Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “variance reduction”

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 127 records · Page 7
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Development of MCNP Training Modules for Safeguards Practitioners [Abstract]

The Monte-Carlo N-Particle (MCNP) software developed at LANL is the most widely used neutron transport code in the world. It is an essential tool for a variety of applications including detector development and design, nuclear fuel burnup simulation, criticality safety, and nondestructive assay system optimization. For this reason, it is indispensable within the safeguards and materials control & accountability (MC&A) communities. Multiple MCNP training courses have been created and taught over the last several decades by the MCNP development team at LANL, however there are no existing courses that cover specialized topics considered fundamental to NDA and safeguards models. To fill this gap, the MCNP team and Safeguards Science and Technology group at LANL have co-created a set of training modules customized to meet the specialized needs of the safeguards and MC&A communities. The basic modules cover concepts such as NDA system optimization, He-specific and other capture tallies, and tools for improved theoretical understanding. An advanced module was also created to cover topics including variance reduction for active interrogation simulations, use of the LANL MCNPTools post-processor, PTRAC (particle tracking) and list-mode data simulations, and fuel burnup simulations. The training modules teach to the latest and most state-of-the-art MCNP features and tools released by the development team at LANL and are intended to be taught jointly by the developers and safeguards experts. Ultimately, we hope that creation of these modules will serve to capture and convey the safeguards modeling and MCNP expertise at LANL, and that we will be able to share the modules more broadly with the MC&A and safeguards communities.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2025

MOOSE ProbML: Parallelizable Probabilistic Machine Learning and Uncertainty Quantification Capabilities

The Multiphysics Object Oriented Simulation Environment (MOOSE) is a widely used open- source finite element software for performing multiphysics multiscale simulations in a massively parallel fashion. Recently, the computational team at Idaho National Laboratory (INL) has implemented Probabilistic Machine Learning (ProbML) capabilities in MOOSE—in a parallelized fashion—and enable active learning with large-scale computational models for tasks such as surrogate model development, scale bridging, forward/inverse uncertainty quantification (UQ), Bayesian optimization, etc. This presentation summarizes these developments in MOOSE along with demonstrations on several real applications relevant to nuclear energy. At the fundamental level, samplers like Monte Carlo/Latin Hypercube, variance reduction, parallelized Markov Chain Monte Carlo (MCMC) support uncertainty propagation in both forward and inverse settings. These samplers can be integrated with the Gaussian processes (GP) suite in MOOSE, which offer several variants like scalar GPs, multi-output GPs, and deep GPs, to enable active learning. These GPs can be tuned using gradient-based optimization methods like Adam and its variants or gradient-free methods like the elliptical slice sampler (a variant of MCMC adept under Gaussian settings) for more complex covariance kernels or likelihoods whose gradient computations can be cumbersome. A variety of batch acquisition functions permit parallelized evaluation of the computational model and support different learning objectives with high efficiency like Bayesian inference, global surrogate development, optimization, etc. Furthermore, libtorch integration supports training, evaluation, and re-training of neural networks and other complex machine learning models in active learning settings. The impacts of these developments are shown on several real applications: (1) nuclear fuel inverse UQ and model inadequacy assessment using the Kennedy O’Hagan framework; (2) uncertainty aware surrogate modeling for additive manufacturing to predict field quantities; (3) nuclear reactor rare events analysis; and (4) complex fluid flow prediction using a global surrogate with quantified prediction uncertainty. Finally, the outlook of MOOSE ProbML is discussed for both outer-loop and inner-loop computations in the broad view to accelerate fuels and materials qualification, address gaps in knowledge and data, and assess new reactor/fuel systems.

11 - NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
NASA NTRS2000

Three Averaging Techniques for Reduction of Antenna Temperature Variance Measured by a Dicke Mode, C-Band Radiometer

As new radiometer technologies provide the possibility of greatly improved spatial resolution, their performance must also be evaluated in terms of expected sensitivity and absolute accuracy. As aperture size increases, the sensitivity of a Dicke mode radiometer can be maintained or improved by application of any or all of three digital averaging techniques: antenna data averaging with a greater than 50% antenna duty cycle, reference data averaging, and gain averaging. An experimental, noise-injection, benchtop radiometer at C-band showed a 68.5% reduction in Delta-T after all three averaging methods had been applied simultaneously. For any one antenna integration time, the optimum 34.8% reduction in Delta-T was realized by using an 83.3% antenna/reference duty cycle.

