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 919 records · Page 51

A tutorial review of machine learning-based model predictive control methods

Abstract This tutorial review provides a comprehensive overview of machine learning (ML)-based model predictive control (MPC) methods, covering both theoretical and practical aspects. It provides a theoretical analysis of closed-loop stability based on the generalization error of ML models and addresses practical challenges such as data scarcity, data quality, the curse of dimensionality, model uncertainty, computational efficiency, and safety from both modeling and control perspectives. The application of these methods is demonstrated using a nonlinear chemical process example, with open-source code available on GitHub. The paper concludes with a discussion on future research directions in ML-based MPC.

Wu, Zhe [Department of Chemical and Biomolecular E↗

Branch-Based Model for the Diameters of the Pulmonary Airways: Accounting for Departures From Self-Consistency and Registration Errors

We examine a previously published branch-based approach to modeling airway diameters that is predicated on the assumption of self-consistency across all levels of the tree. We mathematically formulate this assumption, propose a method to test it and develop a more general model to be used when the assumption is violated. We discuss the effect of measurement error on the estimated models and propose methods that account for it. The methods are illustrated on data from MRI and CT images of silicone casts of two rats, two normal monkeys and one ozone-exposed monkey. Our results showed substantial departures from self-consistency in all five subjects. When departures from selfconsistency exist we do not recommend using the self-consistency model, even as an approximation, as we have shown that it may likely lead to an incorrect representation of the diameter geometry. Measurement error has an important impact on the estimated morphometry models and needs to be accounted for in the analysis.

59 BASIC BIOLOGICAL SCIENCES↗

Validation of the AltRAM Physics Models for use with Compressed Natural Gas

The Alternative Fuels Risk Assessment Models (A1tRAM) toolkit combines Quantitative Risk Assessment (QRA) with simulations of unignited dispersion, ignited turbulent diffusion flames, and indoor accumulation with delayed ignition of fuels. The models of the physical phenomena need to be validated for each of the fuels in the toolkit. This report shows the validation for methane which is being used as a surrogate for natural gas. For the unignited dispersion model, seven previously published experiments from credible sources were used to validate. The validation looked at gas concentrations with respect to the distance from the release point. Four of these were underexpanded jets (i.e. release velocity equal to or greater than local speed of sound) and the other three subsonic releases. The methane plume model in AltRAM matched both varieties well, with higher accuracy for the underexpanded releases. For the jet flame model, we compared the heat flux and thermal radiation data reported from five separate turbulent jet flame experiments to the quantities calculated by A1tRAM. Four of the five datasets were for underexpanded diffusion jets flames. While the results still match well enough to give a good estimate of what is occurring, the error is higher than what was seen with the plume model. For the underexpanded flames A1tRAM provided reasonable approximations, which would lead to conservative risk assessments. Some modeling errors can be attributed to environmental effects (i.e. wind) since most large scale flame experiments are conducted outdoors. A1tRAM has been shown to be a reasonably accurate tool for calculating the concentration or flame properties of natural gas releases. Improvements could still be made for the plume of subsonic releases and radiative heat fluxes to reduce the conservative nature of these predictions. These models can provide valuable information for the risk assessment of natural gas infrastructure.

03 NATURAL GAS↗

Assessment of Model Confidence of a Laser Source Model in xRAGE Using Omega Direct-Drive Implosion Experiments

Los Alamos National Laboratory is interested in developing high-energy-density physics validation capabilities for its multiphysics code xRAGE. xRAGE was recently updated with the laser package Mazinisin to improve predictability. We assess the current implementation and coupling of the laser package via validation of laser-driven, direct-drive spherical capsule experiments from the Omega laser facility. Here, the ASME V&V 20-2009 standard is used to determine the model confidence of xRAGE, and considerations for high-energy-density physics are identified. With current modeling capabilities in xRAGE, the model confidence is overwhelmed by significant systematic errors from the experiment or model. Validation evidence suggests cross-beam energy transfer as a dominant source of the systematic error.

