Search NASA⌕ Search

SEARCH · Search NASA

Results for “model error”

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 883 records · Page 49

Likelihood-based signal and noise analysis for docking of models into cryo-EM maps

Fast, reliable docking of models into cryo-EM maps requires understanding of the errors in the maps and the models. Likelihood-based approaches to errors have proven to be powerful and adaptable in experimental structural biology, finding applications in both crystallography and cryo-EM. Indeed, previous crystallographic work on the errors in structural models is directly applicable to likelihood targets in cryo-EM. Likelihood targets in Fourier space are derived here to characterize, based on the comparison of half-maps, the direction- and resolution-dependent variation in the strength of both signal and noise in the data. Because the signal depends on local features, the signal and noise are analysed in local regions of the cryo-EM reconstruction. The likelihood analysis extends to prediction of the signal that will be achieved in any docking calculation for a model of specified quality and completeness. A related calculation generalizes a previous measure of the information gained by making the cryo-EM reconstruction.

59 BASIC BIOLOGICAL SCIENCES↗

A 15-year record of CO emissions constrained by MOPITT CO observations

Long-term measurements from satellites and surface stations have demonstrated a decreasing trend of tropospheric carbon monoxide (CO) in the Northern Hemisphere over the past decade. Likely explanations for this decrease include changes in anthropogenic, fires, and/or biogenic emissions or changes in the primary chemical sink hydroxyl radical (OH). Using remotely sensed CO measurements from the Measurement of Pollution in the Troposphere (MOPITT) satellite instrument, in situ methyl chloroform (MCF) measurements from the World Data Centre for Greenhouse Gases (WDCGG) and the adjoint of the GEOS-Chem model, we estimate the change in global CO emissions from 2001 to 2015. We show that the loss rate of MCF varied by 0.2 % in the past 15 years, indicating that changes in global OH distributions do not explain the recent decrease in CO. Our two-step inversion approach for estimating CO emissions is intended to mitigate the effect of bias errors in the MOPITT data as well as model errors in transport and chemistry, which are the primary factors contributing to the uncertainties when quantifying CO emissions using these remotely sensed data. Our results confirm that the decreasing trend of tropospheric CO in the Northern Hemisphere is due to decreasing CO emissions from anthropogenic and biomass burning sources. In particular, we find decreasing CO emissions from the United States and China in the past 15 years, and unchanged anthropogenic CO emissions from Europe since 2008. We find decreasing trends of biomass burning CO emissions from boreal North America, boreal Asia and South America, but little change over Africa. In contrast to prior results, we find that a positive trend in CO emissions is likely for India and southeast Asia.

54 ENVIRONMENTAL SCIENCES↗

A Fast Time-Stepping Strategy for Dynamical Systems Equipped with a Surrogate Model

Simulation of complex dynamical systems arising in many applications is computationally challenging due to their size and complexity. Model order reduction, machine learning, and other types of surrogate modeling techniques offer cheaper and simpler ways to describe the dynamics of these systems but are inexact and introduce additional approximation errors. In order to overcome the computational difficulties of the full complex models, on one hand, and the limitations of surrogate models, on the other, this work proposes a new accelerated time-stepping strategy that combines information from both. This approach is based on the multirate infinitesimal general-structure additive Runge--Kutta framework. The inexpensive surrogate model is integrated with a small time step to guide the solution trajectory, and the full model is treated with a large time step to occasionally correct for the surrogate model error and ensure convergence. Here, we provide a theoretical error analysis, and several numerical experiments, to show that this approach can be significantly more efficient than using only the full or only the surrogate model for the integration.

