Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “model interpretability”

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 361 records · Page 20
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Soil incubation methods lead to large differences in inferred methane production temperature sensitivity

Abstract Quantifying the temperature sensitivity of methane (CH 4 ) production is crucial for predicting how wetland ecosystems will respond to climate warming. Typically, the temperature sensitivity (often quantified as a Q 10 value) is derived from laboratory incubation studies and then used in biogeochemical models. However, studies report wide variation in incubation-inferred Q 10 values, with a large portion of this variation remaining unexplained. Here we applied observations in a thawing permafrost peatland (Stordalen Mire) and a well-tested process-rich model ( ecosys ) to interpret incubation observations and investigate controls on inferred CH 4 production temperature sensitivity. We developed a field-storage-incubation modeling approach to mimic the full incubation sequence, including field sampling at a particular time in the growing season, refrigerated storage, and laboratory incubation, followed by model evaluation. We found that CH 4 production rates during incubation are regulated by substrate availability and active microbial biomass of key microbial functional groups, which are affected by soil storage duration and temperature. Seasonal variation in substrate availability and active microbial biomass of key microbial functional groups led to strong time-of-sampling impacts on CH 4 production. CH 4 production is higher with less perturbation post-sampling, i.e. shorter storage duration and lower storage temperature. We found a wide range of inferred Q 10 values (1.2–3.5), which we attribute to incubation temperatures, incubation duration, storage duration, and sampling time. We also show that Q 10 values of CH 4 production are controlled by interacting biological, biochemical, and physical processes, which cause the inferred Q 10 values to differ substantially from those of the component processes. Terrestrial ecosystem models that use a constant Q 10 value to represent temperature responses may therefore predict biased soil carbon cycling under future climate scenarios.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Machine learning and deep learning for mineralogy interpretation and CO 2 saturation estimation in geological carbon Storage: A case study in the Illinois Basin

Carbon capture and storage (CCS) is a promising approach to simultaneously maintaining energy security and reducing carbon dioxide (CO 2 ) emissions under the current energy portfolio that is dominated by fossil fuel energy. Pre-injection formation characterization and post-injection CO 2 monitoring are two critical tasks to guarantee storage efficiency in CCS. The CCS projects in the Illinois Basin, the first large-scale CO 2 injection into saline aquifers in the United States, employed conventional and the latest pulsed neutron logging (PNL) tools for mineralogy interpretation and CO 2 saturation estimation, which provide valuable references for future CCS projects. Because of the inherent fuzziness of petrophysical measurements and complex subsurface heterogeneity, interpreting well-logging data is time-consuming, and its accuracy can be user-biased. In recent years, data-driven methods have been widely used to capture the non-linear patterns between input features and interpretation results. This work applied and evaluated four commonly used machine learning (ML) models, including ridge regression (RR), random forest (RF), gradient boosting regression (GBR), support vector regression (SVR), and one deep learning (DL) model, the artificial neural network (ANN). We optimized the hyperparameters of the four ML models and the DL model using the simulated annealing algorithm and the grid search strategy, respectively. The input features of the mineralogy interpretation models were eleven conventional well-logging parameters, and the label data (i.e., ground truth) were the porosity and volumetric fractions of six minerals, including quartz, feldspar, dolomite, calcite, clay, and iron minerals. The results demonstrated that the GBR and RF models were superior in predicting volumetric fractions of minerals and porosity; label data with low coefficient of variation (CV) values tended to yield better performance. For CO 2 saturation estimation, the RF was the best-performing model, followed by SVR, ANN, GBR, and RR. Furthermore, we conducted feature importance ranking using the permutation importance algorithm and found that the formation sigma and well pressure were the most important features in this study. In conclusion, the study of CCS projects in the Illinois Basin bridges the gap between the limited knowledge and understanding of geological carbon storage and the increasing demand for reliable, cost-effective, and sustainable energy solutions.

