SEARCH · Engineering Papers

Results for “model weights”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

SourceResearch connectionFrom yearThrough year

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

At least 289 records · Page 16

All sources NASA NTRS DOE OSTI

DOE OSTI2020

CSB-RNN: A Faster-Than-Realtime RNN Acceleration Framework with Compressed Structured Blocks

Recurrent Neural Networks (RNN) is widely applied to temporal sequence analysis, where real-time performance is usually in demand. However, RNN suffers a heavy computational workload as the model comes with a large weight matrix. To alleviate the pain, model compression (pruning) schemes have been proposed for RNN that pruning the redundant (near-zero) weight-values. On the one hand, the non-structured pruning methods achieve a considerable pruning rate while bringing the computational irregularity, which is un-friendly to parallel-hardware. On the other hand, the existing structured pruning methods consider the hardware parallelism; However, they suffer a poor pruning rate due to the restrict constraints on pruning structure. This paper presents CSB-RNN, an optimized full-stack RNN framework with the novel compressed structured block (CSB) technique. The CSB-pruned RNN model comes with both fine-granularity that benefits the pruning rate and regular structure that facilitates the hardware-parallelism. Further, we propose a novel hardware architecture for inferencing the CSB-pruned model. Different from conventional parallel hardware, this architecture solves the block-workload imbalance issue and achieves an over 95% hardware utilization. With the experiments on 10 RNN models in 5 application domains, the CSB-RNN realizes 7×-20× lossless compression and up to 50× acceptable lossy-compression, which is 2×-7× to the prior art. With the addition of the novel hardware, the compressed-RNN inference reaches a super real-time latency of 10-400µs with FPGA implementation.

Shi, Runbin↗

NASA NTRS2023

Marin County Wildland Fires: Examining Fuel Load and Land Cover Change to Inform Fire Prevention and Suppression Decisions in Marin County, CA

Heightened occurrence of severe wildfires in the Western United States is increasing the need to better understand regions of high potential wildfire severity and develop methodologies for identifying the best locations for fuels reduction and active wildfire suppression, especially in populated regions such as Marin County, California. Marin County, located in the San Francisco Bay Area, has had significant development in the wildland-urban interface and periods of highly wildfire-prone conditions. The NASA DEVELOP team collaborated with Fire Foundry (a Marin-based fire service workforce development program) and the Marin County Fire Department to develop models to assist with fire management. Using data from Sentinel-2A, PlanetScope, ECOSTRESS, a county-wide LiDAR mapping effort, Landsat 7 Enhanced Thematic Mapper (ETM+), and Landsat 8 Operational Land Imager (OLI), our team developed a number of input data layers for three different models to evaluate wildfire severity. One model performed a suitability analysis with weights based on scientific literature; another model utilized a U-Net Convolutional Neural Network trained on previous fires in Marin and neighboring Sonoma County to predict the difference normalized burn severity; and the third inputted data layers into the FlamMap tool that outputs risk categories. We compared model outputs and performed a weighted overlay analysis to identify specific locations where a fireline could be constructed to interrupt the progress of an active fire. These tools will assist partners in preparing for and managing active wildfire situations.

Remote sensing↗

DOE OSTI2025

Neural refinement of sample weights

Monte Carlo simulations are an essential tool in particle physics data analysis. Events are typically generated alongside weights that redistribute the cross section of the simulated process across the phase space. These weights can be negative, and several post hoc methods have been developed to eliminate or mitigate the negative values. All of these methods share the common strategy of approximating the average weight as a function of phase space. We introduce an alternative approach, which, instead of reweighting to the average, refines the initial weights with a scaling transformation, utilizing a phase space-dependent factor. Since this new refinement method does not need to model the full weight distribution, it can be more accurate. High-dimensional and unbinned phase space is processed using neural networks for the refinement method. In addition to the refinement method, we introduce a new resampling protocol, which can be used in conjunction with any weight transformation to not only preserve the average weight but also the statistical uncertainties of the initial distribution. Using both realistic and synthetic examples, we show that the new neural refinement method is able to match or exceed the accuracy of similar weight transformations and that the new resampling protocol is simpler in implementation than previous methods while exhibiting equivalent statistical properties.