Surrogate models↗

Exact and Model Exchange-Correlation Potentials for Open-Shell Systems

The conventional approaches to the inverse density functional theory problem typically assume nondegeneracy of the Kohn–Sham (KS) eigenvalues, greatly hindering their use in open-shell systems. Here, we present a generalization of the inverse density functional theory problem that can seamlessly admit degenerate KS eigenvalues. Additionally, we allow for fractional occupancy of the Kohn–Sham orbitals to also handle noninteracting ensemble-v-representable densities, as opposed to just noninteracting pure-v-representable densities. We present the exact exchange-correlation (XC) potentials for six open-shell systems–four atoms (Li, C, N, and O) and two molecules (CN and CH 2 )–using accurate ground-state densities from configuration interaction calculations. We compare these exact XC potentials with model XC potentials obtained using nonlocal (B3LYP, SCAN0) and local/semilocal (SCAN, PBE, PW92) XC functionals. Although the relative errors in the densities obtained from these DFT functionals are of $O$(10 –3 to 10 –2 ), the relative errors in the model XC potentials remain substantially large–$O$(10 –1 to 10 0 ).

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

SU-F-T-105: The Predicted Radiation Response of Non-Small Cell Lung Cancer for SRS, SBRT and HDR Brachytherapy

Purpose: The data on the α/β ratio of non-small cell lung cancer (NSCLC) is scarce in the literature. We have previously proposed a generalized LQ (gLQ) model to address the high dose dilemma of the LQ model. In this study, we applied the gLQ model to both the patients and in vitro cell irradiation data treated with a large range of doses, and investigated the α/β ratio in NSCLC. Methods: 150 patients with T1T2 and non-T1T2 stages were treated with stereotactic body radiotherapy (SBRT). In vitro datasets of 14 NSCLC cell lines from the National Cancer Institute published in Eur J Cancer Clin Oncol. 25(3):527–534 (1989) and 7 NSCLC cell lines published in Cancer Res 57:4285–300 (1997) were included. The gLQ model was used to fit datasets. The least χ2 method was adopted to determine the goodness of fit. Errors of the model parameters were determined by propagating minimal χ2. The α/β ratios from both the patients and these in vitro NSCLC cell lines were obtained. Results: The average of α/β ratios for T1T2 and non-T1T2 NSCLC was 1.45 Gy. The same type of cell lines irradiated with different modalities but almost the same dose rate yielded approximately the same α/β ratio. The average of α/β ratios for NSCLC cell lines in this study was 5.45 Gy. Conclusion: The difference in the α/β ratios between the patients and in vitro cell data is expected and the lower α/β ratio for patients suggests the higher radiosensitivity, which could be associated with higher tumor perfusion or other tumor microenvironmental effects. The α/β ratios derived from the gLQ model can be used in high dose regions or high fraction sizes and are useful to extend our clinical experience accumulated from conversional low-dose fractionation to high dose irradiation schedules.

60 APPLIED LIFE SCIENCES↗

TU-FG-BRB-10: A New Approach to Proton Radiography Using the Beamline X-Ray Flat Panel

Purpose: Proton radiography, which images the patients with the same type of particles that they are to be treated with, is a promising approach for image guidance and range uncertainties reduction. This study aimed to realize quality proton radiography by measuring dose rate functions (DRF) in time domain using a single flat panel and retrieve water equivalent path length (WEPL) from them. Methods: An amorphous silicon flat panel (PaxScan™ 4030CB, Varian Medical Systems, Inc., Palo Alto, CA) was placed behind phantoms to measure DRFs from a proton beam modulated by the modulator wheel. To retrieve WEPL and RSP, calibration models based on the intensity of DRFs only, root mean square (RMS) of DRFs only and the intensity weighted RMS were tested. The quality of obtained WEPL images (in terms of spatial resolution and level of details) and the accuracy of WEPL were compared. Results: RSPs for most of the Gammex phantom inserts were retrieved within ± 1% errors by calibration models based on the RMS and intensity weighted RMS. The mean percentage error for all inserts was reduced from 1.08% to 0.75% by matching intensity in the calibration model. In specific cases such as the insert with a titanium rod, the calibration model based on RMS only fails while the that based on intensity weighted RMS is still valid. The quality of retrieved WEPL images were significantly improved for calibration models including intensity matching. Conclusion: For the first time, a flat panel, which is readily available in the beamline for image guidance, was tested to acquire quality proton radiography with WEPL accurately retrieved from it. This technique is promising to be applied for image-guided proton therapy as well as patient specific RSP determination to reduce uncertainties of beam ranges.

