Search NASA⌕ Search

SEARCH · Search NASA

Results for “Function Estimation”

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 397 records · Page 22
All sourcesNASA NTRSDOE OSTI
NASA NTRS1984

The electrical conductivity of the upper mantle as estimated from satellite magnetic field data

The electrical conductivity of the upper mantle is estimated from low-latitude magnetic field variations caused by large fluctuations in the equatorial ring current. The data base is derived from magnetic field magnitude data measured by satellites OGO 2, 4, and 6, which offer better global coverage than land-based observatories. The procedures of analysis consist of: (1) separation of the disturbance field into internal and external parts relative to the surface of the earth, (2) estimation of an electromagnetic response function Q(omega) which relates the internally generated magnetic field variations to the external variations due to the ring current, and (3) interpretation of the estimated response function using theoretical response functions for assumed conductivity profiles. Special consideration is given to possible oceanic and ionospheric effects. Best estimates of the geomagnetic response function Q(omega) for 0.2 to 2.0 cpd indicate an upper mantle conductivity of the order of 0.01 S/m.

Didwall, E. M.↗
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↗
NASA NTRSDate not supplied

Assessment of Sensor Footprint Size and Comparison of Commercial Smallsat Images

Science users of commercial satellite data build their studies on the properties of the satellite data they work with, pixel size being one of the determining factors. However, pixel size and nominal image footprint size can be different, impacting the scale of features discernable in the satellite images. Here, we assess geometric properties such as nominal image footprint size for Planet Labs' SuperDove series, MAXAR's WorldView series, and potentially BlackSky. Additionally, we assess temporal change in nominal footprint size for the SuperDove series. Nominal Image footprint size is assessed over the CalVal sites in Baotou, China; Shadnagar, India; and Big Spring, TX. Edge spread functions are constructed along the black/white transitions, from which the line spread functions are constructed and modulation transfer functions of the scene are estimated. Scene full width half maximum (FWHM), which represents sensor footprint size, is estimated from the line spread function. The average nominal footprint size is 3.3 pixels for SuperDove series, 1.5 pixels for WorldView-2, and 1.3 pixels for WorldView-3. The SuperDove series nominal footprint size improves with time in orbit, from an average of 3.4 pixels soon after launch to an average of 3.2 pixels one or more years after launch.

PlanetScope↗
NASA NTRS1989

On the signal-to-noise ratio in IUE high-dispersion spectra

An observational and data reduction technique for fixed pattern noise (FPN) and random noise (RN) in fully extracted IUE high-dispersion spectra is described in detail, along with actual empirical values of signal-to-noise ratio (S/N) achieved. A co-addition procedure, involving SWP and LWR cameras observations of the same spectrum at different positions in the image format, provides a basis to disentangle FPN from RN, allowing each average amplitude, within a given wavelength interval, to be estimated as a function of average flux number. Empirical curves, derived with the noise algorithm, make it possible to estimate the S/N in individual spectra at the wavelengths investigated. The average S/N at the continuum level in well-exposed stellar spectra varies from 10 to 20, for the orders analyzed, depending on position in the spectral format. The co-addition procedure yields an improvement in S/N by factors ranging from 2.3 to 2.9. Direct measurements of S/N in narrow, line-free wavelength intervals of individual and co-added spectra for weak-lined stars yield comparable, or in some cases somewhat higher, S/N values and improvement factors.

Leckrone, David S.↗
NASA NTRS1982

A structure function representation theorem with applications to frequency stability estimation

Random processes with stationary nth differences serve as models for oscillator phase noise. A theorem which obtains the structure function (covariance of the nth differences) of such a process in terms of the differences of a single function of one time variable is proven. In turn, this function can easily be obtained from the spectral density of the process. The theorem is used for computing the variance of two estimators of frequency stability.

Greenhall, C. A.↗
DOE OSTI2019

Public charging infrastructure for plug-in electric vehicles: What is it worth?

