Search NASA⌕ Search

SEARCH · Search NASA

Results for “Distributed 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 1,603 records · Page 89
All sourcesNASA NTRSDOE OSTI
NASA NTRS1971

Estimating Weibull parameters for materials

The statistical analysis of strength and fracture of materials in general, with application to fiber composites are discussed. The weakest link model is considered in a fairly general form, and the resulting equations are demonstrated by using a Weibull distribution for flaws. This distribution appears naturally in a variety of problems, and therefore additional attention is devoted to analysis and statistical estimation connected with this distribution. Special working charts are included to facilitate interpretation of observed data and estimation of parameters. Implications of the size effect are considered for various kinds of flaw distributions. Failure and damage in a fiber-reinforced system are described. Some useful graphs are included for predicting the strength of such a system. Recent data on organic-fiber (PRD 49) composite material is analyzed by the Weibull distribution with the methods presented.

Robinson, E. Y.↗
NASA NTRS1974

An empirical analysis of the distribution of the duration of overshoots in a stationary gaussian stochastic process

This analysis utilizes computer simulation and statistical estimation. Realizations of stationary gaussian stochastic processes with selected autocorrelation functions are computer simulated. Analysis of the simulated data revealed that the mean and the variance of a process were functionally dependent upon the autocorrelation parameter and crossing level. Using predicted values for the mean and standard deviation, by the method of moments, the distribution parameters was estimated. Thus, given the autocorrelation parameter, crossing level, mean, and standard deviation of a process, the probability of exceeding the crossing level for a particular length of time was calculated.

Parrish, R. S.↗
DOE OSTI2025

Calculating Methane Emissions from Offshore Facilities Using Bottom-Up Methods

With changing demands in regulation, understanding methane emissions from offshore oil and gas production infrastructure has become increasingly important. Reported emissions from facilities in the Gulf of Mexico range from zero to thousands of tons of methane per hour, but these is currently no clear understanding of how this range compares to expected emissions from normally operating facilities. To generate realistic emission estimates, we create two bottom-up models that simulate emissions from facilities operating in the Gulf of Mexico. We estimate type 1 prototypical facilities (typically unmanned, older, lower-producing platforms in shallow water with little processing equipment, compressors, or storage tanks) to emit an average of 13 kg CH 4 h −1 , which corresponds to a loss of 2.7% of the average facility production. Type 2 prototypical facilities (continuously manned, higher production and operate in deeper water with processing equipment, oil storage tanks, compressors and power generation) emit an average of 88 kg CH 4 h −1 , which corresponds to a loss of 2.5% of production. The average measured emission from type 1 facilities was 18 kg CH 4 h −1 with a median production loss estimated at 8%. The average measured emission from type 2 facilities was 36 kg CH 4 h −1 with a median production loss estimated at 2.4%. Using emission factors that consider the long-tail emission distribution partly reconciles the difference between modelled and measured emission estimates, but we suggest the current the fugitive emission estimate may be an underestimate and more data on the number and size of fugitive emissions could explain differences between the modelled and measured emission estimate. We suggest the bottom-up approach described here that uses production data coupled with facility equipment could be used to identify facilities that have abnormally large measured emissions, caused by methodological failure or larger than expected fugitive emissions, which should be targeted for further evaluation resulting in remeasurement or identification of source type so that a more accurate estimates can be made on the absolute emission.

bottom-up↗
NASA NTRS2012

The Kinematic and Microphysical Control of Storm Integrated Lightning Flash Extent