58 GEOSCIENCES↗
DOE OSTI2025

Systematic study of the validity of the eikonal model including uncertainties

Nuclear reactions at intermediate beam energies are often interpreted using the eikonal model. In the analysis of complex reaction probes, where few-body reaction methods are needed, the eikonal method may be used as an efficient way for describing the fragment-target reaction process. In this work, we perform a systematic study to test the validity of the eikonal approximation for nucleon-nucleus reactions. We also quantify uncertainties due to the nucleon optical potential on reaction observables. We inspect the validity of the eikonal model and its semiclassical correction by comparing it to exact solutions (obtained from solving the optical-model equation with a finite-differences method) for a wide range of reactions. We also study the effect of relativistic corrections, both kinematic and dynamic, by effectively incorporating the relativistic effects at intermediate energies. The uncertainties from a Bayesian global optical potential (KDUQ) are propagated to the observables of interest. Our study includes neutron and proton reactions on 27 Al , 40 Ca , 90 Zr , and 208 Pb , for a wide range of energies 𝐸 lab = 0–400 MeV. We calculate neutron-total cross sections (elastic and reactions) as well as proton-absorption cross sections as a function of beam energy, using the eikonal model, the eikonal model with a semiclassical correction, and the exact solution. Here, we also compute angular distributions for the methods above. Our results show that for the proton-absorption cross section, the eikonal model can be used down to around 60 MeV and the semiclassical correction extends its use to 30 MeV. However, the validity of the eikonal model for the neutron-total cross section only goes down to ≈120 MeV, a range extended to ≈ 50 MeV when using the semiclassical correction. We find the semiclassical correction to the eikonal model to be less effective in describing the angular distributions. The 1⁢𝜎 uncertainty intervals on the observables we studied is less than 5% for most of the energies considered, but increases rapidly for higher energies, namely energies outside the range of KDUQ (𝐸 lab > 200MeV).

Cluster models↗
DOE OSTI2023

Differentiable physics-enabled closure modeling for Burgers’ turbulence

Abstract Data-driven turbulence modeling is experiencing a surge in interest following algorithmic and hardware developments in the data sciences. We discuss an approach using the differentiable physics paradigm that combines known physics with machine learning to develop closure models for Burgers’ turbulence. We consider the one-dimensional Burgers system as a prototypical test problem for modeling the unresolved terms in advection-dominated turbulence problems. We train a series of models that incorporate varying degrees of physical assumptions on an a posteriori loss function to test the efficacy of models across a range of system parameters, including viscosity, time, and grid resolution. We find that constraining models with inductive biases in the form of partial differential equations that contain known physics or existing closure approaches produces highly data-efficient, accurate, and generalizable models, outperforming state-of-the-art baselines. Addition of structure in the form of physics information also brings a level of interpretability to the models, potentially offering a stepping stone to the future of closure modeling.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Code for the manuscript titled "Blackout Diffusion: Generative Diffusion Models in Discrete-State Spaces"

We would like to disclose two scripts, written in Jupyter notebook, in which we implement the "Blackout Diffusion Process" described in the manuscript "Blackout Diffusion: Generative Diffusion Models in Discrete-State Spaces" (LA-UR-23-20509), to be submitted to the International Conference of Machine Learning (ICML). The abstract of the manuscript is append below. == Typical generative diffusion models rely on a Gaussian diffusion process for training the backward transformations, which can then be used generate samples from Gaussian noise. However, real world data often takes place in discrete-state spaces, which occur in many scientific applications. Here, we develop a theoretical formulation for arbitrary discrete-state Markov processes in the forward diffusion process. We relate the theory to the existing continuous-state Gaussian diffusion and identify the corresponding reverse-time stochastic process and score function in the continuous-time setting, and the reverse-time mapping in the discrete-time setting. As an example of this framework, we introduce "Blackout Diffusion", which learns to produce samples from an empty image instead of from noise. Numerical experiments on the CIFAR-10 dataset confirm the feasibility of generative diffusion modeling in a discrete space. Generalizing from specific (Gaussian) forward processes to a more general framework also sheds light on how to interpret generative diffusion models and their mathematical structure, which we comment on.

Lin, Yen Ting↗
DOE OSTI2022

Genetic programming for understanding cognitive biases that generate polarization in social networks

Recent studies have applied agent-based models to infer human-interpretable explanations of individual-scale behaviors that generate macro-scale patterns in complex social systems. Genetic programming has proven to be an ideal explainable AI tool for this purpose, where primitives may be expressed in an interpretable fashion and assembled into agent rules. Evolutionary model discovery (EMD) is a tool that combines genetic programming and random forest feature importance analysis, to infer individual-scale, human-interpretable explanations from agent-based models. We deploy EMD to investigate the cognitive biases behind the emergence of ideological polarization within a population. An agent-based model is developed to simulate a social network, where agents are able to create or sever links with one another, and update an internal ideological stance based on their neighbors' stances. Agent rules govern these actions and constitute of cognitive biases. A set of 7 cognitive biases are included as genetic program primitives in the search for rules that generate hyper-polarization among the population of agents. We find that heterogeneity in cognitive biases is more likely to generate polarized social networks. Highly polarized social networks are likely to emerge when individuals with confirmation bias are exposed to those with either attentional bias, egocentric bias, or cognitive dissonance.