Artificial neural networks↗

NASA NTRS2014

Propulsion System Models for Rotorcraft Conceptual Design

The conceptual design code NDARC (NASA Design and Analysis of Rotorcraft) was initially implemented to model conventional rotorcraft propulsion systems, consisting of turboshaft engines burning jet fuel, connected to one or more rotors through a mechanical transmission. The NDARC propulsion system representation has been extended to cover additional propulsion concepts, including electric motors and generators, rotor reaction drive, turbojet and turbofan engines, fuel cells and solar cells, batteries, and fuel (energy) used without weight change. The paper describes these propulsion system components, the architecture of their implementation in NDARC, and the form of the models for performance and weight. Requirements are defined for improved performance and weight models of the new propulsion system components. With these new propulsion models, NDARC can be used to develop environmentally-friendly rotorcraft designs.

Propulsion System Models↗

DOE OSTI2022

Forecasting distributed energy resources adoption for power systems

Failing to incorporate accurate distributed energy resource penetration forecasts into long-term resource and transmission planning can lead to cost inefficiencies at best and system failures at worst. We have developed an open-source tool that employs an advanced Bass specification to calibrate and forecast technology adoption. The advanced specification includes geographic clustering, exogenously estimated market size, and dynamic time steps. Training on historical adoption of rooftop photovoltaics at the U.S. county-level and using detailed techno-economic estimates, our model achieves a two-year average mean-absolute-percentage-error of 19% in predicting system counts at the county-level, weighted by population. Model error was negatively correlated with market maturity - the error was 12% for counties in states with at least 28 W-per-capita of installed capacity. The advanced specification significantly reduces unweighted forecasting percent error compared to a conventional Bass specification: from 196% to 25% for capacity and from 226% to 22% for system count.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

NASA NTRS1999

Optimal System Realization in Frequency Domain

Several approaches are presented to identify an experimental system model directly from frequency response data. The formulation begins with a matrix-fraction description as the model structure. Frequency weighting such as exponential weighting is introduced to solve a weighted least-squares problem to obtain the coefficient matrices for the matrix-fraction description. A multi-variable state-space model can then be formed using the coefficient matrices of the matrix-fraction description. An approach is introduced to fine-tune the model using non-linear programming methods to minimize the desired cost function. The method deals with the model in the real Schur or modal form and reassigns a subset of system poles using a nonlinear optimizer. At every optimization step, the input and output influence matrices are refined through least-squares procedures. The proposed approaches are used to identify an analytical model for a NASA testbed from experimental data.

Juang, Jer-Nan↗

DOE OSTI2020

HYDRODYNAMIC ANALYSIS OF MACROALGAE LOCAL MODEL USING COMPUTATIONAL FLUID DYNAMICS

In this article, a local scale, fully nonlinear coupled fluid-structural interaction (FSI) sugar kelp model has been developed using a computational fluid dynamics (CFD) method. In this model, to be consistent with available experimental data, the sugar kelp is approximated as elongated rectangles with smoothed isosceles triangles at the ends and a single kelp model with one end fixed in a channel with constant current model is developed. Several different current speeds are simulated, and the resulting drag forces and calculated drag coefficients are validated by comparison with experimental data from the literature. In a previous study, a global scale model was developed using a computational structural dynamics (CSD) method to simulate macroalgae farming system and guide the system configuration design. In the global scale model, the hydrodynamic forces are calculated using Morison’s equation and the kinematics and dynamics of the sugar kelp are simplified and the group of kelps attached to the long line is modeled as a slender structure with the same length and an effective diameter such that the volumes are consistent with the real physical system. This simplified model matches the weight and buoyancy but adjusting the hydrodynamic properties when the general hydrodynamic coefficients are employed. Therefore, optimal hydrodynamic coefficients used in global scale model were determined to obtain the hydrodynamic force more accurately. The validated local scale model is then be applied to determine the hydrodynamic coefficients of the simplified sugar kelp model for global dynamic analysis.

