Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “training data”

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 901 records · Page 50
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Consistent performance of large language models in rare disease diagnosis across ten languages and 4917 cases

Background Large language models (LLMs) are increasingly used medicine for diverse applications including differential diagnostic support. The training data used to create LLMs such as the Generative Pretrained Transformer (GPT) predominantly consist of English-language texts, but LLMs could be used across the globe to support diagnostics if language barriers could be overcome. Initial pilot studies on the utility of LLMs for differential diagnosis in languages other than English have shown promise, but a large-scale assessment on the relative performance of these models in a variety of European and non-European languages on a comprehensive corpus of challenging rare-disease cases is lacking. Methods We created 4917 clinical vignettes using structured data captured with Human Phenotype Ontology (HPO) terms with the Global Alliance for Genomics and Health (GA4GH) Phenopacket Schema. These clinical vignettes span a total of 360 distinct genetic diseases with 2525 associated phenotypic features. We used translations of the Human Phenotype Ontology together with language-specific templates to generate prompts in English, Chinese, Czech, Dutch, French, German, Italian, Japanese, Spanish, and Turkish. We applied GPT-4o, version gpt-4o-2024-08-06, and the medically fine-tuned Meditron3-70B to the task of delivering a ranked differential diagnosis using a zero-shot prompt. An ontology-based approach with the Mondo disease ontology was used to map synonyms and to map disease subtypes to clinical diagnoses in order to automate evaluation of LLM responses. Findings For English, GPT-4o placed the correct diagnosis at the first rank 19.9% and within the top-3 ranks 27.0% of the time. In comparison, for the nine non-English languages tested here the correct diagnosis was placed at rank 1 between 16.9% and 20.6%, within top-3 between 25.4% and 28.6% of cases. The Meditron3 model placed the correct diagnosis within the first 3 ranks for 20.9% of cases in English and between 19.9% and 24.0% for the other nine languages. Interpretation The differential diagnostic performance of LLMs across a comprehensive corpus of rare-disease cases was largely consistent across the ten languages tested. This suggests that the utility of LLMs in clinical settings may extend to non-English clinical settings.

Artificial intelligence↗
DOE OSTI2018

Automatic signal decoding and sensor stability of a 3-electrode mixed-potential sensor for NO x /NH 3 quantification

Sensors to detect mixtures of NO x /NH 3 are needed to monitor emissions of diesel automobiles where a selective catalytic reduction system uses an NH 3 mediated reaction to reduce NO x . We report on the application of a three electrode La 0.8 Sr 0.2 CrO 3 , Au 0.5 Pd 0.5 , Pt mixed potential sensor using yttria-stabilized-zirconia (YSZ) as a solid electrolyte to NO x /NH 3 sensing. Artificial neural networks were used to automatically decode the concentrations of NO x /NH 3 and errors of less than 15% are achieved. The optimal architecture for ANN decoding and the maximum density of training data points are also determined. The stability of the sensor was monitored by electrochemical impedance spectroscopy. The impedance associated with YSZ oxygen ion conduction and the electrochemical reactions at the three-phase interface are tracked for a period of over 100 days.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Design and analysis of optimal pre-cooling in residential buildings

Existing pre-cooling strategies provide a means of shifting or reducing the peak demand and/or energy cost in residential buildings. However, majority of them are rule-based and therefore may not be optimal in terms of cost saving, leaving room for improvement. In this paper, an integer linear programming problem that accounts for the thermal properties of a specific home, HVAC system capacity, utility rate structure, and weather conditions and makes use of a home thermal model is formulated. This problem determines the HVAC on/off control signal that minimizes the 24-h energy cost while maintaining thermal comfort and calculates the corresponding optimal indoor air temperature. The model is constructed using home thermal properties identified via data training in real-time. Through simulation, the energy performance of the proposed optimal pre-cooling strategy is investigated and compared with three rule-based operation strategies from the literature. It is found that the optimal strategy requires the least energy consumption without sacrificing thermal comfort. The superb energy performance of the optimal strategy is attributed to a longer runtime of the HVAC system in cool outdoor air conditions and to the elimination of deadband in HVAC operation, which is required by the rule-based strategies, to allow the indoor air temperature to stay near the thermal comfort upper bound as much as possible. In terms of energy cost, the rule-based operation strategies require 3.52, 1.90, and 2.79, respectively, while the optimal strategy only requires 1.52. These figures represent a saving of 56.82%, 20.00%, and 45.52%, respectively. The results suggest that the optimal strategy is indeed significantly more effective than the existing rule-based operation strategies.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2023