Gunaratne, Chathika↗
DOE OSTI2021

The ezSpectra suite: An easy‐to‐use toolkit for spectroscopy modeling

Abstract A molecule's spectrum encodes information about its structure and electronic properties. It is a unique fingerprint that can serve as a molecular ID. Quantum chemistry calculations provide key ingredients for interpreting spectra, but modeling the spectra rarely ends there; it requires additional steps that entail combined treatments of electronic and nuclear degrees of freedom and account for specifics of the experimental setup (light energy, polarization, averaging over molecular orientations, temperature, etc.). This Software Focus article describes the ezSpectra suite, which currently comprises two stand‐alone open‐source codes: ezFCF and ezDyson . ezFCF calculates Franck–Condon factors, which yield vibrational progressions for polyatomic molecules, within the double‐harmonic approximation. ezDyson calculates absolute cross‐sections for photodetachment/photoionization processes and photoelectron angular distributions using Dyson orbitals computed by a quantum chemistry program. This article is categorized under: Electronic Structure Theory > Ab Initio Electronic Structure Methods Theoretical and Physical Chemistry > Spectroscopy Software > Simulation Methods

Gozem, Samer↗
DOE OSTI2025

Predicting the Evolution of Shallow Cumulus Clouds With a Lotka‐Volterra Like Model

Abstract In numerical weather prediction and climate models, boundary‐layer clouds are controlled by a wide range of subgrid‐scale processes. However, understanding the nature of these processes and their role in the evolution of the cloud size distribution as a whole has been elusive. To address this issue, we adopt a novel empirical framework from the field of population dynamics to model the evolution of cloud size statistics by using the shallow cumulus properties obtained from a large‐eddy simulation (LES). Our approach involves representing the cloud size distribution and the total cloud area using a revised Lotka‐Volterra model and ridge linear model, respectively. The physical interpretation of the total cloud area and coefficients obtained from the optimization of the models reveals three stages probably interpreted by dominant processes: the formation of new clouds, the growth of single clouds, and a steady state with organized transitions involving the growth and decay of multiple clouds. Furthermore, we showcase the potential of this framework to serve as a component of scale‐aware parameterizations of shallow‐convective clouds in atmospheric models.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Search for chargino-neutralino production in events with Higgs and W bosons using 137 fb$^{−1}$ of proton-proton collisions at $ \sqrt{s} $ = 13 TeV

A search for electroweak production of supersymmetric (SUSY) particles in final states with one lepton, a Higgs boson decaying to a pair of bottom quarks, and large missing transverse momentum is presented. The search uses data from proton-proton collisions at a center-of-mass energy of 13 TeV collected using the CMS detector at the LHC, corresponding to an integrated luminosity of 137 fb$^{−1}$. The observed yields are consistent with backgrounds expected from the standard model. The results are interpreted in the context of a simplified SUSY model of chargino-neutralino production, with the chargino decaying to a W boson and the lightest SUSY particle (LSP) and the neutralino decaying to a Higgs boson and the LSP. Charginos and neutralinos with masses up to 820 GeV are excluded at 95% confidence level when the LSP mass is small, and LSPs with mass up to 350 GeV are excluded when the masses of the chargino and neutralino are approximately 700 GeV.[graphic not available: see fulltext]

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2020

A comparison of computational models for predicting yield sooting index

Sooting propensity, a measurement of how much particulate matter is produced when a fuel is burned, is a property of significant interest among researchers who are striving to discover the next generation of cleaner, more efficient fuels and fuel additives. Many compounds are not viable as fuels and/or fuel additives, and as a result, designing cleaner-burning biofuels using only experimental techniques is inefficient. Predictive models have been instrumental in reducing this inherent difficulty, providing researchers with a tool to preemptively screen compounds before production and testing. The present work compares the accuracies and interpretabilities of existing models used to predict a particular measure of sooting propensity, Yield Sooting Index (YSI). These models include artificial neural networks, graph neural networks, and multivariate equations. A novel equation for predicting YSI based on atom path count and bond order is proposed, which can highlight key structural components that contribute to YSI. It was found that artificial neural networks slightly outperform graph neural networks and greatly outperform multivariate equations in blind (test set) prediction accuracy; however, graph neural networks and multivariate equations provide significantly more interpretability as to how compound structure relates to YSI. Predictions of YSI are compared to experimental measurements for previously un-tested compounds with cetane numbers comparable to diesel fuel (50-60) (butyl decanoate, ethyl decanoate, 1,4-bis(ethenoxymethyl)cyclohexane, and 5-heptyloxolan-2-one), and it was found that these compounds produce significantly less soot compared to diesel fuel.

