Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Data Extrapolation”

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 433 records · Page 24
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Structure-aware graph neural network based deep transfer learning framework for enhanced predictive analytics on diverse materials datasets

Abstract Modern data mining methods have demonstrated effectiveness in comprehending and predicting materials properties. An essential component in the process of materials discovery is to know which material(s) will possess desirable properties. For many materials properties, performing experiments and density functional theory computations are costly and time-consuming. Hence, it is challenging to build accurate predictive models for such properties using conventional data mining methods due to the small amount of available data. Here we present a framework for materials property prediction tasks using structure information that leverages graph neural network-based architecture along with deep-transfer-learning techniques to drastically improve the model’s predictive ability on diverse materials (3D/2D, inorganic/organic, computational/experimental) data. We evaluated the proposed framework in cross-property and cross-materials class scenarios using 115 datasets to find that transfer learning models outperform the models trained from scratch in 104 cases, i.e., ≈90%, with additional benefits in performance for extrapolation problems. We believe the proposed framework can be widely useful in accelerating materials discovery in materials science.

Chemistry↗
DOE OSTI2021

The updated ITPA global H-mode confinement database: description and analysis

The multi-machine ITPA Global H-mode Confinement Database has been upgraded with new data from JET with the ITER-like wall and ASDEX Upgrade with the full tungsten wall. This paper describes the new database and presents results of regression analysis to estimate the global energy confinement scaling in H-mode plasmas using a standard power law. Various subsets of the database are considered, focusing on type of wall and divertor materials, confinement regime (all H-modes, ELMy H or ELM-free) and ITER-like constraints. Apart from ordinary least squares, two other, robust regression techniques are applied, which take into account uncertainty on all variables. Regression on data from individual devices shows that, generally, the confinement dependence on density and the power degradation are weakest in the fully metallic devices. Using the multi-machine scalings, predictions are made of the confinement time in a standard ELMy H-mode scenario in ITER. The uncertainty on the scaling parameters is discussed with a view to practically useful error bars on the parameters and predictions. One of the derived scalings for ELMy H-modes on an ITER-like subset is studied in particular and compared to the IPB98(y,2) confinement scaling in engineering and dimensionless form. Transformation of this new scaling from engineering variables to dimensionless quantities is shown to result in large error bars on the dimensionless scaling. Regression analysis in the space of dimensionless variables is therefore proposed as an alternative, yielding acceptable estimates for the dimensionless scaling. The new scaling, which is dimensionally correct within the uncertainties, suggests that some dependencies of confinement in the multi- machine database can be reconciled with parameter scans in individual devices. This includes vanishingly small dependence of confinement on line-averaged density and normalized plasma pressure (β), as well as a noticeable, positive dependence on effective atomic mass and plasma triangularity. Extrapolation of this scaling to ITER yields a somewhat lower confinement time compared to the IPB98(y, 2) prediction, possibly related to the considerably weaker dependence on major radius in the new scaling (slightly above linear). Further studies are needed to compare more flexible regression models with the power law used here. In addition, data from more devices concerning possible ‘hidden variables’ could help to determine their influence on confinement, while adding data in sparsely populated areas of the parameter space may contribute to further disentangling some of the global confinement dependencies in tokamak plasmas.

Database↗
DOE OSTI2021

Long-Term Creep-Rupture Behavior of Alloy Inconel 740/740H

To explore potential application of Ni-based alloys for power generation at the higher temperatures and pressures needed to achieve high thermal to electrical-energy conversion efficiency, an extensive creep-rupture dataset covering up to 875 °C and almost 70,000 h for Inconel 740/740H was analyzed using Larson–Miller parameter and Wilshire approaches. The results were used to assess the relative effectiveness of the two analytical methods, both in describing the experimental data and, because of the breadth of the dataset, using analyses of its shorter-time data to make creep lifetime predictions for much more extended times, which were then directly compared to the measured rupture times. The respective methods were also used to predict creep-limited lifetimes relevant to power production (that is, 100,000 h or greater). Despite the complexity of the precipitation-strengthened Inconel 740/740H alloy and the generalized parametric approach of these methodologies, the predictions based on such were reasonably accurate when the entire dataset was analyzed. However, when the analysis was confined to only data for conditions yielding creep-rupture times < 5,000 h( about 65 % of the entire dataset), the Wilshire correlation yielded better prediction for longer time lifetimes due to the inherent instability of the specific Larson–Miller formalism used in this analysis when extrapolated significantly outside its analysis range.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Atomic masses with machine learning for the astrophysical r process