Neural message-passing for objective-based uncertainty quantification and optimal experimental design

Various real-world scientific applications involve the mathematical modeling of complex uncertain systems with numerous unknown parameters. Accurate parameter estimation is often practically infeasible in such systems, as the available training data may be insufficient and the cost of acquiring additional data may be high. In such cases, based on a Bayesian paradigm, we can design robust operators retaining the best overall performance across all possible models and design optimal experiments that can effectively reduce uncertainty to enhance the performance of such operators maximally. While objective-based uncertainty quantification (objective-UQ) based on MOCU (mean objective cost of uncertainty) provides an effective means for quantifying uncertainty in complex systems, the high computational cost of estimating MOCU has been a challenge in applying it to real-world scientific/engineering problems. In this work, we propose a novel scheme to reduce the computational cost for objective-UQ via MOCU based on a data-driven approach. We adopt a neural message-passing model for surrogate modeling, incorporating a novel axiomatic constraint loss that penalizes an increase in the estimated system uncertainty. As an illustrative example, we consider the optimal experimental design (OED) problem for uncertain Kuramoto models, where the goal is to predict the experiments that can most effectively enhance robust synchronization performance through uncertainty reduction. We show that our proposed approach can accelerate MOCU-based OED by four to five orders of magnitude, without any visible performance loss compared to the state-of-the-art. The proposed approach applies to general OED tasks, beyond the Kuramoto model.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Multi-Objective design of interlocking metasurfaces using conditional diffusion models

Unit cell design remains a major challenge for interlocking metasurfaces, a promising joining technology for dissimilar materials, due to the complex, competing, multivariate design space and the need for rapid adaptation to varying performance requirements. This study explores Conditional Diffusion Models as a design optimization tool for interlocking metasurfaces. Given the complex, competing, multivariate design space for interlocking metasurfaces, unit cell design remains a major challenge for this joining technology. We trained a conditional diffusion model on 25,000 finite element analysis-simulated interlocking metasurface unit cells to generate designs with tailored thermo-mechanical properties (tensile strength, shear strength, and thermal conductivity) based on specified performance criteria. The model demonstrated a success rate of approximately 72 % in producing designs that met specified property bounds. The conditional diffusion model generated both thermally resistive and conductive designs, revealing clear trends in design characteristics: taller, dendritic structures were advantageous for tensile loads, while shorter, robust designs excelled in shear applications. Our findings indicate that the model's performance is more influenced by the breadth of the design space than by the quantity of training data, highlighting the importance of expansive design domains for generating innovative solutions. This work establishes conditional diffusion models as a highly efficient and adaptable tool for rapid interlocking metasurface unit cell design, paving the way for advancements in multi-material joining technologies, as well as highlighting the justification to leverage conditional diffusion models as design tools across complex design domains.

Conditional diffusion models↗
DOE OSTI2015

An algebraic method for constructing stable and consistent autoregressive filters

In this paper, we introduce an algebraic method to construct stable and consistent univariate autoregressive (AR) models of low order for filtering and predicting nonlinear turbulent signals with memory depth. By stable, we refer to the classical stability condition for the AR model. By consistent, we refer to the classical consistency constraints of Adams–Bashforth methods of order-two. One attractive feature of this algebraic method is that the model parameters can be obtained without directly knowing any training data set as opposed to many standard, regression-based parameterization methods. It takes only long-time average statistics as inputs. The proposed method provides a discretization time step interval which guarantees the existence of stable and consistent AR model and simultaneously produces the parameters for the AR models. In our numerical examples with two chaotic time series with different characteristics of decaying time scales, we find that the proposed AR models produce significantly more accurate short-term predictive skill and comparable filtering skill relative to the linear regression-based AR models. These encouraging results are robust across wide ranges of discretization times, observation times, and observation noise variances. Finally, we also find that the proposed model produces an improved short-time prediction relative to the linear regression-based AR-models in forecasting a data set that characterizes the variability of the Madden–Julian Oscillation, a dominant tropical atmospheric wave pattern.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2020

Active- and transfer-learning applied to microscale-macroscale coupling to simulate viscoelastic flows