47 OTHER INSTRUMENTATION↗

Unraveling the 2021 Central Tennessee flood event using a hierarchical multi-model inundation modeling framework

Flood prediction systems need hierarchical atmospheric, hydrologic, and hydraulic models to predict rainfall, runoff, streamflow, and floodplain inundation. The accuracy of such systems depends on the error propagation through the modeling chain, sensitivity to input data, and choice of models. In this study, we used multiple precipitation forcings (hindcast and forecast) to drive hydrologic and hydrodynamic models to analyze the impacts of various drivers on the estimates of flood inundation depth and extent. We implement this framework to unravel the August 2021 extreme flooding event that occurred in Central Tennessee, USA. We used two radar-based quantitative precipitation estimates (STAGE4 and MRMS) as well as quantitative precipitation forecasts (QPF) from the National Weather Service Weather Prediction Center (WPC) to drive a series of models in the hierarchical framework, including the Variable Infiltration Capacity (VIC) land surface model, the Routing Application for Parallel Computation of Discharge (RAPID) river routing model, and the AutoRoute and TRITON inundation models. An evaluation with observed high-water marks demonstrates that the framework can reasonably simulate flood inundation. Despite the complex error propagation mechanism of the modeling chain, we show that inundation estimates are most sensitive to rainfall estimates. Most notably, QPF significantly underestimates flood magnitudes and inundations leading to unanticipated severe flooding for all stakeholders involved in the event. Finally, we discuss the implications of the hydrodynamic modeling framework for real-time flood forecasting.

54 ENVIRONMENTAL SCIENCES↗

Development of a Turbulent Liquid Spray Atomization Model for Diesel Engine Simulations (Final Technical Report)

This project addresses the systematic lack of predictive capabilities by spray models within engine CFD codes. We develop a new modeling approach to predict the breakup of diesel sprays based on recent literature showing that liquid turbulence plays a fundamental role in spray atomization. A new body of quantitative validation data is also developed as a critical element of the project, leveraging the joint capabilities of Georgia Tech’s high-pressure continuous-flow spray chamber and Argonne National Lab’s near-nozzle x-ray diagnostics at the Advanced Photon Source. This project contributes spatially-resolved measurements of drop size distribution within well-characterized diesel injectors, Spray A and D, from the Engine Combustion Network (ECN) to the engine combustion community for the first time. Utilizing this new body of measurements, we validate and demonstrate a new spray model for diesel sprays, termed the KH-Faeth model, that predicts global and local spray characteristic more accurately than the widely adopted and employed KH model. Predicted drop size distributions are seen to predict measured drops sizes both quantitatively and predictively, with accurate response in droplet size distributions over a wide range of ambient density, injection pressure, and injector nozzle size (Spray A and D) without model tuning. The KH-Faeth model can reduce error in the predicted centerline droplet size profile by up to 80% for ECN Spray D simulations when compared to use of the widely employed KH model.

33 ADVANCED PROPULSION SYSTEMS↗

Process-oriented evaluation using ARM observations to improve clouds and radiation in a global atmospheric model

The goal of the research is to improve understanding of cloud, radiation and precipitation processes and their representation in numerical models for weather and climate using observations from the ARM/ASR program. Global models need to represent all cloud types across different weather regimes from the tropics to the poles and ARM provides long-term observations at different geographical sites that cover these different regimes. Combined with satellite data to give a global context, the observations are used to evaluate the systematic errors in a global model for predictions of cloud, radiation and precipitation. The focus is on understanding the links to underlying physical processes to pinpoint the sources of error and improve the model representation of these processes. The research covers boundary layer cloud over the continental U.S., marine stratocumulus and shallow cumulus in the subtropics, mixed-phase clouds over the Arctic and Southern Ocean, and high-altitude cirrus ice clouds in the tropics. The model used is the operational global weather prediction system at the European Centre for Medium-range Weather Forecasts (ECMWF). The research has led to numerous developments, several of which have been implemented operationally leading to improvements in everyday forecasts. Many of the systematic errors are common across different modelling systems and the wider implications from the research for other weather and climate models are highlighted.