The objective of this preliminary study is to investigate the kinematic and microphysical control of lightning properties, particularly those that may govern the production of nitrogen oxides (NOx) in thunderstorms, such as flash rate, type and extent. The mixed-phase region is where the noninductive charging (NIC) process is thought to generate most storm electrification during rebounding collisions between ice particles in the presence of supercooled water. As a result, prior radar-based studies have demonstrated that lightning flash rate is well correlated to kinematic and microphysical properties in the mixed-phase region of thunderstorms such as updraft volume, graupel mass, or ice mass flux. There is also some evidence that lightning type is associated with the convective state. Intracloud (IC) lightning tends to dominate during the updraft accumulation of precipitation ice mass while cloud-to-ground (CG) lightning is more numerous during the downdraft-driven descent of radar echo associated with graupel and hail. More study is required to generalize these relationships, especially regarding lightning type, in a wide variety of storm modes and meteorological conditions. Less is known about the co-evolving relationship between storm kinematics, microphysics, morphology and three-dimensional flash extent, despite its importance for lightning NOx production. To address this conceptual gap, the NASA MSFC Lightning Nitrogen Oxides Model (LNOM) is applied to North Alabama Lightning Mapping Array (NALMA) and Vaisala National Lightning Detection NetworkTM (NLDN) observations following ordinary convective cells through their lifecycle. LNOM provides estimates of flash type, channel length distributions, lightning segment altitude distributions (SADs) and lightning NOx production profiles. For this study, LNOM is applied in a Lagrangian sense to well isolated convective cells on 3 April 2007 (single cell and multi-cell hailstorm, non-severe multicell) and 6 July 2007 (non-severe multi-cell) over Northern Alabama. The LNOM lightning characteristics are compared to the evolution of updraft and precipitation properties inferred from dual-Doppler and polarimetric radar analyses applied to observations from a nearby Doppler radar network, including the UA Huntsville Advanced Radar for Meteorological and Operational Research (ARMOR, C-band, polarimetric). The LNOM estimated SAD and lightning NOx production profiles are placed in the context of radar derived profiles of vertical motion, precipitation types and amounts. Finally, these analyses are used to determine if storm integrated flash channel extent is as well correlated to volumetric updraft and precipitation ice characteristics in the mixed phase region as flash rate for these individual convective cells.

Carey, Lawrence↗
DOE OSTI2022

A General Decentralized Dynamic State Estimation With Synchronous Generator Magnetic Saturation

We report the saturation of the nonlinear magnetic circuit of synchronous generators is often neglected when performing Kalman filter-based dynamic state estimation (DSE), yielding significant estimation bias. This letter addresses this problem and proposes a generalized DSE framework to handle magnetic saturation. Moreover, this letter derives a state initialization procedure that improves the Kalman filter tracking speed. The framework is flexible in dealing with different saturation functions and generator models. Numerical results on the Texas 2000-bus system verify the effectiveness of the proposed method.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2019

Convergence Rates for Empirical Estimation of Binary Classification Bounds

Bounding the best achievable error probability for binary classification problems is relevant to many applications including machine learning, signal processing, and information theory. Many bounds on the Bayes binary classification error rate depend on information divergences between the pair of class distributions. Recently, the Henze–Penrose (HP) divergence has been proposed for bounding classification error probability. We consider the problem of empirically estimating the HP-divergence from random samples. We derive a bound on the convergence rate for the Friedman–Rafsky (FR) estimator of the HP-divergence, which is related to a multivariate runs statistic for testing between two distributions. The FR estimator is derived from a multicolored Euclidean minimal spanning tree (MST) that spans the merged samples. We obtain a concentration inequality for the Friedman–Rafsky estimator of the Henze–Penrose divergence. We validate our results experimentally and illustrate their application to real datasets.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Solving high-dimensional inverse problems using amortized likelihood-free inference with noisy and incomplete data

Here, we present a likelihood-free probabilistic inversion method based on normalizing flows for high-dimensional inverse problems. The proposed method is composed of two complementary networks: a summary network for data compression and an inference network for parameter estimation. The summary network encodes raw observations into a fixed-size vector of summary features, while the inference network generates samples of the approximate posterior distribution of the model parameters based on these summary features. The posterior samples are produced in a deep generative fashion by sampling from a latent Gaussian distribution and passing these samples through an invertible transformation. We construct this invertible transformation by sequentially alternating conditional invertible neural network and conditional neural spline flow layers. The summary and inference networks are trained simultaneously. We apply the proposed method to an inversion problem in groundwater hydrology to estimate the posterior distribution of the log-conductivity field conditioned on spatially sparse time-series observations of the system’s hydraulic head responses. The conductivity field is represented with 706 degrees of freedom in the considered problem. Comparison with the likelihood-based iterative ensemble smoother PEST-IES method demonstrates that the proposed method accurately estimates the parameter posterior distribution and the observations’ predictive posterior distribution at a fraction of the inference time of PEST-IES.