Active- and transfer-learning are applied to microscale dynamics of polymer flows for the multiscale discovery of effective constitutive approximations required in viscoelastic flow simulation. The result is macroscopic rheology directly connected to a microstructural model. Micro and macroscale simulations are adaptively coupled by means of Gaussian process regression (GPR) to run the expensive microscale computations only as necessary. This multiscale method is demonstrated with flows of a polymer solution as a model system. At the microscale level dissipative particle dynamics (DPD) is employed to model the fluid as a suspension of bead-spring micro-structures subjected to steady shear flow. The results yield the non-Newtonian viscosity and the first normal stress difference at strain rates as training data used in a GPR model. DPD parameters are calibrated with respect to experimental data for a real polymer solution. Compliance with these data requires adjustment of the DPD model's cutoff radius, which then becomes a function of the second invariant of the strain rate tensor. The FENE-P model is chosen for the macroscale description using the spectral element method (SEM) to simulate channel flow and flow past a circular cylinder. The DPD results at the lowest possible shear strain rate yield an estimate of the zero-shear rate viscosity, which allows the initiation of the macroscale flow by SEM as a Newtonian fluid. The resulting strain-rate field is surveyed to determine additional shear strain rate sampling points for the DPD system. This new information allows an initial fitting of parameters of the constitutive equation followed by new SEM simulations at the macroscale. Additionally, guided by active-learning GPR to select new sampling points, this process continues until convergence is achieved. The effectiveness of this new simulation paradigm for viscoelastic flows is tested with different macroscale operating conditions. The effective closure learned in the channel simulation is then transferred directly to the flow past a circular cylinder at low Reynolds number, where the results show that only two additional DPD simulations are required to achieve a satisfactory constitutive model. With an increase of the Reynolds number, the active-learning scheme automatically detects the inaccuracy of the learned constitutive model, and initiates additional DPD simulations for the extra data needed to once again close the microscale-macroscale coupled system. This new paradigm of active- and transfer-learning for multiscale modeling is readily applicable to other microscale-macroscale coupled simulations of complex fluids and other materials. Furthermore, the coupling between microscale and macroscale solvers can be seamlessly implemented with our open source multiscale universal interface (MUI) library.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2023

DRIPS: A framework for dimension reduction and interpolation in parameter space

Reduced-order models are often used to describe the behavior of complex systems, whose simulation with a full model is too expensive, or to extract salient features from the full model’s output. We introduce a new model-reduction framework DRIPS (dimension reduction and interpolation in parameter space) that combines the offline local model reduction with the online parameter interpolation of reduced-order bases (ROBs). The offline step of this framework relies on dynamic mode decomposition (DMD) to build a low-rank linear surrogate model, equipped with a local ROB, for quantities of interest derived from the training data generated by repeatedly solving the (nonlinear) high-fidelity model for multiple parameter points. The online step consists of the construction of a parametric reduced-order model for each target/test point in the parameter space, with the interpolation of ROBs done on a Grassman manifold and the interpolation of reduced-order operators done on a matrix manifold. The DMD component enables DRIPS to model (typically low-dimensional) quantities of interest directly, without having to access the (typically high-dimensional and possibly nonlinear) operators in a high-fidelity model that governs the dynamics of the underlying high-dimensional state variables, as required in projection-based reduced-order modeling. A series of numerical experiments suggests that DRIPS yields a model reduction, which is computationally more efficient than the commonly used projection-based proper orthogonal decomposition; it does so without requiring a prior knowledge of the governing equation for quantities of interest. Furthermore, for the nonlinear systems considered, DRIPS is more accurate than Gaussian-process interpolation (Kriging).

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2024

Physics-informed KNN milling stability model with process damping effects

This paper describes a k-nearest neighbors, or KNN, model for milling stability including process damping effects. A physics-based, frequency domain milling stability solution is used to generate the training data, but does not incorporate process damping effects. The data set is then updated using limited tests to capture the process damping behavior. A “stair step” approach is used to select the test points, where a first spindle speed-axial depth combination is selected based on the physics-based stability map, subsequent tests are defined using the previous test result, and data points are updated by knowledge of process damping behavior and the test results. Furthermore, the KNN modeling approach demonstrates the ability to predict both stable and unstable results, including process damping behavior.

42 ENGINEERING↗
DOE OSTI2017

The exploration of nonlinear elasticity and its efficient parameterization for crystalline materials

Conventional approaches to analyzing the very large coherency strains that can occur during solid-state phase transformations are founded in linear elasticity and rely on infinitesimal strain metrics. Despite this, there are many technologically important examples where misfit strains of multi-phase mixtures are very large during their synthesis and/or application. Here in this paper, we present a framework for constructing strain-energy expressions and stress-strain relationships beyond the linear-elastic limit for crystalline solids. This approach utilizes group theoretical concepts to minimize both the number of free parameters in the strain-energy expression and amount of first-principles training data required to parameterize strain-energy models that are invariant to all crystal symmetries. Within this framework, the strain-energy and elastic stiffness can be described to high accuracy in terms of a set of conventional symmetry-adapted finite strain metrics that we define independent of crystal symmetry. As an illustration, we use first-principles electronic structure data to parameterize strain energy polynomials and employ them to explore the strain- energy surfaces of HCP Zr and Mg, as well as several important Zr-H and Mg-Nd phases that are known to precipitate coherently within the HCP matrices of Zr and Mg.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Leveraging explainable AI to characterize floating-point exceptions in linear solvers