54 ENVIRONMENTAL SCIENCES↗

The Method of Finite Averages: A rigorous upscaling methodology for heterogeneous porous media

Rigorous upscaling techniques offer accurate and computationally-efficient strategies for modeling the average behaviors of multi-physical, multiscale phenomena in geological porous media. However, such techniques often rely on a variety of methodological assumptions that prohibit their rigorous application to practical systems (e.g., systems involving heterogeneous porous media, system-scale boundary conditions, and fine-scale dynamics that are not diffusion-dominant). In this work, we aim to formulate an upscaling methodology with few methodological assumptions to provide high levels of model generality and foster the utilization of rigorously-derived upscaled models in practice. In particular, we introduce the Method of Finite Averages (MoFA), a novel upscaling methodology for rigorously modeling heterogeneous porous media and system-scale boundary conditions. We then detail MoFA’s implementation for the advective–diffusive transport of a single species and compare the methodology with classic numerical techniques, as well as other rigorous upscaling techniques, to highlight MoFA’s unique combination of rigor and generality. We then validate the derived model while demonstrating its benefits in three numerical experiments. The results suggest that (1.) the applicability and a priori error guarantees of MoFA models do not directly depend on system geometry, (2.) a model’s applicability and error guarantees can be can arbitrarily expanded and reduced, respectively, with further computational expense, and (3.) downscaling with MoFA provides an efficient strategy for generating accurate pore-scale solutions from upscaled results. Ultimately, the results evidence that upscaled models can be rigorously derived for heterogeneous porous media systems and resolved in a fraction of the time it takes to perform the equivalent pore-scale simulations.

58 GEOSCIENCES↗

Analyzing Hardware and Software Common Cause Failures in Digital Instrumentation and Control Systems using Dual Error Propagation Method

This paper develops a methodology for quantifying software common cause failures (CCFs) in digital instrumentation and control (I&C) systems of nuclear power plants. To support the transition of analog I&C systems to digital in nuclear power plants, probabilistic risk assessment (PRA) techniques are used. The hardware components of the I&C systems have reliability databases that can be used in the PRA studies. However, the failure data for redundant software components of the systems is sparse. Failure of components constitutes a CCF, wherein two or more components or systems fail due to a single shared cause and coupling mechanism. This paper proposes a quantification approach that can simultaneously model hardware and software components, incorporate the CCFs of software systems in the models, and bridge the gap between the failure quantification of models and the development of CCF parametric databases. We demonstrate the dual error propagation method (DEPM) by developing I&C systems failure models for a representative digital reactor trip system. The DEPM models are built to simulate the control and data flows within the systems and can accommodate failure states. By expanding DEPM to software CCFs, we generated alpha factor parameter estimates for each of the modeled error propagation mechanisms.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Mesh-based super-resolution of fluid flows with multiscale graph neural networks

A graph neural network (GNN) approach is introduced in this work which enables mesh-based three-dimensional super-resolution of fluid flows. In this framework, the GNN is designed to operate not on the full mesh-based field at once, but on localized meshes of elements (or cells) directly. To facilitate mesh-based GNN representations in a manner similar to spectral (or finite) element discretizations, a baseline GNN layer (termed a message passing layer, which updates local node properties) is modified to account for synchronization of coincident graph nodes, rendering compatibility with commonly used element-based mesh connectivities. Furthermore, the architecture is multiscale in nature, and is comprised of a combination of coarse-scale and fine-scale message passing layer sequences (termed processors) separated by a graph unpooling layer. The coarse-scale processor embeds a query element (alongside a set number of neighboring coarse elements) into a single latent graph representation using coarse-scale synchronized message passing over the element neighborhood, and the fine-scale processor leverages additional message passing operations on this latent graph to correct for interpolation errors. Demonstration studies are performed using hexahedral mesh-based data from Taylor–Green Vortex and backward-facing step flow simulations at Reynolds numbers of 1600 and 3200. Through analysis of both global and local errors, the results ultimately show how the GNN is able to produce accurate super-resolved fields compared to targets in both coarse-scale and multiscale model configurations. Reconstruction errors for fixed architectures were found to increase in proportion to the Reynolds number. Geometry extrapolation studies on a separate cavity flow configuration show promising cross-mesh capabilities of the super-resolution strategy.

