruments on the SOHO Satellite

Microchannel plate based detectors with cross delay line image readout have been rapidly implemented for the SUMER and UVCS instruments aboard the Solar Orbiting Heliospheric Observatory (SOHO) mission to be launched in July 1995. In October 1993 a fast track program to build and characterize detectors and detector control electronics was initiated. We present the detector system design for the SOHO UVCS and SUMER detector programs, and results from the detector test program. Two deliverable detectors have been built at this point, a demonstration model for UVCS, and the flight Ly alpha detector for UVCS, both of which are to be delivered in the next few weeks. Test results have also been obtained with one other demonstration detector system. The detector format is 26mm x 9mm, with 1024 x 360 digitized pixels, using a low resistance Z stack of microchannel plates (MCP's) and a multilayer cross delay line anode (XDL). This configuration provides gains of approximately 2 x 10(exp 7) with good pulse height distributions (less than 50% FWHM) under uniform flood illumination, and background levels typical for this configuration (approximately 0.6 event cm (exp -2)sec(exp -1)). Local counting rates up to about 400 events/pixel/sec have been achieved with no degradation of the MCP gain. The detector and event encoding electronics achieves about 25 millimeter FVHM with good linearity (plus or minus approximately 1 pixel) and is stable to high global counting rates (greater than 4 x 10(exp 5) events sec(exp -1)). Flat field images are dominated by MCP fixed pattern noise and are stable, but the MCP multifiber modulation usually expected is uncharacteristically absent. The detector and electronics have also successfully passed both thermal vacuum and vibration tests.

Seigmund, O. H. W.↗

Correlation between eigenvalues and sorted diagonal elements of a large dimensional matrix

We show the functional dependence of eigenvalues in terms of sorted diagonal elements of a Hamiltonian matrix in the nuclear shell model (NSM), a matrix with uniform distribution and that with normal distribution. For a realistic two-body interaction, its relation is approximately expressed by a linear function, especially for the most elements in the intermediate region. We also derive their functional dependences for the uniform distribution and the normal distribution analytically. As a special case, the functional relation for the normal distribution turns out to be approximated by a hyperbolic-tangent function.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Substorm recurrence during steady and variable solar wind driving: Evidence for a normal mode in the unloading dynamics of the magnetosphere

Farrugia et al (1993) have recently studied substorm activity driven by the passage of an interplanetary magnetic cloud during which the interplanetary magnetic field turned southward for approximately 18 hours. It was shown that both the epsilon and the VB(sub s) parameters varied slowly on the timescale of a substorm but changed considerably over the interval as a whole. The substorm occurrence rate did not reflect the variation in magnetospheric energy loading rate as measured by these parameters but, rather, remained roughly constant with a 50-min average period. Klimas et al. (1992) showed that the Faraday loop analog model of geomagnetic activity predicts this single unloading rate under various constant loading rates. However, various model parameters were adjusted to yield a 1-hour unloading period in agreement with the Bargatze et al. (1985) linear prediction filters and in approximate agreement with the Farrugia et al. (1993) results. It has since been found necessary to add a slow relaxation mechanism to the Faraday loop model to allow for its approach to a ground state during long periods of inactivity. It is proposed that the relaxation mechanism is provided by slow convection of magnetic flux out of the magnetotail to the dayside magnetosphere. In addition, a rudimentary representation of magnetotail-ionosphere coupling has been added to enable comparison of model output to measured AL. The present study is of the modified Faraday loop model response to solar wind input from the Bargatze et al. data set with comparison of its output to concurrent AL. This study has removed the degree of freedom in parameter choice that had earlier allowed adjustments toward the 1-hour unloading period and has, instead, yielded the 1-hour unloading period under various constant loading rates. It is demonstrated that the second peak of the bimodal Bargatze et al. linear prediction filters at approximately equal 1-hour lag and the approximately constant substorm recurrence rate observed by Farrugia et al. can be interpreted as both being due to the existence of a normal unloading recurrence period in the dynamics of the magnetosphere.

Klimas, A. J.↗

Calculation of excitation energies from the CC2 linear response theory using Cholesky decomposition

A new implementation of the approximate coupled cluster singles and doubles CC2 linear response model is reported. It employs a Cholesky decomposition of the two-electron integrals that significantly reduces the computational cost and the storage requirements of the method compared to standard implementations. Our algorithm also exploits a partitioning form of the CC2 equations which reduces the dimension of the problem and avoids the storage of doubles amplitudes. We present calculation of excitation energies of benzene using a hierarchy of basis sets and compare the results with conventional CC2 calculations. The reduction of the scaling is evaluated as well as the effect of the Cholesky decomposition parameter on the quality of the results. The new algorithm is used to perform an extrapolation to complete basis set investigation on the spectroscopically interesting benzylallene conformers. A set of calculations on medium-sized molecules is carried out to check the dependence of the accuracy of the results on the decomposition thresholds. Moreover, CC2 singlet excitation energies of the free base porphin are also presented.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

New Algorithms for Estimating Spacecraft Position Using Scanning Techniques for Deep Space Network Antennas