Linear solver packages are central to many scientific, engineering, and machine learning applications. When floating-point exceptions occur in these solvers, e.g., division by zero or overflow, numerical results are compromised and become unreliable. Existing static and dynamic analysis tools can detect such exceptions, but they do not explain why the exceptions occur in terms of the solver inputs. Here, we present a study to characterize the inputs that cause numerical exceptions in linear solver packages. Our approach uses explainable AI (XAI) to find the most relevant characteristics of input matrices that explain the occurrence of exceptions in the solvers. Since training data in this domain is scarce, we perform extensive data gathering and data augmentation to obtain exception-inducing inputs. Our approach uses a repair strategy on the features blamed by XAI to validate that such features indeed explain the exceptions. We compare the LIME and SHAP XAI techniques using a dozen matrix features with three classifiers. We evaluate the approach on three widely used linear solver packages and find that some input characteristics can explain the occurrence of exceptions 100% of the time, in specific solvers and preconditioners.

Explainable AI↗
DOE OSTI2025

Direct sensitivity analysis on the parameterization of crystal plasticity models

Various methods for calibrating crystal plasticity finite element (CPFE) models lead to non-unique input parameter values, which subsequently introduce uncertainty in the predicted mechanical response. Sensitivity analysis (SA) conducted on crystal plasticity models is used to identify how variability in these parameters contribute to output uncertainty. Traditional SA on CPFE parameters uses simplified surrogate models to save computational time. However, the accuracy of the surrogate models depends on the quantity of training data used, and any modeling error can propagate into the SA results, potentially affecting their reliability. In this work, the elementary effects test (EET) method, a global SA technique using direct CPFE simulations was employed, and the results obtained were compared with the First Order Second Moment (FOSM) method. ExaConstit, an open-source GPU-enabled CPFE code, was used to perform the simulations and direct SA. The EET method was accurately able to capture the non-linear effects of all the input parameters on the output and is a valuable approach for reliably attributing parameter sensitivities in CPFE models. Based on the results, efficient strategies to perform future parameter calibration and SA are discussed. Additionally, the SA trends observed in different single crystal orientations closely mirrored the activity of the slip systems.

Elementary Effects Test↗
DOE OSTI2025

An accelerated framework for predicting creep rupture lifetimes in engineering alloys

Confidently predicting high-temperature deformation, including creep and creep rupture, is paramount for the design and commercialization of candidate materials for advanced nuclear energy systems. To accelerate creep quantification, we introduce a framework that enables rapid, cost-effective, and reliable prediction of creep rupture lifetimes, minimizing reliance on time-intensive bulk creep testing. Unlike conventional creep analysis, which requires extensive time and resources, our method leverages a maximum of four short-term bulk creep tests as training data for prediction. This framework combines high-throughput nanoindentation up to 700 °C with these targeted bulk tests to inform our creep rupture model in order to predict rupture lifetimes. The strong agreement between our predictions and conventional experimental data demonstrates the effectiveness of our approach for accelerated creep analysis and lifetime prediction of structural components in high-temperature applications. Our multi-pronged approach motivates further integration of computational tools and advanced instrumentation to establish a universal framework for understanding high-temperature material responses.

36 MATERIALS SCIENCE↗
DOE OSTI2014

On the equivalence of the RTI and SVM approaches to time correlated analysis

Recently two papers on how to perform passive neutron auto-correlation analysis on time gated histograms formed from pulse train data, generically called time correlation analysis (TCA), have appeared in this journal [1,2]. For those of us working in international nuclear safeguards these treatments are of particular interest because passive neutron multiplicity counting is a widely deployed technique for the quantification of plutonium. The purpose of this letter is to show that the skewness-variance-mean (SVM) approach developed in [1] is equivalent in terms of assay capability to the random trigger interval (RTI) analysis laid out in [2]. Mathematically we could also use other numerical ways to extract the time correlated information from the histogram data including for example what we might call the mean, mean square, and mean cube approach. The important feature however, from the perspective of real world applications, is that the correlated information extracted is the same, and subsequently gets interpreted in the same way based on the same underlying physics model.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2021

Fast & accurate emulation of two-body scattering observables without wave functions