Backward-facing step↗

Fitting the constitution type Ia supernova data with the redshift-binned parametrization method

In this work, we explore the cosmological consequences of the recently released Constitution sample of 397 Type Ia supernovae (SNIa). By revisiting the Chevallier-Polarski-Linder (CPL) parametrization, we find that, for fitting the Constitution set alone, the behavior of dark energy (DE) significantly deviates from the cosmological constant {lambda}, where the equation of state (EOS) w and the energy density {rho}{sub {lambda}} of DE will rapidly decrease along with the increase of redshift z. Inspired by this clue, we separate the redshifts into different bins, and discuss the models of a constant w or a constant {rho}{sub {lambda}} in each bin, respectively. It is found that for fitting the Constitution set alone, w and {rho}{sub {lambda}} will also rapidly decrease along with the increase of z, which is consistent with the result of CPL model. Moreover, a step function model in which {rho}{sub {lambda}} rapidly decreases at redshift z{approx}0.331 presents a significant improvement ({delta}{chi}{sup 2}=-4.361) over the CPL parametrization, and performs better than other DE models. We also plot the error bars of DE density of this model, and find that this model deviates from the cosmological constant {lambda} at 68.3% confidence level (CL); this may arise from some biasing systematic errors in the handling of SNIa data, or more interestingly from the nature of DE itself. In addition, for models with same number of redshift bins, a piecewise constant {rho}{sub {lambda}} model always performs better than a piecewise constant w model; this shows the advantage of using {rho}{sub {lambda}}, instead of w, to probe the variation of DE.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Relating flow resistance to equivalent roughness

Describing flow resistance using the physical properties of an underlying surface is a recalcitrant problem in overland flow models. If discharge measurements are available, an equivalent roughness (e.g., Manning’s n) can be calibrated to represent the effects of surface properties within the domain with a single numerical value. Alternatively, the flow resistance can be estimated from discharge and velocity measured at a point, typically a runoff plot outlet. However, such experimental estimates are often inconsistent with the equivalent roughness determined from calibration to discharge, even if both derive from the same dataset. For example, if Manning’s equation is used to parameterize flow resistance, the Manning’s n obtained by calibrating a model to discharge differs from the value of n calculated from measured flow and velocity at the hillslope outlet. Here, this discrepancy is resolved by deriving a correction factor relating experimentally-determined flow resistance to the equivalent roughness. The derived correction factor is tested for four commonly-used resistance formulations using 129 rainfall simulator experiments. The correction factor is necessary to reproduce measured velocities, and yields minor improvements in discharge prediction. Plain Language Summary: Accurate runoff prediction is needed for land and water management in dryland regions, where sporadic and limited rainfall necessitate efficient water use and drought mitigation strategies. The skill of runoff models is known to be hindered by out ability to estimate flow resistance, which is the quantity that describes how energy is lost from flowing water to the underlying surface. Typically, models represent flow resistance with an equivalent roughness, e.g., Manning’s n, that is adjusted until the model can reproduce available discharge observations at watershed scale. However, the flow resistance measured in plot-scale experiments (1–10 m) often exceeds equivalent roughness coefficients by a factor of 10. This means that the direct use of plot-scale experimental data to parameterize runoff models could cause errors in discharge and runoff velocity predictions. Here, we resolve these differences by deriving an analytic correction factor that relates flow resistance to the equivalent roughness required for models to reproduce experimental velocity and discharge data. This correction factor is tested using rainfall simulator data from 129 experiments performed in the US Southwest covering a wide range of precipitation intensities, soil textures and vegetation types. Use of the correction factor substantially improves model prediction of flow velocity, which is needed for reproducing the timing of flood events and the estimation of erosion.

54 ENVIRONMENTAL SCIENCES↗

Accuracy of Partial Core Corrections Using Fourier Transforms in Pseudopotential–Density Functional Theory

Partial core corrections can be important in obtaining an accurate description of nonlinear exchange-correlation functionals and improving the transferability of pseudopotentials. We show that a widely used procedure, which calculates partial core charge density, $ρ^{partial}_{core}$, in Fourier space and then converts it to real space with fast Fourier transforms, can lead to sizable numerical errors of exchange-correlation potentials in the vacuum region. Such errors occur in modeling low-dimensional materials or surfaces with supercells. The loss of accuracy originates from the slow-decaying feature of core charge density in reciprocal space. Numerical errors on the order of 1 eV in the Kohn–Sham energies of unoccupied states can occur in pseudopotential–density functional calculations. The direct calculation of the partial core charge in real space can avoid the numerical errors caused by Fourier transforms.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Microsecond Molecular Dynamics Simulations of Proteins Using a Quasi-Equilibrium Solvation Shell Model