The astrophysical r process plays a vital role in the production of heavy elements. Modeling of the r process is sensitive to masses and further requires knowledge of masses beyond current experimental reach. Therefore, simulations of the r process offer a unique test bed for predicting mass extrapolations. We take a Machine-Learning (ML) approach to model the masses across the entire chart of nuclides. For the first time, we simulate r-process nucleosynthesis with a physics-based ML mass model. We compare simulated abundances to solar data in order to evaluate the model's performance far from stability. The resulting r-process abundances up to thorium and uranium qualitatively match those of the observed solar system abundance pattern, with the characteristic peaks well positioned. We propagate the mass uncertainties obtained from the ML model to r-process abundance yields to estimate an uncertainty band associated with our approach. The size of the uncertainty band is approximately one order of magnitude which aligns with the uncertainty reported using alternative techniques.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Can a deep-learning model make fast predictions of vacancy formation in diverse materials?

The presence of point defects, such as vacancies, plays an important role in materials design. Here, we explore the extrapolative power of a graph neural network (GNN) to predict vacancy formation energies. We show that a model trained only on perfect materials can also be used to predict vacancy formation energies (E vac ) of defect structures without the need for additional training data. Such GNN-based predictions are considerably faster than density functional theory (DFT) calculations and show potential as a quick pre-screening tool for defect systems. To test this strategy, we developed a DFT dataset of 530 E vac consisting of 3D elemental solids, alloys, oxides, semiconductors, and 2D monolayer materials. We analyzed and discussed the applicability of such direct and fast predictions. We applied the model to predict 192 494 E vac for 55 723 materials in the JARVIS-DFT database. Our work demonstrates how a GNN-model performs on unseen data.

2D materials↗
DOE OSTI2023

Bayesian averaging for ground state masses of atomic nuclei in a Machine Learning approach

We present global predictions of the ground state mass of atomic nuclei based on a novel Machine Learning algorithm. We combine precision nuclear experimental measurements together with theoretical predictions of unmeasured nuclei. This hybrid data set is used to train a probabilistic neural network. In addition to training on this data, a physics-based loss function is employed to help refine the solutions. The resultant Bayesian averaged predictions have excellent performance compared to the testing set and come with well-quantified uncertainties which are critical for contemporary scientific applications. We assess extrapolations of the model’s predictions and estimate the growth of uncertainties in the region far from measurements.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2022

Experimental Studies and Energy Modeling for Evaluating Performance of Various Deep Wall Retrofits

The Pacific Northwest National Laboratory, Oak Ridge National Laboratory, and the University of Minnesota conducted a three-year, multipart study on residential retrofit wall assemblies. The project, which was funded by the U.S. Department of Energy’s Building Technology Office, aimed to compare a range of residential wall retrofit systems that prioritized affordability, durability, and energy savings potential. The research team identified, constructed, tested, simulated, and analyzed the feasibility and economics of 16 wall retrofit assemblies (14 test configurations and two baseline configurations) that can be applied to the exterior side of existing walls (either with or without the existing cladding). The 16 wall assemblies were installed in an in-situ laboratory environment, to evaluate the ease of construction and study the thermal and hygrothermal performance of the walls. This paper presents the methodology used to evaluate the thermal performance of the walls and discusses the energy modeling results of the study. The results from the experiments were used to calibrate a THERM model of each wall assembly, which was then applied to a whole building using the EnergyPlus 8.6 simulation engine. A residential prototype building was used to extrapolate whole-building energy savings in each U.S. climate zone. To capture the conditions of the largest number of homes in the United States, the most frequent building characteristics (e.g., attic insulation level, window specifications, foundation insulation, etc.) were extracted from ResStock data and applied to the prototype model. Results from the energy modeling showed that the climate zones with the highest potential for retrofit savings are those which are heating-dominated (i.e., Cold and Very Cold climate designations). In these climate zones, heating and cooling energy savings due to the wall retrofits alone ranged from 21.5% to 38.2%.