As more and more nonlinear estimation techniques become available, our interest is in finding out what performance improvement, if any, they can provide for practical nonlinear problems that have been traditionally solved using linear methods. In this paper we examine the problem of estimating spacecraft position using conical scan (conscan) for NASA's Deep Space Network antennas. We show that for additive disturbances on antenna power measurement, the problem can be transformed into a linear one, and we present a general solution to this problem, with the least square solution reported in literature as a special case. We also show that for additive disturbances on antenna position, the problem is a truly nonlinear one, and we present two approximate solutions based on linearization and Unscented Transformation respectively, and one 'exact' solution based on Markov Chain Monte Carlo (MCMC) method. Simulations show that, with the amount of data collected in practice, linear methods perform almost the same as MCMC methods. It is only when we artificially reduce the amount of collected data and increase the level of noise that nonlinear methods show significantly better accuracy than that achieved by linear methods, at the expense of more computation.

Metropolis-Hastings (MH)↗

Observing Solar Radio Bursts from the Lunar Surface

Locating low frequency radio observatories on the lunar surface has a number of advantages. Here, we describe the Radio Observatory for Lunar Sortie Science (ROLSS), a concept for a low frequency, radio imaging interferometric array designed to study particle acceleration in the corona and inner heliosphere. ROLSS would be deployed during an early lunar sortie or by a robotic rover as part of an unmanned landing. The prime science mission is to image type II and type III solar radio bursts with the aim of determining the sites at and mechanisms by which the radiating particles are accelerated. Secondary science goals include constraining the density of the lunar ionosphere by searching for a low radio frequency cutoff of the solar radio emissions and constraining the low energy electron population in astrophysical sources. Furthermore, ROLSS serves a pathfinder function for larger lunar radio arrays. Key design requirements on ROLES include the operational frequency and angular resolution. The electron densities in the solar corona and inner heliosphere are such that the relevant emission occurs below 10 MHz, essentially unobservable from Earth's surface due to the terrestrial ionospheric cutoff. Resolving the potential sites of particle acceleration requires an instrument with an angular resolution of at least 2 deg, equivalent to a linear array size of approximately 500 meters. Operations would consist of data acquisition during the lunar day, with regular data downlinks. The major components of the ROLSS array are 3 antenna arms arranged in a Y shape, with a central electronics package (CEP). Each antenna arm is a linear strip of polyimide film (e.g., Kapton (TM)) on which 16 single polarization dipole antennas are located by depositing a conductor (e.g., silver). The arms also contain transmission lines for carrying the radio signals from the science antennas to the CEP.

MacDowall, R. J.↗

High-Fidelity Calibration and Characterization of Hyperspectral Computed Tomography System

This document archives the results developed by the Lab Directed Research and Development (LDRD) project sponsored by Sandia National Laboratories (SNL). In this work, a numerical study was performed to show the feasibility of approximating the non-linear operator of SNL's unique high-energy hyperspectral computed tomography system as a sequence of linear operators. The four main results gained from this work include the development of a simulation test-bed using a particle-transport Monte Carlo approach; the demonstration to assemble a linear operator of almost-arbitrary resolution for a given narrow energy window, developing a compression approach to dramatically reduce the size of the linear operator via a spline approach, and the demonstration of using the linear operator to perform processing of x-ray data; in this case, the development of an iterative reconstruction method. This numerical study has indicated that if these results can be replicated on the SNL system, the improved performance could be revolutionary as the method to approximate the nonlinear operator for a hyperspectral CT system is not feasible to perform on a traditional CT system.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Solution algorithms for non-linear singularly perturbed optimal control problems

The applicability and usefulness of several classical and other methods for solving the two-point boundary-value problem which arises in non-linear singularly perturbed optimal control are assessed. Specific algorithms of the Picard, Newton and averaging types are formally developed for this class of problem. The computational requirements associated with each algorithm are analysed and compared with the computational requirement of the method of matched asymptotic expansions. Approximate solutions to a linear and a non-linear problem are obtained by each method and compared.

Ardema, M. D.↗

Kernel Manifolds: Nonlinear‐Augmentation Dimensionality Reduction Using Reproducing Kernel Hilbert Spaces

This paper generalizes recent advances on quadratic manifold (QM) dimensionality reduction by developing kernel methods-based nonlinear-augmentation dimensionality reduction. QMs, and more generally feature map-based nonlinear corrections, augment linear dimensionality reduction with a nonlinear correction term in the reconstruction map to overcome approximation accuracy limitations of purely linear approaches. While feature map-based approaches typically learn a least squares optimal polynomial correction term, we generalize this approach by learning an optimal nonlinear correction from a user-defined reproducing kernel Hilbert space. Our approach allows one to impose arbitrary nonlinear structure on the correction term, including polynomial structure, and includes feature map and radial basis function-based corrections as special cases. Furthermore, our method has relatively low training cost and has monotonically decreasing error as the latent space dimension increases. In conclusion, we compare our approach to proper orthogonal decomposition and several recent QM approaches on data from several example problems.

kernel methods↗