In this work, we describe the development and implementation of a quasi-equilibrium hydration shell model of biomolecular solvation with adaptive boundaries. Applying the model to microsecond-long molecular dynamics simulations of several protein systems of varying complexity, we find that the model simulation results are of comparable quality to those obtained from simulations of fully solvated systems, but at a reduced computational cost. We discuss the dominant sources of error in the model and outline directions for future improvements.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Machine Learning Based Multi-Physical-Model Blending for Enhancing Renewable Energy Forecast -- Improvement via Situation Dependent Error Correction

With increasing penetration of solar and wind energy to the total energy supply mix, the pressing need for accurate energy forecasting has become well-recognized. Here we report the development of a machine-learning based model blending approach for statistically combining multiple meteorological models for improving the accuracy of solar/wind power forecast. Importantly, we demonstrate that in addition to parameters to be predicted (such as solar irradiance and power), including additional atmospheric state parameters which collectively define weather situations as machine learning input provides further enhanced accuracy for the blended result. Functional analysis of variance shows that the error of individual model has substantial dependence on the weather situation. The machine-learning approach effectively reduces such situation dependent error thus produces more accurate results compared to conventional multi-model ensemble approaches based on simplistic equally or unequally weighted model averaging. Validation over an extended period of time results show over 30% improvement in solar irradiance/power forecast accuracy compared to forecasts based on the best individual model.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Simulating organic aerosol in Delhi with WRF-Chem using the volatility-basis-set approach: exploring model uncertainty with a Gaussian process emulator

The nature and origin of organic aerosol in the atmosphere remain unclear. The gas–particle partitioning of semi-volatile organic compounds (SVOCs) that constitute primary organic aerosols (POAs) and the multigenerational chemical aging of SVOCs are particularly poorly understood. The volatility basis set (VBS) approach, implemented in air quality models such as WRF-Chem (Weather Research and Forecasting model with Chemistry), can be a useful tool to describe emissions of POA and its chemical evolution. However, the evaluation of model uncertainty and the optimal model parameterization may be expensive to probe using only WRF-Chem simulations. Gaussian process emulators, trained on simulations from relatively few WRF-Chem simulations, are capable of reproducing model results and estimating the sources of model uncertainty within a defined range of model parameters. In this study, a WRF-Chem VBS parameterization is proposed; we then generate a perturbed parameter ensemble of 111 model runs, perturbing 10 parameters of the WRF-Chem model relating to organic aerosol emissions and the VBS oxidation reactions. This allowed us to cover the model's uncertainty space and to compare outputs from each run to aerosol mass spectrometer observations of organic aerosol concentrations and O:C ratios measured in New Delhi, India. The simulations spanned the organic aerosol concentrations measured with the aerosol mass spectrometer (AMS). However, they also highlighted potential structural errors in the model that may be related to unsuitable diurnal cycles in the emissions and/or failure to adequately represent the dynamics of the planetary boundary layer. While the structural errors prevented us from clearly identifying an optimized VBS approach in WRF-Chem, we were able to apply the emulator in the following two periods: the full period (1–29 May) and a subperiod period of 14:00–16:00 h LT (local time) on 1–29 May. The combination of emulator analysis and model evaluation metrics allowed us to identify plausible parameter combinations for the analyzed periods. We demonstrate that the methodology presented in this study can be used to determine the model uncertainty and to identify the appropriate parameter combination for the VBS approach and hence to provide valuable information to improve our understanding of OA production.