A lack of charging infrastructure is an important barrier to the growth of the plug-in electric vehicle (PEV) market. Public charging infrastructure has tangible and intangible value, such as reducing range anxiety or building confidence in the future of the PEV market. Quantifying the value of public charging infrastructure can inform analysis of investment decisions and can help predict the impact of charging infrastructure on future PEV sales. Estimates of willingness to pay (WTP) based on stated preference surveys are limited by consumers' lack of familiarity with PEVs. As an alternative, we focus on quantifying the tangible value of public PEV chargers in terms of their ability to displace gasoline use for PHEVs and to enable additional electric (e-) vehicle miles for BEVs, thereby mitigating the limitations of shorter range and longer recharging time. Simulation studies provide data that can be used to quantify e-miles enabled by public chargers and the value of additional e-miles can be inferred from econometric estimates of WTP for increased vehicle range. Functions are synthesized that estimate the WTP for public charging infrastructure by plug-in hybrid and battery electric vehicles, conditional on vehicle range, annual vehicle travel, pre-existing charging infrastructure, energy prices, vehicle efficiency, and household income. A case study based on California's public charging network in 2017 indicates that, to the purchaser of a new BEV with a 100-mile range and home recharging, existing public fast chargers are worth about $1500 for intraregional travel, and fast chargers along intercity routes are valued at over $6500.

33 ADVANCED PROPULSION SYSTEMS↗
DOE OSTI2022

Uncertainty Quantification for Point Spread Functions in X-Ray Radiography

X-ray radiographers have tools for estimating the point spread function (PSF) which determines the extent of the blur in images. Knowledge of uncertainties in their estimations can be extremely valuable, and existing methods do not provide this information. We have developed a new technique using a Bayesian framework that will provide useful uncertainties to the estimated PSF.

point spread function↗
NASA NTRS1987

Mars - VLA observations of the northern hemisphere and the north polar region at wavelengths of 2 and 6 cm

Calculations based on 2- and 6-cm observations of Mars with the A configuration of the VLA have yielded a whole-disk effective dielectric constant of 2.34 + or - 0.05, implying a subsurface density of 1.24 + or - 0.11 g/cu cm at 2 cm, as well as 1.45 + or - 0.10 g/cu cm effective density and 2.70 + or - 0.10 dielectric constant at 6 cm. These parameters have also been estimated as a function of latitude over the 15 deg S - 60 deg N range; subsurface radio absorption length was estimated to be about 15 wavelengths at most of these latitudes. Most of the subsurface density calculations yielded results in the 1-2-g/cu cm range, implying that the subsurface is not very different from the surface observed by Viking and Mariner spacecraft; the decrease in correlation with depth is in keeping with slow variation of the subsurface in the near-subsurface region.

Rudy, Donald J.↗
DOE OSTI2024

A Cellular Automaton Simulation for Predicting Phase Evolution in Solid-State Reactions

New computational tools for solid-state synthesis recipe design are needed in order to accelerate the experimental realization of novel functional materials proposed by high-throughput materials discovery workflows. This work contributes a cellular automaton simulation framework for predicting the time-dependent evolution of intermediate and product phases during solid-state reactions as a function of precursor choice and amount, reaction atmosphere, and heating profile. The simulation captures the effects of reactant particle spatial distribution, particle melting, and reaction atmosphere. Reaction rates based on rudimentary kinetics are estimated using density functional theory data from the Materials Project and machine learning estimators for the melting point and the vibrational entropy component of the Gibbs free energy. The resulting simulation framework allows for the prediction of the likely outcome of a reaction recipe before any experiments are performed. We analyze five experimental solid-state recipes for BaTiO 3 , CaZrN 2 , and YMnO 3 found in the literature to illustrate the performance of the model in capturing reaction selectivity and reaction pathways as a function of temperature and precursor choice. This simulation framework offers an easier way to optimize existing recipes, aid in the identification of intermediates, and design effective recipes for yet unrealized inorganic solids in silico .

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2019

GMLS-Nets: A Framework for Learning from Unstructured Data

Data fields sampled on irregularly spaced points arise in many applications in the sciences and engineering. For regular grids, Convolutional Neural Networks (CNNs) have been successfully used to gaining benefits from weight sharing and invariances. We generalize CNNs by introducing methods for data on unstructured point clouds based on Generalized Moving Least Squares (GMLS). GMLS is a nonparametric technique for estimating linear bounded functionals from scattered data, and has recently been used in the literature for solving partial differential equations. By parameterizing the GMLS estimator, we obtain learning methods for operators with unstructured stencils. In GMLS-Nets the necessary calculations are local, readily parallelizable, and the estimator is supported by a rigorous approximation theory. We show how the framework may be used for unstructured physical data sets to perform functional regression to identify associated differential operators and to regress quantities of interest. The results suggest the architectures to be an attractive foundation for data-driven model development in scientific machine learning applications.