Nagda, Harshil↗
DOE OSTI2020

Calibration of CDM-Based Creep Constitutive Model Using Accelerated Creep Test (ACT) Data

In conventional creep testing (CCT) a specimen is subject to constant load and temperature for a long-duration until creep rupture occurs. Conventional testing can be costly when considering the number of experiments needed to characterize the creep response of a material over a range of stress and temperature. To predict long-term creep-rupture properties, the time-temperature-stress superposition principle (TTSSP) approach has been employed where stress and/or temperature is applied at an elevated level; the result of which are extrapolated down to low stress and/or temperature conditions. These methods have been successful in predicting minimum-creep-strain-rate (MCSR) and stress-rupture (SR) but suffer from an inability to predict the creep deformation curve or account for changes in deformation mechanisms or aging that occurs at long-duration. An accelerated technique, termed the Stepped isostress method (SSM) allows the accelerated testing of materials to determine their creep deformation response. Unlike TTSSP tests, the SSM test employs a single specimen where the stress is periodically step increased until rupture. The SSM creep deformation curve is processed (time and strain shifted) to produce an accelerated creep deformation curve that represent the creep deformation curve at the initial stress level in SSM. A processing procedure for metals has yet to be developed. The research objective of this study is to develop a processing procedure for SSM test data using a creep-damage constitutive model. Triplicate SSM tests were conducted on Ni-based superalloy Inconel 718 at 650°C with stress being periodically increased until rupture. Triplicate CCT tests were conducted at the initial stress level of the SSM tests. The Sine-hyperbolic (Sinh) creep-damage model was employed in this study. The Sinh creep-damage constitutive model is based on coupled creep strain rate and damage evolution equations; where both rates are dependent on the current state of damage. Calibration is two-step: analytical and numerical optimization. Each stepped creep deformation curve is tackled quasi-analytically to determine MCSR and SR related material constants and accumulated damage. The damage accumulated at the end of each step was then passed onto subsequent steps to calibrate the MCSR, rupture prediction, and damage evolution. Numerical optimization was applied to optimize model constants involved in the creep strain constitutive equations in order to generate best-fitted Sinh creep deformation curves. The Sinh model predictions were compared to the SSM and CCT data. The Sinh model satisfactorily predicts the SSM data and thus the calibrated material constants provides a good estimate of rupture found in the CCT data. Calibration using SSM data reduces the number of tests needed to calibrate a model; significantly reducing costs. A single SSM test replaces numerous creep tests at different stresses.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Improved accuracy and transferability of molecular-orbital-based machine learning: Organics, transition-metal complexes, non-covalent interactions, and transition states

We report molecular-orbital-based machine learning (MOB-ML) provides a general framework for the prediction of accurate correlation energies at the cost of obtaining molecular orbitals. The application of Nesbet’s theorem makes it possible to recast a typical extrapolation task, training on correlation energies for small molecules and predicting correlation energies for large molecules, into an interpolation task based on the properties of orbital pairs. We demonstrate the importance of preserving physical constraints, including invariance conditions and size consistency, when generating the input for the machine learning model. Numerical improvements are demonstrated for different datasets covering total and relative energies for thermally accessible organic and transition-metal containing molecules, non-covalent interactions, and transition-state energies. MOB-ML requires training data from only 1% of the QM7b-T dataset (i.e., only 70 organic molecules with seven and fewer heavy atoms) to predict the total energy of the remaining 99% of this dataset with sub-kcal/mol accuracy. This MOB-ML model is significantly more accurate than other methods when transferred to a dataset comprising of 13 heavy atom molecules, exhibiting no loss of accuracy on a size intensive (i.e., per-electron) basis. It is shown that MOB-ML also works well for extrapolating to transition-state structures, predicting the barrier region for malonaldehyde intramolecular proton-transfer to within 0.35 kcal/mol when only trained on reactant/product-like structures. Finally, the use of the Gaussian process variance enables an active learning strategy for extending the MOB-ML model to new regions of chemical space with minimal effort. We demonstrate this active learning strategy by extending a QM7b-T model to describe non-covalent interactions in the protein backbone–backbone interaction dataset to an accuracy of 0.28 kcal/mol.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Lessons Learned in Employing Data Analytics to Predict Oxidation Kinetics and Spallation Behavior of High-Temperature NiCr-Based Alloys