Mackenzie, Anne I.↗
NASA NTRS1991

On the effectiveness of Geosat altimeter corrections

Measurements of sea level by a satellite altimeter require a number of corrections for atmospheric and geophysical effects. The paper examines the effectiveness of a large suite of correction terms for Geosat by studying how each correction affects the variances of altimetric sea-surface height differences. Most corrections were found to provide a significant reduction in variance, the most important exception being certain short-period wet-troposphere correlations. The present procedure of using height differences implies nothing about the time-invariant parts of corrections.

Ray, R. D.↗
DOE OSTI2025

A Regularized Variance-Reduced Modified Extragradient Method for Stochastic Hierarchical Games

We consider an N -player hierarchical game in which the i th player’s objective comprises of an expectation-valued term, parametrized by rival decisions, and a hierarchical term. Such a framework allows for capturing a broad range of stochastic hierarchical optimization problems, Stackelberg equilibrium problems, and leader-follower games. We develop an iteratively regularized and smoothed variance-reduced modified extragradient framework for iteratively approaching hierarchical equilibria in a stochastic setting. We equip our analysis with rate statements, complexity guarantees, and almost-sure convergence results. We then extend these statements to settings where the lower-level problem is solved inexactly and provide the corresponding rate and complexity statements. Our model framework encompasses many game theoretic equilibrium problems studied in the context of power markets. We present a realistic application to the study of virtual power plants, emphasizing the role of hierarchical decision making and regularization. Preliminary numerics suggest that empirical behavior compares well with theoretical guarantees.

Tikhonov regularization↗
DOE OSTI2022

Accelerated statistical failure analysis of multifidelity TRISO fuel models

Statistical nuclear fuel failure analysis is critical for the design and development of advanced reactor technologies. Although Monte Carlo Sampling (MCS) is a standard method of statistical failure analysis for fuels, the low failure probabilities of some advanced fuel forms and the correspondingly large number of required model evaluations limit its application to low-fidelity (e.g., 1-D) fuel models. In this paper, we present four other statistical methods for fuel failure analysis in Bison, considering tri-structural isotropic (TRISO)-coated particle fuel as a case study. The statistical methods considered are Latin hypercube sampling (LHS), adaptive importance sampling (AIS), subset simulation (SS), and the Weibull theory. Using these methods, we analyzed both 1-D and 2-D representations of TRISO models to compute failure probabilities and the distributions of fuel properties that result in failures. The results of these methods compare well across all TRISO models considered. Overall, SS and the Weibull theory were deemed the most efficient, and can be applied to both 1-D and 2-D TRISO models to compute failure probabilities. Moreover, since SS also characterizes the distribution of parameters that cause TRISO failures, and can consider failure modes not described by the Weibull criterion, it may be preferred over the other methods. Finally, a discussion on the efficacy of different statistical methods of assessing nuclear fuel safety is provided.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2022

Reliability estimation of an advanced nuclear fuel using coupled active learning, multifidelity modeling, and subset simulation

Tristructural isotropic (TRISO)-coated particle fuel is a robust nuclear fuel and determining its reliability is critical for the success of advanced nuclear technologies. However, TRISO failure probabilities are small and the associated computational models are expensive. Here we used coupled active learning, multifidelity modeling, and subset simulation to estimate the failure probabilities of TRISO fuels using several 1D and 2D models. With multifidelity modeling, we replaced expensive high-fidelity (HF) model evaluations with information fusion from two low-fidelity (LF) models. For the 1D TRISO models, we considered three multifidelity modeling strategies: only Kriging, Kriging LF prediction plus Kriging correction, and deep neural network (DNN) LF prediction plus Kriging correction. While the results across these multifidelity modeling strategies compared satisfactorily, strategies employing information fusion from two LF models called the HF model least often. Next, for the 2D TRISO model, we considered two multifidelity modeling strategies: DNN LF prediction plus Kriging correction (data-driven) and 1D TRISO LF prediction plus Kriging correction (physics-based). The physics-based strategy, as expected, consistently required the fewest calls to the HF model. However, the data-driven strategy had a lower overall simulation time since the DNN predictions are instantaneous, and the 1D TRISO model requires a non-negligible simulation time.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2023