42 ENGINEERING↗
DOE OSTI2019

GMLS-Nets: Scientific Machine Learning Methods for Unstructured Data

Data fields sampled on irregularly spaced points arise in many applications in the sciences and engineering. For regular grids, Convolutional Neural Networks (CNNs) have been successfully used to gaining benefits from weight sharing and invariances. We generalize CNNs by introducing methods for data on unstructured point clouds based on Generalized Moving Least Squares (GMLS). GMLS is a nonparametric technique for estimating linear bounded functionals from scattered data, and has recently been used in the literature for solving partial differential equations. By parameterizing the GMLS estimator, we obtain learning methods for operators with unstructured stencils. In GMLS-Nets the necessary calculations are local, readily parallelizable, and the estimator is supported by a rigorous approximation theory. We show how the framework may be used for unstructured physical data sets to perform functional regression to identify associated differential operators and to regress quantities of interest. The results suggest the architectures to be an attractive foundation for data-driven model development in scientific machine learning applications

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

GMLS-Nets: A Framework for Learning from Unstructured Data

Data fields sampled on irregularly spaced points arise in many applications in the sciences and engineering. For regular grids, Convolutional Neural Networks (CNNs) have been successfully used to gaining benefits from weight sharing and invariances. We generalize CNNs by introducing methods for data on unstructured point clouds based on Generalized Moving Least Squares (GMLS). GMLS is a non-parametric technique for estimating linear bounded functionals from scattered data, and has recently been used in the literature for solving partial differential equations. By parameterizing the GMLS estimator, we obtain learning methods for operators with unstructured stencils. In GMLS-Nets the necessary calculations are local, readily parallelizable, and the estimator is supported by a rigorous approximation theory. We show how the framework may be used for unstructured physical data sets to perform functional regression to identify associated differential operators and to regress quantities of interest. The results suggest the architectures to be an attractive foundation for data-driven model development in scientific machine learning applications

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Statistical estimates of the binary properties of rotational variables

ABSTRACT We present a model to estimate the average primary masses, companion mass ranges, the inclination limit for recognizing a rotational variable, and the primary mass spreads for populations of binary stars. The model fits a population’s binary mass function distribution and allows for a probability that some mass functions are incorrectly estimated. Using tests with synthetic data, we assess the model’s sensitivity to each parameter, finding that we are most sensitive to the average primary mass and the minimum companion mass, with less sensitivity to the inclination limit and little to no sensitivity to the primary mass spread. We apply the model to five populations of binary spotted rotational variables identified in ASAS-SN, computing their binary mass functions using RV data from APOGEE. Their average primary mass estimates are consistent with our expectations based on their CMD locations ($\sim 0.75 \, {\rm M}_{\odot }$ for lower main sequence primaries and $\sim 0.9$–$1.2 \, {\rm M}_{\odot }$ for RS CVn and sub-subgiants). Their companion mass range estimates allow companion masses down to $M_2/M_1\simeq 0.1$, although the main sequence population may have a higher minimum mass fraction ($\sim 0.4$). We see weak evidence of an inclination limit $\gtrsim 50^{\circ }$ for the main sequence and sub-subgiant groups and no evidence of an inclination limit in the other groups. No groups show strong evidence for a preferred primary mass spread. We conclude by demonstrating that the approach will provide significantly better estimates of the primary mass and the minimum mass ratio and reasonable sensitivity to the inclination limit with 10 times as many systems.