hydrodynamic modeling, macroalgae, CFD model↗

DOE OSTI2017

On the Bayesian calibration of computer model mixtures through experimental data, and the design of predictive models

For many real systems, several computer models may exist with different physics and predictive abilities. To achieve more accurate simulations/predictions, it is desirable for these models to be properly combined and calibrated. Here, we propose the Bayesian calibration of computer model mixture method which relies on the idea of representing the real system output as a mixture of the available computer model outputs with unknown input dependent weight functions. The method builds a fully Bayesian predictive model as an emulator for the real system output by combining, weighting, and calibrating the available models in the Bayesian framework. Moreover, it fits a mixture of calibrated computer models that can be used by the domain scientist as a mean to combine the available computer models, in a flexible and principled manner, and perform reliable simulations. It can address realistic cases where one model may be more accurate than the others at different input values because the mixture weights, indicating the contribution of each model, are functions of the input. Inference on the calibration parameters can consider multiple computer models associated with different physics. The method does not require knowledge of the fidelity order of the models. We provide a technique able to mitigate the computational overhead due to the consideration of multiple computer models that is suitable to the mixture model framework. We implement the proposed method in a real-world application involving the Weather Research and Forecasting large-scale climate model.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

Urban morphology from a landscape perspective: How building morphology distribution land models (BMDLM) emulate pattern and process

Urban form (e.g., building morphology such as height or footprint) can be used to predict environmental footprints, such as energy/water consumption and carbon emissions. Although progress has been made in predicting building characteristics to fill gaps in observation or derive 3-D representations, the relationships between morphology and other variables such as land use and population are poorly understood. Understanding these relationships may enable projections for how cities will evolve with landscapes in the future. A suite of random forest models, the Building Morphology Distribution Land Models (BMDLM), was developed to determine how well building morphology for two distinct statistical measures (central tendency and frequency) can be predicted using land use (e.g., zoning) and population at different resolutions. Clark County, Nevada and Los Angeles County, California are explored as case studies. Generally, 1-km models outperformed 30-m models. Frequency distribution models had the best performance, especially in LA County. Frequency models significantly outperformed spatial autocorrelative models using inverse distance weighting (IDW). BMDLM offers a new take on modeling urban form in which generalized landscape patterns are characterized to understand the influence of population and zoning on urban development, as described by urban scaling theory.

Sturtevant, Jillian [Baylor Univ., Waco, TX (Unite↗

DOE OSTI2012

A BAYESIAN APPROACH TO LOCATING THE RED GIANT BRANCH TIP MAGNITUDE. II. DISTANCES TO THE SATELLITES OF M31

In 'A Bayesian Approach to Locating the Red Giant Branch Tip Magnitude (Part I)', a new technique was introduced for obtaining distances using the tip of the red giant branch (TRGB) standard candle. Here we describe a useful complement to the technique with the potential to further reduce the uncertainty in our distance measurements by incorporating a matched-filter weighting scheme into the model likelihood calculations. In this scheme, stars are weighted according to their probability of being true object members. We then re-test our modified algorithm using random-realization artificial data to verify the validity of the generated posterior probability distributions (PPDs) and proceed to apply the algorithm to the satellite system of M31, culminating in a three-dimensional view of the system. Further to the distributions thus obtained, we apply a satellite-specific prior on the satellite distances to weight the resulting distance posterior distributions, based on the halo density profile. Thus in a single publication, using a single method, a comprehensive coverage of the distances to the companion galaxies of M31 is presented, encompassing the dwarf spheroidals Andromedas I-III, V, IX-XXVII, and XXX along with NGC 147, NGC 185, M33, and M31 itself. Of these, the distances to Andromedas XXIV-XXVII and Andromeda XXX have never before been derived using the TRGB. Object distances are determined from high-resolution tip magnitude posterior distributions generated using the Markov Chain Monte Carlo technique and associated sampling of these distributions to take into account uncertainties in foreground extinction and the absolute magnitude of the TRGB as well as photometric errors. The distance PPDs obtained for each object both with and without the aforementioned prior are made available to the reader in tabular form. The large object coverage takes advantage of the unprecedented size and photometric depth of the Pan-Andromeda Archaeological Survey. Finally, a preliminary investigation into the satellite density distribution within the halo is made using the obtained distance distributions. For simplicity, this investigation assumes a single power law for the density as a function of radius, with the slope of this power law examined for several subsets of the entire satellite sample.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2021