General Multifidelity Surrogate Models: Framework and Active-Learning Strategies for Efficient Rare Event Simulation

Estimating the probability of failure for complex real-world systems using high-fidelity computational models is often prohibitively expensive, especially when the probability is small. Exploiting low-fidelity models can make this process more feasible, but merging information from multiple low-fidelity and high-fidelity models poses several challenges. Here, this paper presents a robust multi-fidelity surrogate modeling strategy in which the multi-fidelity surrogate is assembled using an active learning strategy using an on-the-fly model adequacy assessment set within a subset simulation framework for efficient reliability analysis. The multi-fidelity surrogate is assembled by first applying a Gaussian process correction to each low-fidelity model and assigning a model probability based on the model's local predictive accuracy and cost. Three strategies are proposed to fuse these individual surrogates into an overall surrogate model based on model averaging and deterministic/stochastic model selection. The strategies also dictate which model evaluations are necessary. No assumptions are made about the relationships between low-fidelity models, while the high-fidelity model is assumed to be the most accurate and most computationally expensive model. Through two analytical and two numerical case studies, including a case study evaluating the failure probability of Tristructural isotropic-coated (TRISO) nuclear fuels, the algorithm is shown to be highly accurate while drastically reducing the number of high-fidelity model calls (and hence computational cost).

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
NASA NTRS1991

On the sea-state bias of the Geosat altimeter

The sea-state bias in a satellite altimeter's range measurement is caused by the influence of ocean waves on the radar return pulse; it results in an estimate of sea level that is too low according to some function of the wave height. This bias is here estimated for Geosat by correlating collinear differences of altimetric sea-surface heights with collinear differences of significant wave heights (H1/3). Corrections for satellite orbit error are estimated simultaneously with the sea-state bias. Based on twenty 17-day repeat cycles of the Geosat Exact Repeat Mission, the solution for the sea-state bias is 2.6 + or - 0.2 percent of H1/3. The least-squares residuals, however, show a correlation with wind speed U, so the traditional model of the bias has been supplemented with a second term: H1/3 + alpha-2H1/3U. This second term produces a small, but statistically significant, reduction in variance of the residuals. Both systematic and random errors in H1/3 and U tend to bias the estimates of alpha-1 and alpha-2, which complicates comparisons of the results with ground-based measurements of the sea-state bias.

Ray, Richard D.↗
NASA NTRS1994

Correlated environmental corrections in TOPEX/POSEIDON, with a note on ionospheric accuracy

Estimates of the effectiveness of an altimetric correction, and interpretation of sea level variability as a response to atmospheric forcing, both depend upon assuming that residual errors in altimetric corrections are uncorrelated among themselves and with residual sea level, or knowing the correlations. Not surprisingly, many corrections are highly correlated since they involve atmospheric properties and the ocean surface's response to them. The full corrections (including their geographically varying time mean values), show correlations between electromagnetic bias (mostly the height of wind waves) and either atmospheric pressure or water vapor of -40%, and between atmospheric pressure and water vapor of 28%. In the more commonly used collinear differences (after removal of the geographically varying time mean), atmospheric pressure and wave height show a -30% correlation, atmospheric pressure and water vapor a -10% correlation, both pressure and water vapor a 7% correlation with residual sea level, and a bit surprisingly, ionospheric electron content and wave height a 15% correlation. Only the ocean tide is totally uncorrelated with other corrections or residual sea level. The effectiveness of three ionospheric corrections (TOPEX dual-frequency, a smoothed version of the TOPEX dual-frequency, and Doppler orbitography and radiopositioning integrated by satellite (DORIS) is also evaluated in terms of their reduction in variance of residual sea level. Smooth (90-200 km along-track) versions of the dual-frequency altimeter ionosphere perform best both globally and within 20 deg in latitude from the equator. The noise variance in the 1/s TOPEX inospheric samples is approximately (11 mm) squared, about the same as noise in the DORIS-based correction; however, the latter has its error over scales of order 10(exp 3) km. Within 20 deg of the equator, the DORIS-based correction adds (14 mm) squared to the residual sea level variance.