Phillips, Anya (ORCID:000900051914974X)↗
DOE OSTI2021

Nucleon momentum fraction, helicity and transversity from 2+1-flavor lattice QCD

A detailed analysis of the systematic uncertainties in the calculation of the isovector momentum fraction, ( x ) u - d , helicity moment, ( x ) Δ u - Δ d , and the transversity moment, ( x ) δu - δd , of the nucleon is presented using high-statistics data on seven ensembles of gauge configurations generated by the JLab/W&M/LANL/MIT collaborations using 2 + 1-flavors of dynamical Wilson-clover quarks. The much higher statistics have facilitated better control over all systematics compared to previous lattice calculations. The least understood systematic — excited-state contamination — is quantified by studying the variation of the results as a function of different estimates of the mass gap of the first excited state, obtained from two- and three-point correlation functions, and as a function of the pion mass M π . The final results are obtained using a simultaneous fit in the lattice spacing a , pion mass M π and the finite volume parameter M π L keeping leading order corrections. The data show no significant dependence on the lattice spacing and some evidence for finite-volume corrections. Our final results, in the $ \overline{\mathrm{MS}} $ scheme at 2 GeV, are ( x ) u - d = 0 . 155(17)(20), ( x ) Δ u - Δ d = 0 . 183(14)(20) and ( x ) δu - δd = 0 . 220(18)(20), where the first error is the overall analysis uncertainty assuming excited-state contributions have been removed, and the second is an additional systematic uncertainty due to possible residual excited-state contributions. These results are consistent with phenomenological global fit values.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRSDate not supplied

The Containment Assurance Risk Framework of the Mars Sample Return Program

The Mars Sample Return campaign aims at bringing rock and atmospheric samples from Mars to Earth through a series of robotic missions. These missions would collect the samples being cached and deposited on Martian soil by the Perseverance rover, place them in a container, and launch them into Martian orbit for subsequent capture by an orbiter that would bring them back. Given there exists a non-zero probability that the samples contain biological material, precautions are being taken to design systems that would break the chain of contact between Mars and Earth. These include techniques such as sterilization of Martian particles, redundant containment vessels, and a robust reentry capsule capable of accurate landings without a parachute. Requirements exist that the probability of containment not assured of Martian-contaminated material into Earth’s biosphere be less than one in a million. To demonstrate compliance with this strict requirement, a statistical framework was developed to assess the likelihood of containment loss during each sample return phase and make a statement about the total combined mission probability of containment not assured. The work presented here describes this framework, which considers failure modes or fault conditions that can initiate failure sequences ultimately leading to containment not assured. Reliability estimates are generated from databases, design heritage, component specifications, or expert opinion in the form of probability density functions or point estimates and provided as inputs to the mathematical models that simulate the different failure sequences. The probabilistic outputs are then combined following the logic of several fault trees to compute the ultimate probability of containment not assured. Given the multidisciplinary nature of the problem and the different types of mathematical models used, the statistical tools needed for analysis are required to be computationally efficient. While standard Monte Carlo approaches are used for fast models, a multi-fidelity approach to rare event probabilities is proposed for expensive models. In this paradigm, inexpensive low-fidelity models are developed for computational acceleration purposes while the expensive high-fidelity model is kept in the loop to retain accuracy in the results. This work presents an example of end-to-end application of this framework highlighting the computational benefits of a multi-fidelity approach.

Giuseppe Cataldo↗
NASA NTRS1985

Completely monotone regression estimates of software failure rates

A method for estimating the present failure rate of a program is presented. A crude nonparameter estimate of the failure rate function is obtained from past failure times. This estimate is then smoothed by fitting a completely monotonic function, which is the solution of a quadratic programming problem. The value of the smoothed function at present time is used as the estimate of present failure rate. Results of a Monte Carlo study of performance are given.

Miller, D. R.↗
NASA NTRS1983

A structure function representation theorem with applications to frequency stability estimation

Random processes with stationary nth differences serve as models for oscillator phase noise. The theorem proved here allows one to obtain the structure function (covariances of the nth differences) of such a process in terms of the differences of a single function of one time variable. In turn, this function can easily be obtained from the spectral density of the process. The theorem is used for computing the variance of two estimators of frequency stability.

Greenhall, C. A.↗
NASA NTRS1985

Approximation of Optimal Infinite Dimensional Compensators for Flexible Structures

The infinite dimensional compensator for a large class of flexible structures, modeled as distributed systems are discussed, as well as an approximation scheme for designing finite dimensional compensators to approximate the infinite dimensional compensator. The approximation scheme is applied to develop a compensator for a space antenna model based on wrap-rib antennas being built currently. While the present model has been simplified, it retains the salient features of rigid body modes and several distributed components of different characteristics. The control and estimator gains are represented by functional gains, which provide graphical representations of the control and estimator laws. These functional gains also indicate the convergence of the finite dimensional compensators and show which modes the optimal compensator ignores.

Gibson, J. S.↗