Machine learning (ML) can offer many advantages in predicting material properties over traditional materials development methods based solely on limited experimental investigations or physical-based simulations with the capability to reduce development cost, risk, and time. However, so far, limited efforts have been made to predict alloy oxidation kinetics and spallation behavior via ML due to the lack of consistently measured and sufficient experimental data and the inherent complexity in oxidation behavior of multicomponent high-temperature alloys. A previous study reported the ability of ML to predict oxidation kinetics of NiCr-based alloys as a function of alloy composition and operating conditions. Here, the performance of a ML model in predicting rate constants and spallation probability was evaluated in light of the roles of the data distribution of the experimental dataset (data analytics), the alloy composition, the exposure environment and the chosen oxidation approach to extracting kinetic values from the measured mass changes (but using either a simple parabolic law or a statistical cyclic oxidation model). Potential strategies to improve the predictions and enhance the extrapolative capability of the previously trained model will be discussed.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Nuclear Safeguards: Feature Extraction for Machine Learning Enrichment Analysis of UF 6 Cylinders

With the increasing international interest in using nuclear material as sources of energy, comes the growing concern that we may see an increase in proliferation threats. Uranium hexafluoride (UF 6 ) is used in the nuclear fuel cycle for uranium enrichment. Inspectors from the International Atomic Energy Agency (IAEA) monitor the enrichment levels of UF 6 stored in transportation cylinders, to ensure the enrichments match that of a facility operator’s declarations. However, only a characteristic subset of UF 6 cylinders can be measured by inspectors from the IAEA during these inspections. In turn, the inspector may not identify a cylinder whose enrichment levels do not match the facility declarations. Therefore, our team attempts to develop a machine learning network that could determine the enrichment percentage of cylinders as they enter and exit facilities. The machine learning model must be robust against spectral variations due to factors that are internal and external to the UF 6 cylinder. Such factors include but are not limited to the speed and distance of the moving vehicle, cylinder type, cylinder orientation, and fill level. Many features in the spectral continuum can be used to identify and correct for some of these variations. As a consequence, the model must extract numerous features in the continuum. We investigate new approaches in continuum subtraction to verify enrichment percentage: interpolation and extrapolation of lines at the notable characteristic spectral peaks. It was determined that there was no clear answer on which method proved superior, therefore the decision was made to implement both new methods into the current codebase. We currently rely on synthetic training and testing data to analyze the results and fine tune our models but hope to test the system on measured data soon.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗
DOE OSTI2022

Accurate localization of Kosterlitz-Thouless-type quantum phase transitions for one-dimensional spinless fermions

We investigate the charge-density wave (CDW) transition for one-dimensional spinless fermions at half band filling with nearest-neighbor electron transfer amplitude t and interaction V. The model is equivalent to the anisotropic XXZ Heisenberg model for which the Bethe Ansatz provides an exact solution. For V>V c =2t, the CDW order parameter and the single-particle gap are finite but exponentially small, as is characteristic for a Kosterlitz-Thouless transition. It is notoriously difficult to locate such infinite-order phase transitions in the phase diagram using approximate analytical and numerical approaches. Second-order Hartree-Fock theory is qualitatively applicable for all interaction strengths, and predicts the CDW transition to occur at V$^{(2)}_{c,2}$≈1.5t. Second-order Hartree Fock theory is almost variational because the density of quasiparticle excitations is small. We apply the density-matrix renormalization group (DMRG) for periodic boundary conditions for system sizes up to 514 sites, which permits a reliable extrapolation of all physical quantities to the thermodynamic limit, apart from the critical region. We investigate the ground-state energy, the gap, the order parameter, the momentum distribution, the quasiparticle density, and the density-density correlation function to locate V c from the DMRG data. In conclusion, tracing the breakdown of the Luttinger liquid and the peak in the quasiparticle density at the band edge permits us to reproduce V c with an accuracy of one percent.

1-dimensional spin chains↗
DOE OSTI2024

Solar PV Transmission Setbacks: Ordinances (2022) and Extrapolated Trends

This dataset represents solar energy setback requirements from transmission. A setback requirement is a minimum distance from transmission infrastructure that an energy project may be developed. As of April 2022, no ordinances were discovered for any counties. Such ordinances are likely to arise as regulations continue to expand. Therefore, this dataset applies a 30-meter setback, sourced from trends in other infrastructure. A TIF data file and a PNG map of the data are provided, showing areas where solar energy is prohibited or permitted across the contiguous United States. For further details and citation, please refer to the publication linked below: Lopez, Anthony, Pavlo Pinchuk, Michael Gleason, Wesley Cole, Trieu Mai, Travis Williams, Owen Roberts, Marie Rivers, Mike Bannister, Sophie-Min Thomson, Gabe Zuckerman, and Brian Sergi. 2024. Solar Photovoltaics and Land-Based Wind Technical Potential and Supply Curves for the Contiguous United States: 2023 Edition. Golden, CO: National Renewable Energy Laboratory. NREL/TP-6A20-87843.