60 APPLIED LIFE SCIENCES↗

GEOMETRIC AND DYNAMICAL MODELS OF REVERBERATION MAPPING DATA

We present a general method to analyze reverberation (or echo) mapping data that simultaneously provides estimates for the black hole mass and for the geometry and dynamics of the broad-line region (BLR) in active galactic nuclei (AGNs). While previous methods yield a typical scale size of the BLR or a reconstruction of the transfer function, our method directly infers the spatial and velocity distribution of the BLR from the data, from which a transfer function can be easily derived. Previous echo mapping analysis requires an independent estimate of a scaling factor known as the virial coefficient to infer the mass of the black hole, but this is not needed in our more direct approach. We use the formalism of Bayesian probability theory and implement a Markov Chain Monte Carlo algorithm to obtain estimates and uncertainties for the parameters of our BLR models. Fitting of models to the data requires knowledge of the continuum flux at all times, not just the measured times. We use Gaussian Processes to interpolate and extrapolate the continuum light curve data in a fully consistent probabilistic manner, taking the associated errors into account. We illustrate our method using simple models of BLR geometry and dynamics and show that we can recover the parameter values of our test systems with realistic uncertainties that depend upon the variability of the AGN and the quality of the reverberation mapping observing campaign. With a geometry model we can recover the mean radius of the BLR to within {approx}0.1 dex random uncertainty for simulated data with an integrated line flux uncertainty of 1.5%, while with a dynamical model we can recover the black hole mass and the mean radius to within {approx}0.05 dex random uncertainty, for simulated data with a line profile average signal-to-noise ratio of 4 per spectral pixel. These uncertainties do not include modeling errors, which are likely to be present in the analysis of real data, and should therefore be considered as lower limits to the accuracy of the method.

79 ASTRONOMY AND ASTROPHYSICS↗

Optimal calibration accuracy for gravitational-wave detectors

Calibration errors in the response function of a gravitational-wave detector degrade its ability to detect and then to measure the properties of any detected signals. This paper derives the needed levels of calibration accuracy for each of these data-analysis tasks. The levels derived here are optimal in the sense that lower accuracy would result in missed detections and/or a loss of measurement precision, while higher accuracy would be made irrelevant by the intrinsic noise level of the detector. Calibration errors affect the data-analysis process in much the same way as errors in theoretical waveform templates. The optimal level of calibration accuracy is expressed therefore as a joint limit on modeling and calibration errors: increased accuracy in one reduces the accuracy requirement in the other.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Tula: Optimizing Time, Cost, and Generalization in Distributed Large-Batch Training

Distributed training increases the number of batches processed per iteration either by scaling-out (adding more nodes) or scaling-up (increasing the batch-size). However, the largest configuration does not necessarily yield the best performance. Horizontal scaling introduces additional communication overhead, while vertical scaling is constrained by computation cost and device memory limits. Thus, simply increasing the batch-size leads to diminishing returns: training time and cost decrease initially but eventually plateaus, creating a knee-point in the time/cost vs. batch-size pareto curve. The optimal batch-size therefore depends on the underlying model, data and available compute resources. Large batches also suffer from worse model quality due to the well-known “generalization gap”. In this paper, we present Tula, an online service that automatically optimizes time, cost, and convergence quality for large-batch training of convolutional models. It combines parallel-systems modeling with statistical performance prediction to identify the optimal batchsize. Tula predicts training time and cost within 7.5−14% error across multiple models, and achieves up to 20× overall speedup and improves test accuracy by ≈9% on average over standard large-batch training on various vision tasks, thus successfully mitigating the generalization gap and accelerating training at the same time.