Zlotnicki, V.↗
NASA NTRS1998

Evidence for Solar-Cycle Forcing and Secular Variation in the Armagh Observatory Temperature Record

A prominent feature of previous long-term temperature studies has been the appearance of warming since the 1880s, this often being taken as evidence for anthropogenic-induced global warming. In this investigation, the long-term, annual, mean temperature record (1844-1992) of the Armagh Observatory (Armagh, North Ireland), a set of temperature data based on maximum and minimum thermometers that predates the 1880s and correlates well with northern hemispheric and global standards, is examined for evidence of systematic variation, in particular, as related to solar-cycle forcing and secular variation. Indeed, both appear to be embedded within the Armagh data. Removal of these effects, each contributing about 8% to the overall reduction in variance, yields residuals that are randomly distributed. Application of the 10-year moving average to the residuals, furthermore, strongly suggests that the behavior of the residuals is episodic, inferring that (for extended periods of time) temperatures at Armagh sometimes were warmer or cooler (than expected), while at other times they were stable. Comparison of cyclic averages of annual mean temperatures against the lengths of the associated Hale cycles (i.e., the length of two, sequentially numbered, even-odd sunspot cycle pairs) strongly suggests that the temperatures correlate inversely (r = -0.886 at less than 2% level of significance) against the length of the associated Hale cycle. Because sunspot cycle 22 ended in 1996, the present Hale cycle probably will be shorter than average, implying that temperatures at Armagh over this Hale cycle will be warmer (about 9.31 q 0.23 C at the 90% confidence level) than average (= 9.00 C).

Wilson, Robert M.↗
NASA NTRS2015

M2 Internal Tides and Their Observed Wavenumber Spectra from Satellite Altimetry*

A near-global chart of surface elevations associated with the stationary M2 internal tide is empirically constructed from multi-mission satellite altimeter data. An advantage of a strictly empirical mapping approach is that results are independent of assumptions about ocean wave dynamics and, in fact, can be used to test such assumptions. A disadvantage is that present-day altimeter coverage is only marginally adequate to support mapping such short-wavelength features. Moreover, predominantly north-south ground-track orientations and contamination from nontidal oceanographic variability can lead to deficiencies in mapped tides. Independent data from Cryosphere Satellite-2 (CryoSat-2) and other altimeters are used to test the solutions and show positive reduction in variance except in regions of large mesoscale variability. The tidal fields are subjected to two-dimensional wavenumber spectral analysis, which allows for the construction of an empirical map of modal wavelengths. Mode-1 wavelengths show good agreement with theoretical wavelengths calculated from the ocean's mean stratification, with a few localized exceptions (e.g., Tasman Sea). Mode-2 waves are detectable in much of the ocean, with wavelengths in reasonable agreement with theoretical expectations, but their spectral signatures grow too weak to map in some regions.

Ray, R. D.↗
NASA NTRS1979

Large Area Crop Inventory Experiment (LACIE). Evaluation of three-category classification

The author has identified the following signficant results. Examination of both machine estimates and stratified areal estimates produced by clustering and classification reveal no significant differences between the proportion estimates and ground truth estimates. When testing the variances of the machine estimates, a significant reduction in the variances was found when the number of starting dots was increased from 30 to 45. The variances were again reduced, although not significantly, when the number of starting dots was increased from 45 to 60. From these results, 60 starting dots are recommended for a three-category classifier. When examining the variances of the estimates for the four estimation procedures (using 60 dots), no significant differences were found between procedures. Thus, only the machine clustering may be used to produce an estimate, and the stratified areal estimate computations and maximum likelihood classification can be deleted.

Havens, K. A.↗
NASA NTRS1983

Trajectory-Estimation Error Analysis

Orbital and Geodetic Parameter Estimation Error Analysis Program (ORAN) is Bayesian least-squares simulation program for orbital trajectories. ORAN does not process data; rather, computes accuracy of results of data reduction if measurements are processed by minimum-variance data-reduction program. Actual data used to provide time when given measurement was available and estimated noise on that measurement.