Array↗
DOE OSTI2024

Solar PV Railroad Setbacks: Ordinances (2022) and Extrapolated Trends

This dataset represents solar energy setback requirements from railroads. A setback requirement is a minimum distance from a railroad that an energy project may be developed. As of April 2022, no ordinances were discovered for any counties. Such ordinances are likely to arise as regulations continue to expand. Therefore, this dataset applies a 30-meter setback, sourced from trends in other infrastructure. A TIF data file and a PNG map of the data are provided, showing areas where solar energy is prohibited or permitted across the contiguous United States. For further details and citation, please refer to the publication linked below: Lopez, Anthony, Pavlo Pinchuk, Michael Gleason, Wesley Cole, Trieu Mai, Travis Williams, Owen Roberts, Marie Rivers, Mike Bannister, Sophie-Min Thomson, Gabe Zuckerman, and Brian Sergi. 2024. Solar Photovoltaics and Land-Based Wind Technical Potential and Supply Curves for the Contiguous United States: 2023 Edition. Golden, CO: National Renewable Energy Laboratory. NREL/TP-6A20-87843.

Array↗
DOE OSTI2020

Measurement of the cross section for $\text{t}\bar{\text{t}}$ production with additional jets and b jets in pp collisions at $\sqrt{s}=$ 13 TeV

Measurements of the cross section for the production of top quark pairs in association with a pair of jets from bottom quarks $ \left({\sigma}_{\mathrm{t}\overline{\mathrm{t}}\mathrm{b}\overline{\mathrm{b}}}\right) $ and in association with a pair of jets from quarks of any flavor or gluons $ \left({\sigma}_{\mathrm{t}\overline{\mathrm{t}}\mathrm{jj}}\right) $ and their ratio are presented. The data were collected in proton-proton collisions at a center-of-mass energy of 13 TeV by the CMS experiment at the LHC in 2016 and correspond to an integrated luminosity of 35.9 fb$^{−1}$. The measurements are performed in a fiducial phase space and extrapolated to the full phase space, separately for the dilepton and lepton+jets channels, where lepton corresponds to either an electron or a muon. The results of the measurements in the fiducial phase space for the dilepton and lepton+jets channels, respectively, are $ {\sigma}_{\mathrm{t}\overline{\mathrm{t}}\mathrm{jj}} $ = 2.36±0.02 (stat)±0.20 (syst) pb and 31.0±0.2 (stat)±2.9 (syst) pb, and for the cross section ratio 0.017 ± 0.001 (stat) ± 0.001 (syst) and 0.020 ± 0.001 (stat) ± 0.001 (syst). The values of $ {\sigma}_{\mathrm{t}\overline{\mathrm{t}}\mathrm{b}\overline{\mathrm{b}}} $ are determined from the product of the $ {\sigma}_{\mathrm{t}\overline{\mathrm{t}}\mathrm{jj}} $ and the cross section ratio, obtaining, respectively, 0.040±0.002 (stat)±0.005 (syst) pb and 0.62±0.03 (stat)±0.07 (syst) pb. These measurements are the most precise to date and are consistent, within the uncertainties, with the standard model expectations obtained using a matrix element calculation at next-to-leading order in quantum chromodynamics matched to a parton shower.[graphic not available: see fulltext]

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2025

Spectrophotometric determination of the stability of La hydroxyl complexes at near neutral to alkaline pH from 25 to 75 °C