Tyagi, Sahil [ORNL] (ORCID:0009000783144745)↗

Woody plants optimise stomatal behaviour relative to hydraulic risk

Abstract Stomatal response to environmental conditions forms the backbone of all ecosystem and carbon cycle models, but is largely based on empirical relationships. Evolutionary theories of stomatal behaviour are critical for guarding against prediction errors of empirical models under future climates. Longstanding theory holds that stomata maximise fitness by acting to maintain constant marginal water use efficiency over a given time horizon, but a recent evolutionary theory proposes that stomata instead maximise carbon gain minus carbon costs/risk of hydraulic damage. Using data from 34 species that span global forest biomes, we find that the recent carbon‐maximisation optimisation theory is widely supported, revealing that the evolution of stomatal regulation has not been primarily driven by attainment of constant marginal water use efficiency. Optimal control of stomata to manage hydraulic risk is likely to have significant consequences for ecosystem fluxes during drought, which is critical given projected intensification of the global hydrological cycle.

Anderegg, William R. L.↗

BLENDING STUDY FOR SRR SALT DISPOSITION INTEGRATION: TANK 50H SCALE-MODELING AND COMPUTER-MODELING FOR BLENDING PUMP DESIGN, PHASE 2

The Salt Disposition Integration (SDI) portfolio of projects provides the infrastructure within existing Liquid Waste facilities to support the startup and long term operation of the Salt Waste Processing Facility (SWPF). Within SDI, the Blend and Feed Project will equip existing waste tanks in the Tank Farms to serve as Blend Tanks where 300,000-800,000 gallons of salt solution will be blended in 1.3 million gallon tanks and qualified for use as feedstock for SWPF. Blending requires the miscible salt solutions from potentially multiple source tanks per batch to be well mixed without disturbing settled sludge solids that may be present in a Blend Tank. Disturbing solids may be problematic both from a feed quality perspective as well as from a process safety perspective where hydrogen release from the sludge is a potential flammability concern. To develop the necessary technical basis for the design and operation of blending equipment, Savannah River National Laboratory (SRNL) completed scaled blending and transfer pump tests and computational fluid dynamics (CFD) modeling. A 94 inch diameter pilot-scale blending tank, including tank internals such as the blending pump, transfer pump, removable cooling coils, and center column, were used in this research. The test tank represents a 1/10.85 scaled version of an 85 foot diameter, Type IIIA, nuclear waste tank that may be typical of Blend Tanks used in SDI. Specifically, Tank 50 was selected as the tank to be modeled per the SRR, Project Engineering Manager. SRNL blending tests investigated various fixed position, non-rotating, dual nozzle pump designs, including a blending pump model provided by the blend pump vendor, Curtiss Wright (CW). Primary research goals were to assess blending times and to evaluate incipient sludge disturbance for waste tanks. Incipient sludge disturbance was defined by SRR and SRNL as minor blending of settled sludge from the tank bottom into suspension due to blending pump operation, where the sludge level was shown to remain constant. To experimentally model the sludge layer, a very thin, pourable, sludge simulant was conservatively used for all testing. To experimentally model the liquid, supernate layer above the sludge in waste tanks, two salt solution simulants were used, which provided a bounding range of supernate properties. One solution was water (H{sub 2}O + NaOH), and the other was an inhibited, more viscous salt solution. The research performed and data obtained significantly advances the understanding of fluid mechanics, mixing theory and CFD modeling for nuclear waste tanks by benchmarking CFD results to actual experimental data. This research significantly bridges the gap between previous CFD models and actual field experiences in real waste tanks. A finding of the 2009, DOE, Slurry Retrieval, Pipeline Transport and Plugging, and Mixing Workshop was that CFD models were inadequate to assess blending processes in nuclear waste tanks. One recommendation from that Workshop was that a validation, or bench marking program be performed for CFD modeling versus experiment. This research provided experimental data to validate and correct CFD models as they apply to mixing and blending in nuclear waste tanks. Extensive SDI research was a significant step toward bench marking and applying CFD modeling. This research showed that CFD models not only agreed with experiment, but demonstrated that the large variance in actual experimental data accounts for misunderstood discrepancies between CFD models and experiments. Having documented this finding, SRNL was able to provide correction factors to be used with CFD models to statistically bound full scale CFD results. Through the use of pilot scale tests performed for both types of pumps and available engineering literature, SRNL demonstrated how to effectively apply CFD results to salt batch mixing in full scale waste tanks. In other words, CFD models were in error prior to development of experimental correction factors determined during this research, which provided a technique to use CFD models for salt batch mixing and transfer pump operations. This major scientific advance in mixing technology resulted in multi-million dollar cost savings to SRR. New techniques were developed for both experiment and analysis to complete this research. Supporting this success, research findings are summarized in the Conclusions section of this report, and technical recommendations for design and operation are included in this section of the report.