Bayesian model averaging for analysis of lattice field theory results

Statistical modeling is a key component in the extraction of physical results from lattice field theory calculations. Although the general models used are often strongly motivated by physics, many model variations can frequently be considered for the same lattice data. Model averaging, which amounts to a probability-weighted average over all model variations, can incorporate systematic errors associated with model choice without being overly conservative. We discuss the framework of model averaging from the perspective of Bayesian statistics, and give useful formulae and approximations for the particular case of least-squares fitting, commonly used in modeling lattice results. In addition, we frame the common problem of data subset selection (e.g. choice of minimum and maximum time separation for fitting a two-point correlation function) as a model selection problem and study model averaging as a straightforward alternative to manual selection of fit ranges. Numerical examples involving both mock and real lattice data are given.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2015

Automated contouring error detection based on supervised geometric attribute distribution models for radiation therapy: A general strategy

Purpose: One of the most critical steps in radiation therapy treatment is accurate tumor and critical organ-at-risk (OAR) contouring. Both manual and automated contouring processes are prone to errors and to a large degree of inter- and intraobserver variability. These are often due to the limitations of imaging techniques in visualizing human anatomy as well as to inherent anatomical variability among individuals. Physicians/physicists have to reverify all the radiation therapy contours of every patient before using them for treatment planning, which is tedious, laborious, and still not an error-free process. In this study, the authors developed a general strategy based on novel geometric attribute distribution (GAD) models to automatically detect radiation therapy OAR contouring errors and facilitate the current clinical workflow. Methods: Considering the radiation therapy structures’ geometric attributes (centroid, volume, and shape), the spatial relationship of neighboring structures, as well as anatomical similarity of individual contours among patients, the authors established GAD models to characterize the interstructural centroid and volume variations, and the intrastructural shape variations of each individual structure. The GAD models are scalable and deformable, and constrained by their respective principal attribute variations calculated from training sets with verified OAR contours. A new iterative weighted GAD model-fitting algorithm was developed for contouring error detection. Receiver operating characteristic (ROC) analysis was employed in a unique way to optimize the model parameters to satisfy clinical requirements. A total of forty-four head-and-neck patient cases, each of which includes nine critical OAR contours, were utilized to demonstrate the proposed strategy. Twenty-nine out of these forty-four patient cases were utilized to train the inter- and intrastructural GAD models. These training data and the remaining fifteen testing data sets were separately employed to test the effectiveness of the proposed contouring error detection strategy. Results: An evaluation tool was implemented to illustrate how the proposed strategy automatically detects the radiation therapy contouring errors for a given patient and provides 3D graphical visualization of error detection results as well. The contouring error detection results were achieved with an average sensitivity of 0.954/0.906 and an average specificity of 0.901/0.909 on the centroid/volume related contouring errors of all the tested samples. As for the detection results on structural shape related contouring errors, an average sensitivity of 0.816 and an average specificity of 0.94 on all the tested samples were obtained. The promising results indicated the feasibility of the proposed strategy for the detection of contouring errors with low false detection rate. Conclusions: The proposed strategy can reliably identify contouring errors based upon inter- and intrastructural constraints derived from clinically approved contours. It holds great potential for improving the radiation therapy workflow. ROC and box plot analyses allow for analytically tuning of the system parameters to satisfy clinical requirements. Future work will focus on the improvement of strategy reliability by utilizing more training sets and additional geometric attribute constraints.