conditional invertible neural network↗
DOE OSTI2021

Performance Comparison of Circular and Spherical Error Probable Estimators

This report compares the performance of three Circular Error Probable (CEP) estimators: the Grubbs-Patnaik estimator, a new, non-iterative, radial-integration estimator, and a median estimator. It also compares the performance of two Spherical Error Probable (SEP) estimators. The performance of each estimator is assessed in terms of bias, uncertainty, robustness, and computational complexity. Robustness is evaluated with respect to outliers, variations in the underlying statistical distribution characterizing munition impact positions, and impact-position measurement errors. The performance assessments indicate the radial-integration and Grubbs-Patnaik estimators perform nearly identically providing the statistical distribution of impact-position coordinates is jointly normal with zero means. In that case, both estimators outperform the median estimator by about 2% relative to the true CEP in terms of estimator uncertainty. The bias performance of the radial-integration and median estimators is close to zero for jointly normal impacts, however, the Grubbs-Patnaik estimator can be significantly biased for jointly normal impacts with non-zero means. When the statistical distribution characterizing impact positions is known, but not jointly normal, the radial-integration estimator is superior. In this case, the median estimator also outperforms the Grubbs-Patnaik estimator but is not quite as good as the radial-integration estimator. If the statistical distribution characterizing impacts is unknown and not jointly normal, or if distribution parameters are difficult or impractical to estimate, or if test data is corrupted with outliers, then the median estimator dramatically outperforms the other estimators, especially in terms of estimation bias. Unexpectedly, measurement noise did not significantly degrade the performance of any of the estimators, except for cases with signal to noise ratios less than five. Although the Grubbs-Patnaik estimator has remained the gold standard for CEP estimation for over half a century, the performance assessments indicate the new, non-iterative, radial-integration estimator and the median estimator offer significant advantages and, in most practical real-world conditions, are superior estimators. These estimators are also useful for SEP estimation whereas the Grubbs-Patnaik estimator does not extend to three dimensions.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1981

Thermal IR exitance model of a plant canopy

A thermal IR exitance model of a plant canopy based on a mathematical abstraction of three horizontal layers of vegetation was developed. Canopy geometry within each layer is quantitatively described by the foliage and branch orientation distributions and number density. Given this geometric information for each layer and the driving meteorological variables, a system of energy budget equations was determined and solved for average layer temperatures. These estimated layer temperatures, together with the angular distributions of radiating elements, were used to calculate the emitted thermal IR radiation as a function of view angle above the canopy. The model was applied to a lodgepole pine (Pinus contorta) canopy over a diurnal cycle. Simulated vs measured radiometric average temperatures of the midcanopy layer corresponded with 2 C. Simulation results suggested that canopy geometry can significantly influence the effective radiant temperature recorded at varying sensor view angles.

Kimes, D. S.↗
NASA NTRS1972

Estimating Weibull parameters for composite materials.

This paper deals with the statistical analysis of strength and fracture of materials in general, with application to fiber composites. The 'weakest link' model is considered in a fairly general form, and the resulting equations are demonstrated by using a Weibull distribution for flaws. This distribution appears naturally in a variety of problems, and therefore additional attention is devoted to analysis and statistical estimation connected with this distribution. Special working charts are included to facilitate interpretation of observed data and estimation of parameters. Implications of the size effect are considered for various kinds of flaw distributions. The paper describes failure and damage in a fiber-reinforced systems.