The hydrolysis of rare earth elements (REE) potentially controls their mobility during fluid-rock interaction in a broad range of pH and temperature conditions. However, there is still a lack of thermodynamic data for modeling accurately the stability of REE hydroxyl complexes in hydrothermal aqueous fluids. Here, in this study, UV–Vis spectrophotometric experiments were conducted from 25 to 75 °C in near-neutral to alkaline NaOH-bearing aqueous solutions with varying lanthanum (La) concentrations (0 to ∼0.23 mmol/kg). The color indicator m-cresol purple was used to determine in situ pH and derive the average OH− ligand number ($\overrightarrow{n}$) and formation constants for the La hydroxyl complexes (LaOH 2+ , La(OH) 2 + , and La(OH) 3 0 ). From 25 to 50 °C, $\overrightarrow{n}$ ranges between ∼1 and 2 at pH from 7.0 to 9.3. At 75 °C, $\overrightarrow{n}$ ranges between ∼1.5 and 3 at pH from 6.3 to 8.8. These results suggest the predominance of LaOH 2+ and La(OH) 2 + complexes from 25 to 50 °C, and an increased predominance of La(OH) 3 0 at 75 °C. The cumulative formation constants (β n °, n = 1 to 3) are derived for the reaction La 3+ + nOH − = La(OH) n 3-n , and fitted between 25 and 250 °C by combining the UV–Vis and literature solubility data. The resulting logβ n ° are expressed as function of temperature (T in Kelvin): logβ 1 ° = −1.786 + 0.0133 T + 1.049·10 3 /T; logβ 2 ° = −5.797 + 0.0267 T + 2.713·10 3 /T; logβ 3 ° = 6.435 + 0.0223 T + 512.7/T. A comparison between these new fits and existing extrapolations using the Helgeson-Kirkham-Flowers equation of state indicates significant differences in the predicted hydrolysis of La. The latter extrapolations should therefore be updated for the hydrolysis of REE.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Efficient and generalizable nested Fourier-DeepONet for three-dimensional geological carbon sequestration

Geological carbon sequestration (GCS) involves injecting CO2 into subsurface geological formationsfor permanent storage. Numerical simulations could guide decisions in GCS projects by predictingCO 2 migration pathways and the pressure distribution in storage formation. However, these simula-tions are often computationally expensive due to highly coupled physics and large spatial-temporalsimulation domains. Surrogate modelling with data-driven machine learning has become a promis-ing alternative to accelerate physics-based simulations. Among these, the Fourier neural operator(FNO) has been applied to three-dimensional synthetic subsurface models. Despite its good accuracyin simulating CO 2 plume migration, it requires large computational resources in training and alsolacks generalizability. Here, to further improve performance, we have developed a nested Fourier-DeepONet by combining the expressiveness of the FNO with the modularity of a deep operatornetwork (DeepONet). This new framework is twice as efficient as a nested FNO for training and has atleast 80% lower GPU memory requirement due to its flexibility to treat temporal coordinates sepa-rately. These performance improvements are achieved without compromising prediction accuracy.In addition, the generalization and extrapolation ability of nested Fourier-DeepONet beyond thetraining range has been thoroughly evaluated. Nested Fourier-DeepONet outperformed the nestedFNO for extrapolation in time with more than 50% reduced error. It also exhibited good extrapolationaccuracy beyond the training range in terms of reservoir properties, number of wells, and injectionrate.

Lee, Jonathan E. [Department of Chemical and Envir↗
DOE OSTI2023

Dynamic Model Agnostic Reliability Evaluation of Machine-Learning Models Integrated in Instrumentation & Control Systems

In recent years, the field of data-driven neural network-based machine learning (ML) algorithms has grown significantly and spurred research in its applicability to instrumentation and control systems. While they are promising in operational contexts, the trustworthiness of such algorithms is not adequately assessed. Failures of ML-integrated systems are poorly understood; the lack of comprehensive risk modeling can degrade the trustworthiness of these systems. In recent reports by the National Institute for Standards and Technology, trustworthiness in ML is a critical barrier to adoption and will play a vital role in intelligent systems' safe and accountable operation. Thus, in this work, we demonstrate a real-time model-agnostic method to evaluate the relative reliability of ML predictions by incorporating out-of-distribution detection on the training dataset. It is well documented that ML algorithms excel at interpolation (or near-interpolation) tasks but significantly degrade at extrapolation. This occurs when new samples are "far" from training samples. The method, referred to as the Laplacian distributed decay for reliability (LADDR), determines the difference between the operational and training datasets, which is used to calculate a prediction's relative reliability. LADDR is demonstrated on a feedforward neural network-based model used to predict safety significant factors during different loss-of-flow transients. LADDR is intended as a "data supervisor" and determines the appropriateness of well-trained ML models in the context of operational conditions. Ultimately, LADDR illustrates how training data can be used as evidence to support the trustworthiness of ML predictions when utilized for conventional interpolation tasks.

97 MATHEMATICS AND COMPUTING↗