54 ENVIRONMENTAL SCIENCES↗

Characterization of organic aerosol across the global remote troposphere: a comparison of ATom measurements and global chemistry models

The spatial distribution and properties of submicron organic aerosol (OA) are among the key sources of uncertainty in our understanding of aerosol effects on climate. Uncertainties are particularly large over remote regions of the free troposphere and Southern Ocean, where very few data have been available and where OA predictions from AeroCom Phase II global models span 2 to 3 orders of magnitude, greatly exceeding the model spread over source regions. The (nearly) pole-to-pole vertical distribution of non-refractory aerosols was measured with an aerosol mass spectrometer onboard the NASA DC-8 aircraft as part of the Atmospheric Tomography (ATom) mission during the Northern Hemisphere summer (August 2016) and winter (February 2017). This study presents the first extensive characterization of OA mass concentrations and their level of oxidation in the remote atmosphere. OA and sulfate are the major contributors by mass to submicron aerosols in the remote troposphere, together with sea salt in the marine boundary layer. Sulfate was dominant in the lower stratosphere. OA concentrations have a strong seasonal and zonal variability, with the highest levels measured in the lower troposphere in the summer and over the regions influenced by biomass burning from Africa (up to 10 µg sm -3 ). Lower concentrations (~0.1–0.3 µg sm -3 ) are observed in the northern middle and high latitudes and very low concentrations (<0.1 µg sm -3 ) in the southern middle and high latitudes. The ATom dataset is used to evaluate predictions of eight current global chemistry models that implement a variety of commonly used representations of OA sources and chemistry, as well as of the AeroCom-II ensemble. The current model ensemble captures the average vertical and spatial distribution of measured OA concentrations, and the spread of the individual models remains within a factor of 5. These results are significantly improved over the AeroCom-II model ensemble, which shows large overestimations over these regions. However, some of the improved agreement with observations occurs for the wrong reasons, as models have the tendency to greatly overestimate the primary OA fraction and underestimate the secondary fraction. Measured OA in the remote free troposphere is highly oxygenated, with organic aerosol to organic carbon (OA/OC) ratios of ~2.2–2.8, and is 30 %–60 % more oxygenated than in current models, which can lead to significant errors in OA concentrations. The model–measurement comparisons presented here support the concept of a more dynamic OA system as proposed by Hodzic et al. (2016), with enhanced removal of primary OA and a stronger production of secondary OA in global models needed to provide better agreement with observations.

54 ENVIRONMENTAL SCIENCES↗

Modeling of streamflow in a 30 km long reach spanning 5 years using OpenFOAM 5.x

Abstract. Developing accurate and efficient modeling techniques for streamflow at the tens-of-kilometers spatial scale and multi-year temporal scale is critical for evaluating and predicting the impact of climate- and human-induced discharge variations on river hydrodynamics. However, achieving such a goal is challenging because of limited surveys of streambed hydraulic roughness, uncertain boundary condition specifications, and high computational costs. We demonstrate that accurate and efficient three-dimensional (3-D) hydrodynamic modeling of natural rivers at 30 km and 5-year scales is feasible using the following three techniques within OpenFOAM, an open-source computational fluid dynamics platform: (1) generating a distributed hydraulic roughness field for the streambed by integrating water-stage observation data, a rough wall theory, and a local roughness optimization and adjustment strategy; (2) prescribing the boundary condition for the inflow and outflow by integrating precomputed results of a one-dimensional (1-D) hydraulic model with the 3-D model; and (3) reducing computational time using multiple parallel runs constrained by 1-D inflow and outflow boundary conditions. Streamflow modeling for a 30 km long reach in the Columbia River (CR) over 58 months can be achieved in less than 6 d using 1.1 million CPU hours. The mean error between the modeled and the observed water stages for our simulated CR reach ranges from −16 to 9 cm (equivalent to approximately ±7 % relative to the average water depth) at seven locations during most of the years between 2011 and 2019. We can reproduce the velocity distribution measured by the acoustic Doppler current profiler (ADCP). The correlation coefficients of the depth-averaged velocity between the model and ADCP measurements are in the range between 0.71 and 0.83 at 75 % of the survey cross sections. With the validated model, we further show that the relative importance of dynamic pressure versus hydrostatic pressure varies with discharge variations and topography heterogeneity. Given the model's high accuracy and computational efficiency, the model framework provides a generic approach to evaluate and predict the impacts of climate- and human-induced discharge variations on river hydrodynamics at tens-of-kilometers and decadal scales.

58 GEOSCIENCES↗