Putney, B.↗
DOE OSTI2024

Optimization of foreground moment deprojection for semi-blind CMB polarization reconstruction

Abstract Upcoming Cosmic Microwave Background (CMB) experiments, aimed at measuring primordial CMB polarization B-modes, require exquisite control of instrumental systematics and Galactic foreground contamination. Blind minimum-variance techniques, like the Needlet Internal Linear Combination (NILC), have proven effective in reconstructing the CMB polarization signal and mitigating foregrounds and systematics across diverse sky models without suffering from foreground mismodelling errors. Still, residual foreground contamination from NILC may bias the recovered CMB polarization at large angular scales when confronted with the most complex foreground scenarios.By adding constraints to NILC to deproject statistical moments of the Galactic emission, the Constrained Moment ILC (cMILC) method has been demonstrated to further enhance foreground subtraction, albeit with an associated increase in overall noise variance. Faced with this trade-off between foreground bias reduction and overall variance minimization, there is still no recipe on which moments to deproject and which are better suited for blind variance minimization. To address this, we introduce the optimized cMILC (ocMILC) pipeline, which performs full automated optimization of the required number and set of foreground moments to deproject, pivot parameter values, and deprojection coefficients across the sky and angular scales, depending on the actual sky complexity, available frequency coverage, and experiment sensitivity. The optimal number of moments for deprojection, before paying significant noise penalty, is determined through a data diagnosis inspired by the Generalized NILC (GNILC) method.Validated on B-mode simulations of thePICOspace mission concept with four challenging foreground models, ocMILC exhibits lower Galactic foreground contamination compared to NILC and cMILC at all angular scales, with limited noise penalty. This multi-layer optimization enables the ocMILC pipeline to achieve unbiased posteriors of the tensor-to-scalar ratio, regardless of foreground complexity.

Astronomy & Astrophysics↗
DOE OSTI2023

Mitigating Green's function Monte Carlo signal-to-noise problems using contour deformations

The Green's function Monte Carlo (GFMC) method provides accurate solutions to the nuclear many-body problem and predicts properties of light nuclei starting from realistic two- and three-body interactions. Controlling the GFMC fermion-sign problem is crucial, as the signal-to-noise ratio decreases exponentially with Euclidean time, requiring significant computing resources. Inspired by similar scenarios in lattice quantum field theory and spin systems, in this work, we employ integration contour deformations to improve the GFMC signal-to-noise ratio. Machine learning techniques are used to select optimal contours with minimal variance from parameterized families of deformations. As a proof of principle, we consider the deuteron binding energies and Euclidean density response functions. We only observe mild signal-to-noise improvement for the binding energy case. On the other hand, we achieve an order of magnitude reduction of the variance for Euclidean density response functions, paving the way for computing electron- and neutrino-nucleus cross-sections of larger nuclei.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Mitigating Green's function Monte Carlo signal-to-noise problems using contour deformations

The Green's function Monte Carlo (GFMC) method provides accurate solutions to the nuclear many-body problem and predicts properties of light nuclei starting from realistic two- and three-body interactions. Controlling the GFMC fermion sign problem is crucial, as the signal-to-noise ratio decreases exponentially with imaginary time, requiring significant computing resources. Inspired by similar scenarios in lattice quantum field theory and spin systems, in this work, we employ integration contour deformations to improve the GFMC signal-to-noise ratio. Machine learning techniques are used to select optimal contours with minimal variance from parametrized families of deformations. As a proof of principle, we consider the deuteron binding energies and Euclidean density response functions. Here, we only observe mild signal-to-noise improvement for the binding energy case. On the other hand, we achieve an order of magnitude reduction of the variance for Euclidean density response functions, paving the way for computing electron- and neutrino-nucleus cross sections of larger nuclei.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
NASA NTRS1987

Discrete filtering techniques applied to sequential GPS range measurements

The basic navigation solution is described for position and velocity based on range and delta range (Doppler) measurements from NAVSTAR Global Positioning System satellites. The application of discrete filtering techniques is examined to reduce the white noise distortions on the sequential range measurements. A second order (position and velocity states) Kalman filter is implemented to obtain smoothed estimates of range by filtering the dynamics of the signal from each satellite separately. Test results using a simulated GPS receiver show a steady-state noise reduction, the input noise variance divided by the output noise variance, of a factor of four. Recommendations for further noise reduction based on higher order Kalman filters or additional delta range measurements are included.

Vangraas, Frank↗