Robinson, E. Y.↗
DOE OSTI2022

CAMERA: A method for cost-aware, adaptive, multifidelity, efficient reliability analysis

Estimating probability of failure in aerospace systems is a critical requirement for flight certification and qualification. Failure probability estimation involves resolving tails of probability distributions, and Monte Carlo sampling methods are intractable when expensive high-fidelity simulations have to be queried. Here, we propose a method to use models of multiple fidelities that trade accuracy for computational efficiency. Specifically, we propose the use of multifidelity Gaussian process models to efficiently fuse models at multiple fidelity, thereby offering a cheap surrogate model that emulates the original model at all fidelities. Furthermore, we propose a novel sequential acquisition function based experiment design framework that can automatically select samples from appropriate fidelity models to make predictions about quantities of interest at the highest fidelity. We use our proposed approach in an importance sampling setting and demonstrate our method on the failure level set and probability estimation on synthetic test functions and two real-world applications, namely, the reliability analysis of a gas turbine engine blade using a finite element method and a transonic aerodynamic wing test case using Reynolds-averaged Navier-Stokes equations. We show that our method predicts the failure boundary and probability more accurately and at a fraction of the computational cost compared with using just a single expensive high-fidelity model. Finally, we show that our sequential approach is guaranteed to asymptotically converge to the true failure boundary with high probability.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Utilizing the maximum likelihood estimator for flow analysis

We explore the possibility of evaluating flow harmonics by employing the maximum likelihood estimator (MLE). For a given finite multiplicity, the MLE simultaneously furnishes estimations for all the parameters of the underlying distribution function while efficiently suppressing the variance of measures. Also, the method provides a means to assess a specific class of mixed harmonics, which is not straightforwardly feasible by the approaches primarily based on particle correlations. The results are analyzed using the Wald, likelihood ratio, and score tests of hypotheses. Besides, the resultant flow harmonics obtained using MLE are compared with those derived using particle correlations and event plane methods. Here, the dependencies of extracted flow harmonics on the multiplicity of individual events and the total number of events are analyzed. It is shown that the proposed approach works efficiently to deal with the deficiency in detector acceptability. Moreover, we elaborate on a fictitious scenario where the event plane is not a well-defined quantity in the distribution function. For the latter case, the MLE is shown to largely perform better than the two-particle correlation estimator. In this regard, one concludes that the MLE furnishes a meaningful alternative to the existing approaches for flow analysis.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
NASA NTRS1989

Cloud field classification based on textural features

An essential component in global climate research is accurate cloud cover and type determination. Of the two approaches to texture-based classification (statistical and textural), only the former is effective in the classification of natural scenes such as land, ocean, and atmosphere. In the statistical approach that was adopted, parameters characterizing the stochastic properties of the spatial distribution of grey levels in an image are estimated and then used as features for cloud classification. Two types of textural measures were used. One is based on the distribution of the grey level difference vector (GLDV), and the other on a set of textural features derived from the MaxMin cooccurrence matrix (MMCM). The GLDV method looks at the difference D of grey levels at pixels separated by a horizontal distance d and computes several statistics based on this distribution. These are then used as features in subsequent classification. The MaxMin tectural features on the other hand are based on the MMCM, a matrix whose (I,J)th entry give the relative frequency of occurrences of the grey level pair (I,J) that are consecutive and thresholded local extremes separated by a given pixel distance d. Textural measures are then computed based on this matrix in much the same manner as is done in texture computation using the grey level cooccurrence matrix. The database consists of 37 cloud field scenes from LANDSAT imagery using a near IR visible channel. The classification algorithm used is the well known Stepwise Discriminant Analysis. The overall accuracy was estimated by the percentage or correct classifications in each case. It turns out that both types of classifiers, at their best combination of features, and at any given spatial resolution give approximately the same classification accuracy. A neural network based classifier with a feed forward architecture and a back propagation training algorithm is used to increase the classification accuracy, using these two classes of features. Preliminary results based on the GLDV textural features alone look promising.