60 APPLIED LIFE SCIENCES↗

DOE OSTI2021

Population balance modeling of polyurethane foam formation with pressure‐dependent growth kernel

Abstract Polyurethane foams are widely used materials often chosen for their useful characteristics such as low thermal conductivity, ease of application, and high strength‐to‐weight ratios. Computational models are needed to predict the dynamics of the flow and expansion, and the resulting material properties, to improve manufacturing processes. In this paper, a model for PMDI, a water‐blown polyurethane foam, is presented. By extending a kinetics‐based approach by adding bubble‐scale information via a population balance equation (PBE) using the quadrature method of moments, we can track bubble size distributions during foaming. We present results from a three‐dimensional computational fluid dynamics model using arbitrary Lagrangian–Eulerian interface tracking implemented in finite element software. The model compares favorably with experimental data, including dynamics, bubble distributions measured by both camera and diffusion wave spectroscopy, and post‐test bubble size from scanning electron microscopy and density measurements from x‐ray computed tomography.

Ortiz, Weston↗

DOE OSTI2020

Chaos on the hypercube

We analyze the spectral properties of a d-dimensional HyperCubic (HC) lattice model originally introduced by Parisi. The U(1) gauge links of this model give rise to a magnetic flux of constant magnitude φ but random orientation through the faces of the hypercube. The HC model, which also can be written as a model of 2d interacting Majorana fermions, has a spectral flow that is reminiscent of Maldacena-Qi (MQ) model, and its spectrum at φ = 0, actually coincides with the coupling term of the MQ model. As was already shown by Parisi, at leading order in 1/d, the spectral density of this model is given by the density function of the Q-Hermite polynomials, which is also the spectral density of the double-scaled Sachdev-Ye-Kitaev model. Parisi demonstrated this by mapping the moments of the HC model to Q-weighted sums on chord diagrams. We point out that the subleading moments of the HC model can also be mapped to weighted sums on chord diagrams, in a manner that descends from the leading moments. The HC model has a magnetic inversion symmetry that depends on both the magnitude and the orientation of the magnetic flux through the faces of the hypercube. The spectrum for fixed quantum number of this symmetry exhibits a transition from regular spectra at φ = 0 to chaotic spectra with spectral statistics given by the Gaussian Unitary Ensembles (GUE) for larger values of φ. For small magnetic flux, the ground state is gapped and is close to a Thermofield Double (TFD) state.

1/N expansion↗

NASA NTRS1986

Preparation of satellite soundings for insertion in a variational objective analysis model

The performance of a variational analysis model being developed at the Illinois State Water Survey, which allows blending of satellite and conventional soundings in a way which preserves the information content of both data sources, requires preparation of the input data before insertion into the model. Using operationally retrieved Tiros-N TOVS soundings and rawinsonde data, four procedures to prepare satellite soundings for model insertion are demonstrated: (1) determination and removal of biases, (2) determination of standard errors of the satellite temperatures to assist in the assignment of weights in the model, (3) conversion of layer mean temperatures to level temperatures, and (4) determination of the action necessary for correcting for the nonsynoptic nature of the satellite data.

Kidder, Stanley Q.↗

NASA NTRS2006

Application of the Convolution Formalism to the Ocean Tide Potential: Results from the Gravity and Recovery and Climate Experiment (GRACE)

A computationally efficient approach to reducing omission errors in ocean tide potential models is derived and evaluated using data from the Gravity Recovery and Climate Experiment (GRACE) mission. Ocean tide height models are usually explicitly available at a few frequencies, and a smooth unit response is assumed to infer the response across the tidal spectrum. The convolution formalism of Munk and Cartwright (1966) models this response function with a Fourier series. This allows the total ocean tide height, and therefore the total ocean tide potential, to be modeled as a weighted sum of past, present, and future values of the tide-generating potential. Previous applications of the convolution formalism have usually been limited to tide height models, but we extend it to ocean tide potential models. We use luni-solar ephemerides to derive the required tide-generating potential so that the complete spectrum of the ocean tide potential is efficiently represented. In contrast, the traditionally adopted harmonic model of the ocean tide potential requires the explicit sum of the contributions from individual tidal frequencies. It is therefore subject to omission errors from neglected frequencies and is computationally more intensive. Intersatellite range rate data from the GRACE mission are used to compare convolution and harmonic models of the ocean tide potential. The monthly range rate residual variance is smaller by 4-5%, and the daily residual variance is smaller by as much as 15% when using the convolution model than when using a harmonic model that is defined by twice the number of parameters.