08 HYDROGEN↗

Benchmarking the performance of uncertainty quantification methods for neural network-based interatomic potentials

Machine-learned interatomic potentials (ML-IAPs) continue to gain popularity as accurate, computationally efficient replacements for traditional, physics-based interatomic potentials and expensive ab initio methods. Uncertainty quantification (UQ) of ML-IAPs is a growing area of research as UQ is critical in many applications of IAPs, such as developing curated datasets, active learning-based data augmentation, self-improving models, and estimating the uncertainty of molecular dynamics simulations. In this paper, we construct and benchmark a series of different neural network potentials (NNPs) with varying network architectures to determine the performance of these models with respect to both the mean and uncertainty calibration error. Each NNP method is specifically designed to predict either epistemic or aleatoric uncertainty with particular focus on the differences in behavior between the epistemic and aleatoric uncertainty estimates. We benchmark these methods using multiple datasets common in the ML-IAP literature. The results show that the aleatoric uncertainty from single-shot model architectures is a competitive alternative to ensemble-based epistemic uncertainty predictions in regions of sufficient data-density. However, in regions where the representative data is sparse, aleatoric uncertainty models tend to overpredict and epistemic methods tend to underpredict the actual model error. We conclude that the type of UQ is crucial when discussing performance of probabilistic model results as different methods have different performance characteristics depending on the regime in which they are evaluated. Therefore, the type of UQ method should be carefully evaluated against both the data characteristics and requirements for the intended application.

97 MATHEMATICS AND COMPUTING↗

Exploring Autoencoder-based Error-bounded Compression for Scientific Data

Error-bounded lossy compression is becoming an indispensable technique for the success of today's scientific projects with vast volumes of data produced during the simulations or instrument data acquisitions. Not only can it significantly reduce data size, but it also can control the compression errors based on user-specified error bounds. Autoencoder (AE) models have been widely used in image compression, but few AE-based compression approaches support error-bounding features, which are highly required by scientific applications. To address this issue, we explore using convolutional autoencoders to improve error-bounded lossy compression for scientific data, with the following three key contributions. (1) We provide an in-depth investigation of the characteristics of various autoencoder models and develop an error-bounded autoencoder-based framework in terms of the SZ model. (2) We optimize the compression quality for main stages in our designed AE-based error-bounded compression framework, fine-tuning the block sizes and latent sizes and also optimizing the compression efficiency of latent vectors. (3) We evaluate our proposed solution using five real-world scientific datasets and comparing them with six other related works. Experiments show that our solution exhibits a very competitive compression quality from among all the compressors in our tests. In absolute terms, it can obtain a much better compression quality (100%similar to 800% improvement in compression ratio with the same data distortion) compared with SZ2.1 and ZFP in cases with a high compression ratio.

Liu, Jinyang↗

Validation of Multiple Tools for Flat Plate Photovoltaic Modeling Against Measured Data

This report expands upon a previous work by the same authors, published in the 40th IEEE Photovoltaic Specialists conference. In this validation study, comprehensive analysis is performed on nine photovoltaic systems for which NREL could obtain detailed performance data and specifications, including three utility-scale systems and six commercial scale systems. Multiple photovoltaic performance modeling tools were used to model these nine systems, and the error of each tool was analyzed compared to quality-controlled measured performance data. This study shows that, excluding identified outliers, all tools achieve annual errors within +/-8% and hourly root mean squared errors less than 7% for all systems. It is further shown using SAM that module model and irradiance input choices can change the annual error with respect to measured data by as much as 6.6% for these nine systems, although all combinations examined still fall within an annual error range of +/-8.5%. Additionally, a seasonal variation in monthly error is shown for all tools. Finally, the effects of irradiance data uncertainty and the use of default loss assumptions on annual error are explored, and two approaches to reduce the error inherent in photovoltaic modeling are proposed.