Sengupta, Sailes Kumar↗
DOE OSTI2026

Active operator learning with predictive uncertainty quantification for partial differential equations

With the increased prevalence of neural operators being used to provide rapid solutions to partial differential equations (PDEs), understanding the accuracy of model predictions and the associated error levels is necessary for deploying reliable surrogate models in scientific applications. Existing uncertainty quantification (UQ) frameworks employ ensembles or Bayesian methods, which can incur substantial computational costs during both training and inference. Here, we propose a lightweight predictive UQ method tailored for Deep operator networks (DeepONets) that also generalizes to other operator networks. Numerical experiments on linear and nonlinear PDEs demonstrate that the framework’s uncertainty estimates are unbiased and provide accurate out-of-distribution uncertainty predictions with a sufficiently large training dataset. Our framework provides fast inference and uncertainty estimates that can efficiently drive outer-loop analyses that would be prohibitively expensive with conventional solvers. We demonstrate how predictive uncertainties can be used in the context of Bayesian optimization and active learning problems to yield improvements in accuracy and data-efficiency for outer-loop optimization procedures. In the active learning setup, we extend the framework to Fourier Neural Operators (FNO) and describe a generalized method for other operator networks. To enable real-time deployment, we introduce an inference strategy based on precomputed trunk outputs and a sparse placement matrix, reducing evaluation time by more than a factor of five. Our method provides a practical route to uncertainty-aware operator learning in time-sensitive settings.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Detecting Anomalies in Time Series Using Kernel Density Approaches

This paper introduces a novel anomaly detection approach tailored for time series data with exclusive reliance on normal events during training. Our key innovation lies in the application of kernel-density estimation (KDE) to scrutinize reconstruction errors, providing an empirically derived probability distribution for normal events post-reconstruction. This non-parametric density estimation technique offers a nuanced understanding of anomaly detection, differentiating it from prevalent threshold-based mechanisms in existing methodologies. In post-training, events are encoded, decoded, and evaluated against the estimated density, providing a comprehensive notion of normality. In addition, we propose a data augmentation strategy involving variational autoencoder-generated events and a smoothing step for enhanced model robustness. The significance of our autoencoder-based approach is evident in its capacity to learn normal representation without prior anomaly knowledge. Through the KDE step on reconstruction errors, our method addresses the versatility of anomalies, departing from assumptions tied to larger reconstruction errors for anomalous events. Our proposed likelihood measure then distinguishes normal from anomalous events, providing a concise yet comprehensive anomaly detection solution. The extensive experimental results support the feasibility of our proposed method, yielding significantly improved classification performance by nearly 10% on the UCR benchmark data.

Frehner, Robin↗
NASA NTRS2012

The Distribution of Snow Black Carbon observed in the Arctic and Compared to the GISS-PUCCINI Model

In this study, we evaluate the ability of the latest NASA GISS composition-climate model, GISS-E2- PUCCINI, to simulate the spatial distribution of snow BC (sBC) in the Arctic relative to present-day observations. Radiative forcing due to BC deposition onto Arctic snow and sea ice is also estimated. Two sets of model simulations are analyzed, where meteorology is linearly relaxed towards National Centers for Environmental Prediction (NCEP) and towards NASA Modern Era Reanalysis for Research and Applications (MERRA) reanalyses. Results indicate that the modeled concentrations of sBC are comparable with presentday observations in and around the Arctic Ocean, except for apparent underestimation at a few sites in the Russian Arctic. That said, the model has some biases in its simulated spatial distribution of BC deposition to the Arctic. The simulations from the two model runs are roughly equal, indicating that discrepancies between model and observations come from other sources. Underestimation of biomass burning emissions in Northern Eurasia may be the main cause of the low biases in the Russian Arctic. Comparisons of modeled aerosol BC (aBC) with long-term surface observations at Barrow, Alert, Zeppelin and Nord stations show significant underestimation in winter and spring concentrations in the Arctic (most significant in Alaska), although the simulated seasonality of aBC has been greatly improved relative to earlier model versions. This is consistent with simulated biases in vertical profiles of aBC, with underestimation in the lower and middle troposphere but overestimation in the upper troposphere and lower stratosphere, suggesting that the wet removal processes in the current model may be too weak or that vertical transport is too rapid, although the simulated BC lifetime seems reasonable. The combination of observations and modeling provides a comprehensive distribution of sBC over the Arctic. On the basis of this distribution, we estimate the decrease in snow and sea ice albedo and the resulting radiative forcing. We suggest that the albedo reduction due to BC deposition presents significant space-time variations, with highest mean reductions of 1.25% in the Russian Arctic, which are much larger than those in other Arctic regions (0.39% to 0.64 %). The averaged value over the Arctic north of 66degN is 0.4-0.6% during spring, leading to regional surface radiative forcings of 0.7, 1.1 and 1.0Wm(exp−2) in spring 2007, 2008 and 2009, respectively.