We combine Newton's variational method with ideas from eigenvector continuation to construct a fast & accurate emulator for two-body scattering observables. The emulator will facilitate the application of rigorous statistical methods for interactions that depend smoothly on a set of free parameters. Our approach begins with a trial K or T matrix constructed from a small number of exact solutions to the Lippmann–Schwinger equation. Subsequent emulation only requires operations on small matrices. We provide several applications to short-range potentials with and without the Coulomb interaction and partial-wave coupling. It is shown that the emulator can accurately extrapolate far from the support of the training data. When used to emulate the neutron-proton cross section with a modern chiral interaction as a function of 26 free parameters, it reproduces the exact calculation with negligible error and provides an over 300x improvement in CPU time.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2023

A Three-Dimensional, Analytical Wind Turbine Wake Model: Flow Acceleration, Empirical Correlations, and Continuity

A new, three-dimensional, analytical, steady-state wake model is presented that includes local flow acceleration near the rotor, improving the wake description compared to existing models. Wake structures such as the momentum deficit and regions of accelerated flow are concisely described with compound and normal Gaussian functions. Large-eddy simulations (LES) are used as training data to develop the model using two in-line turbines under various inflow conditions parameterized by hub-height wind speed and turbulence intensity. Mass conservation is considered by fixing two components of the wake velocity model and optimizing the third to best satisfy continuity; after which, the model performs comparably if not better than existing work with regards to both relative error and mass consistency. The final model demonstrates a high degree of flexibility, making use of empirical correlations to scale across different inflow conditions. The inclusion of these effects is capable of revealing unused opportunities for enhanced power generation by aligning wake trajectories with these regions of accelerated flow.

modeling↗
DOE OSTI2022

Surrogate modeling for efficiently, accurately and conservatively estimating measures of risk

We present a surrogate modeling framework for conservatively estimating measures of risk from limited realizations of an expensive physical experiment or computational simulation. Risk measures combine objective probabilities with the subjective values of a decision maker to quantify anticipated outcomes. Given a set of samples, we construct a surrogate model that produces estimates of risk measures that are always greater than their empirical approximations obtained from the training data. These surrogate models limit over-confidence in reliability and safety assessments and produce estimates of risk measures that converge much faster to the true value than purely sample-based estimates. We first detail the construction of conservative surrogate models that can be tailored to a stakeholder’s risk preferences and then present an approach, based on stochastic orders, for constructing surrogate models that are conservative with respect to families of risk measures. Our surrogate models include biases that permit them to conservatively estimate the target risk measures. We provide theoretical results that show that these biases decay at the same rate as the L 2 error in the surrogate model. Numerical demonstrations confirm that risk-adapted surrogate models do indeed overestimate the target risk measures while converging at the expected rate.

42 ENGINEERING↗
DOE OSTI2017

Cost-effective treatment of swine wastes through recovery of energy and nutrients

Highlights: • Cost-effective swine waste treatment that recovers energy/nutrients was evaluated. • Ion exchange and struvite precipitation were considered for nutrient recovery. • Generally, more than 90% of N, P, and K{sup +} nutrients were recovered. • Life Cycle Cost Analysis was performed, showing significant advantages. - Abstract: Wastes from concentrated animal feeding operations (CAFOs) are challenging to treat because they are high in organic matter and nutrients. Conventional swine waste treatment options in the U.S., such as uncovered anaerobic lagoons, result in poor effluent quality and greenhouse gas emissions, and implementation of advanced treatment introduces high costs. Therefore, the purpose of this paper is to evaluate the performance and life cycle costs of an alternative system for treating swine CAFO waste, which recovers valuable energy (as biogas) and nutrients (N, P, K{sup +}) as saleable fertilizers. The system uses in-vessel anaerobic digestion (AD) for methane production and solids stabilization, followed by struvite precipitation and ion exchange (IX) onto natural zeolites (chabazite or clinoptilolite) for nutrient recovery. An alternative approach that integrated struvite recovery and IX into a single reactor, termed STRIEX, was also investigated. Pilot- and bench-scale reactor experiments were used to evaluate the performance of each stage in the treatment train. Data from these studies were integrated into a life cycle cost analysis (LCCA) to assess the cost-effectiveness of various process alternatives. Significant improvement in water quality, high methane production, and high nutrient recovery (generally over 90%) were observed with both the AD-struvite-IX process and the AD-STRIEX process. The LCCA showed that the STRIEX system can provide considerable financial savings compared to conventional systems. AD, however, incurs high capital costs compared to conventional anaerobic lagoons and may require larger scales to become financially attractive.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