14 SOLAR ENERGY↗

Verification of the 3-Region Advanced Test Reactor MCNP Model

The verification of the 3-region homogenized fuel Advanced Test Reactor MCNP model. The 3-region model was compared to the 19-plate model found in the 94-CIC report. Flux tallies, energy deposition tallies, and quarter core mesh tallies were used to compare the two models. The 3-region model needed updating in order to make good comparisons between the models. The percent error from the flux and energy deposition tallies data shows that experiment positions inside the flux trap have higher errors than positions outside the fuel ring. The standard deviation data obtained from the mesh tallies shows that the two models agree within two standard deviations throughout the reactor. It is concluded that the model works adequately for what it is used for.

99 GENERAL AND MISCELLANEOUS↗

Verification of the 3-Region Advanced Test Reactor MCNP Model

The verification of the 3-region homogenized fuel Advanced Test Reactor MCNP model. The 3-region model was compared to the 19-plate model found in the 94-CIC report. Flux tallies, energy deposition tallies, and quarter core mesh tallies were used to compare the two models. The 3-region model needed updating in order to make good comparisons between the models. The percent error from the flux and energy deposition tallies data shows that experiment positions inside the flux trap have higher errors than positions outside the fuel ring. The standard deviation data obtained from the mesh tallies shows that the two models agree within two standard deviations throughout the reactor. It is concluded that the model works adequately for what it is used for.

99 GENERAL AND MISCELLANEOUS↗

Calibration and Localization of Optically Pumped Magnetometers Using Electromagnetic Coils

In this paper, we propose a method to estimate the position, orientation, and gain of a magnetic field sensor using a set of (large) electromagnetic coils. We apply the method for calibrating an array of optically pumped magnetometers (OPMs) for magnetoencephalography (MEG). We first measure the magnetic fields of the coils at multiple known positions using a well-calibrated triaxial magnetometer, and model these discreetly sampled fields using vector spherical harmonics (VSH) functions. We then localize and calibrate an OPM by minimizing the sum of squared errors between the model signals and the OPM responses to the coil fields. We show that by using homogeneous and first-order gradient fields, the OPM sensor parameters (gain, position, and orientation) can be obtained from a set of linear equations with pseudo-inverses of two matrices. The currents that should be applied to the coils for approximating these low-order field components can be determined based on the VSH models. Computationally simple initial estimates of the OPM sensor parameters follow. As a first test of the method, we placed a fluxgate magnetometer at multiple positions and estimated the RMS position, orientation, and gain errors of the method to be 1.0 mm, 0.2°, and 0.8%, respectively. Lastly, we calibrated a 48-channel OPM array. The accuracy of the OPM calibration was tested by using the OPM array to localize magnetic dipoles in a phantom, which resulted in an average dipole position error of 3.3 mm. The results demonstrate the feasibility of using electromagnetic coils to calibrate and localize OPMs for MEG.

47 OTHER INSTRUMENTATION↗

Blind photovoltaic modeling intercomparison: A multidimensional data analysis and lessons learned

The Photovoltaic (PV) Performance Modeling Collaborative (PVPMC) organized a blind PV performance modeling intercomparison to allow PV modelers to blindly test their models and modeling ability against real system data. Measured weather and irradiance data were provided along with detailed descriptions of PV systems from two locations (Albuquerque, New Mexico, USA, and Roskilde, Denmark). Participants were asked to simulate the plane-of-array irradiance, module temperature, and DC power output from six systems and submit their results to Sandia for processing. The results showed overall median mean bias (i.e., the average error per participant) of 0.6% in annual irradiation and –3.3% in annual energy yield. While most PV performance modeling results seem to exhibit higher precision and accuracy as compared to an earlier blind PV modeling study in 2010, human errors, modeling skills, and derates were found to still cause significant errors in the estimates.

14 SOLAR ENERGY↗