Dou, T.↗
NASA NTRS1986

Number-counts slope estimation in the presence of Poisson noise

The slope determination of a power-law number flux relationship in the case of photon-limited sampling. This case is important for high-sensitivity X-ray surveys with imaging telescopes, where the error in an individual source measurement depends on integrated flux and is Poisson, rather than Gaussian, distributed. A bias-free method of slope estimation is developed that takes into account the exact error distribution, the influence of background noise, and the effects of varying limiting sensitivities. It is shown that the resulting bias corrections are quite insensitive to the bias correction procedures applied, as long as only sources with signal-to-noise ratio five or greater are considered. However, if sources with signal-to-noise ratio five or less are included, the derived bias corrections depend sensitively on the shape of the error distribution.

Schmitt, Juergen H. M. M.↗
DOE OSTI2023

Quantifying Uncertainty in Ice Particle Velocity–Dimension Relationships Using MC3E Observations

Abstract Ice particle terminal fall velocity ( V t ) is fundamental for determining microphysical processes, yet remains extremely challenging to measure. Current theoretical best estimates of V t are functions of Reynolds number. The Reynolds number is related to the Best number, which is a function of ice particle mass, area ratio ( A r ), and maximum dimension ( D max ). These estimates are not conducive for use in most models since model parameterizations often take the form , where ( α , β ) depend on habit and D max . A previously developed framework is used to determine surfaces of equally plausible ( α , β ) coefficients whereby ice particle size/shape distributions are combined with V t best estimates to determine mass- ( V M ) or reflectivity-weighted ( V Z ) velocities that closely match parameterized V M ,SD or V Z ,SD calculated using the ( α , β ) coefficients using two approaches. The first uses surfaces of equally plausible ( a , b ) coefficients describing mass ( M )–dimension relationships (i.e., ) to calculate mass- or reflectivity-weighted velocity from size/shape distributions that are then used to determine ( α , β ) coefficients. The second investigates how uncertainties in A r , D max , and size distribution N ( D ) affect V M or V Z . For seven of nine flight legs flown on 20 and 23 May 2011 during the Mesoscale Continental Convective Clouds Experiment (MC3E), uncertainty from natural parameter variability—namely, the variability in ice particle parameters in similar meteorological conditions—exceeds uncertainties arising from different A r assumptions or D max estimates. The combined uncertainty between A r , D max , and N ( D ) produced smaller variability in ( α , β ) compared to varying M ( D ), demonstrating M ( D ) must be accurately quantified for model fall velocities. Primary sources of uncertainty vary considerably depending on environmental conditions. Significance Statement Ice particle fall velocity is fundamental for numerous processes within clouds, and hence is a critical property that must be accurately represented in weather and climate models. Using aircraft observations of ice particle shapes and sizes obtained in clouds behind midlatitude thunderstorms, this work develops a new framework for estimating ice particle fall velocities and their uncertainty, including quantifying the importance of different uncertainty sources from cloud microphysics measurements. Natural parameter variability contributes the most uncertainty in ice particle fall velocity estimates, although other sources can also be important contributors to uncertainty in certain conditions. Additional work examining ice particle data is needed to further understand how dependent uncertainty in certain ice particle properties are to local environmental conditions.

Meteorology & Atmospheric Sciences↗