models↗

DOE OSTI2025

Magnon damping and mode softening in quantum double-exchange ferromagnets

We present a comprehensive analysis of the magnetic excitations and electronic properties of fully quantum double-exchange ferromagnets, i.e. systems where ferromagnetic (FM) ordering emerges from the competition between spin, charge, and orbital degrees of freedom, but without the canonical approximation of using classical localized spins. Specifically, we investigate spin excitations within the Kondo lattice-like model, as well as a two-orbital Hubbard Hamiltonian in proximity to the orbital-selective Mott phase. Computational analysis of the magnon dispersion, damping, and spectral weight within these models reveals unexpected phenomena, such as magnon mode softening and the anomalous decoherence of magnetic excitations as observed in earlier experimental efforts, but explained here without the use of the phononic degrees of freedom. We show that these effects are intrinsically linked to incoherent spectral features near the Fermi level, which arise due to the quantum nature of the local (on-site) triplets. This incoherent spectrum leads to a Stoner-like continuum on which spin excitations scatter, governing magnon lifetime and strongly influencing the dynamical spin structure factor. Our study explores the transition from coherent to incoherent magnon spectra by varying the electron density. Furthermore, we demonstrate that the magnitude of the localized spin mitigates decoherence by suppressing the incoherent spectral contributions near the Fermi level. We also discuss the effective J 1 –J 2 spin Hamiltonian, which can accurately describe the large doping region characterized by the magnon-mode softening. Finally, we show that this behavior is also present in multiorbital models with partially filled orbitals, namely, in systems without localized spin moments, provided that the model is in a strong coupling regime. Our results potentially have far-reaching implications for understanding FM ordering in various multi-band systems. These findings establish a previously unknown direct connection between the electronic correlations of those materials and spin excitations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2023

Constraining quenching time-scales in galaxy clusters by forward-modelling stellar ages and quiescent fractions in projected phase space

We forward-model mass-weighted stellar ages (MWAs) and quiescent fractions (f Q ) in projected phase space (PPS), using data from the Sloan Digital Sky Survey, to jointly constrain an infall quenching model for galaxies in log (M vir /M ⊙ ) > 14 galaxy clusters at z ∼ 0. We find the average deviation in MWA from the MWA–M⋆ relation depends on position in PPS, with a maximum difference between the inner cluster and infalling interloper galaxies of ∼1 Gyr. Our model employs infall information from N-body simulations and stochastic star-formation histories from the universemachine model. We find total quenching times of t Q = 3.7 ± 0.4 Gyr and t Q = 4.0 ± 0.2 Gyr after first pericentre, for 9 < log (M ⋆ /M ⊙ ) < 10 and 10 < log (M ⋆ /M ⊙ ) < 10.5 galaxies, respectively. By using MWAs, we break the degeneracy in time of quenching onset and time-scale of star formation rate (SFR) decline. We find that time of quenching onset relative to pericentre is $t_{\mathrm{delay}}=3.5^{+0.6}_{-0.9}$ Gyr and $t_{\mathrm{delay}}=-0.3^{+0.8}_{-1.0}$ Gyr for 9 < log (M ⋆ /M ⊙ ) < 10 and 10 < log (M ⋆ /M ⊙ ) < 10.5 galaxies, respectively, and exponential SFR suppression time-scales are τenv ≤ 1.0 Gyr for 9 < log (M ⋆ /M ⊙ ) < 10 galaxies and τenv ∼ 2.3 Gyr for 10 < log (M ⋆ /M ⊙ ) < 10.5 galaxies. Stochastic star formation histories remove the need for rapid infall quenching to maintain the bimodality in the SFR of cluster galaxies; the depth of the green valley prefers quenching onsets close to first pericentre and a longer quenching envelope, in slight tension with the MWA-driven results. Taken together these results suggest that quenching begins close to, or just after pericentre, but the time-scale for quenching to be fully complete is much longer and therefore ram-pressure stripping is not complete on first pericentric passage.

79 ASTRONOMY AND ASTROPHYSICS↗