09 BIOMASS FUELS↗
DOE OSTI2022

A reduced model for the ITER divertor based on SOLPS solutions for ITER Q = 10 baseline conditions: B. A reduced model based on reversed-direction two point modeling

Edge codes such as SOLPS coupled to neutral codes such as EIRENE have become so comprehensive and sophisticated that they now constitute, in effect, 'code-experiments' that, as for actual experiments, can benefit from interpretation using simple models and conceptual frameworks, i.e. reduced models. The first task is the identification of options for the reduced model control parameters that are best suited for control of the action of the divertor, i.e. for control of target power loading and sputter–erosion, primarily. A strong correlation between the electron temperature at the divertor target, T e,t , and the neutral deuterium D 2 density at the target, n D2,t , flux-tube resolved, has recently been reported for a number of code studies including SOLPS-4.3 modeling of a set of ~50 ITER baseline cases: Q DT = 10, q 95 = 3, P SOL = 100 MW, metallic walls, and Ne seeding (Pitts et al 2019). Part A of the present study reports new results for largely the same ITER cases, confirming the strong correlation reported earlier between local values of T e,t , and (i) n D2,t , and (ii) normalized volumetric losses of power and pressure in the divertor. Strong correlations have now also been found, and are reported here for the first time, between T e,t and all of the divertor target quantities of practical interest. A physical explanation for this surprising result has not as yet been fully identified; nevertheless it has encouraging implications for reduced modeling of the ITER divertor. For such ITER conditions, (i) the global Ne injection rate, Inj Ne (Ne s –1 ), and (ii) the electron temperature at the location on the target where the peak power deposition occurs, T e,t @q ⊥,pk (eV), are found to be promising reduced model control parameters. In this part B, a reduced model for the ITER divertor is developed and described in detail, based on reversed-direction two point modeling, Rev2PM. The input to the reduced model is a value of the variable pair (T e,t @q ⊥,pk , Inj Ne )for a chosen case and the output are values of the various target as well as divertor-entrance quantities of practical interest, e.g. q ⊥,pk , the electron density at the X-point, n e,Xpt , etc. The reduced model was quantitatively characterized using one half of the code cases; it was then used to successfully predict (replicate) the code values of e.g. n e,Xpt for the other half.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

A systematic method for selecting molecular descriptors as features when training models for predicting physiochemical properties

Machine learning has proven to be a powerful tool for accelerating biofuel development. Although numerous models are available to predict a range of properties using chemical descriptors, there is a trade-off between interpretability and performance. Neural networks provide predictive models with high accuracy at the expense of some interpretability, while simpler models such as linear regression often lack in accuracy. In addition to model architecture, feature selection is also critical for developing interpretable and accurate predictive models. We present a method for systematically selecting molecular descriptor features and developing interpretable machine learning models without sacrificing accuracy. Our method simplifies the process of selecting features by reducing feature multicollinearity and enables discoveries of new relationships between global properties and molecular descriptors. To demonstrate our approach, we developed models for predicting melting point, boiling point, flash point, yield sooting index, and net heat of combustion with the help of the Tree-based Pipeline Optimization Tool (TPOT). For training, we used publicly available experimental data for up to 8351 molecules. Our models accurately predict various molecular properties for organic molecules (mean absolute percent error (MAPE) ranges from 3.3% to 10.5%) and provide a set of features that are well-correlated to the property. This method enables researchers to explore sets of features that significantly contribute to the prediction of the property, offering new scientific insights. To help accelerate early stage biofuel research and development, we also integrated the data and models into a open-source, interactive web tool.

09 BIOMASS FUELS↗
DOE OSTI2021

Combining artificial intelligence and physics-based modeling to directly assess atomic site stabilities: from sub-nanometer clusters to extended surfaces

The performance of functional materials is dictated by chemical and structural properties of individual atomic sites. In catalysts, for instance, the thermodynamic stability of constituting atomic sites is a key descriptor from which more complex properties, such as molecular adsorption energies and reaction rates, can be derived. In this study, we present a widely applicable machine learning (ML) approach to instantaneously compute the stability of individual atomic sites in structurally and electronically complex nano-materials. Conventionally, we determine such site stabilities using computationally intensive first-principles calculations. With our approach, we predict the stability of atomic sites in sub-nanometer metal clusters of 3–55 atoms with mean absolute errors in the range of 0.11–0.14 eV. To extract physical insights from the ML model, we introduce a genetic algorithm (GA) for feature selection. This algorithm distills the key structural and chemical properties governing the stability of atomic sites in size-selected nanoparticles, allowing for physical interpretability of the models and revealing structure–property relationships. The results of the GA are generally model and materials specific. In the limit of large nanoparticles, the GA identifies features consistent with physics-based models for metal–metal interactions. By combining the ML model with the physics-based model, we predict atomic site stabilities in real time for structures ranging from sub-nanometer metal clusters (3–55 atom) to larger nanoparticles (147 to 309 atoms) to extended surfaces using a physically interpretable framework. Finally, we present a proof of principle showcasing how our approach can determine stable and active nanocatalysts across a generic materials space of structure and composition.

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

Machine learning assisted modeling of mixing timescale for LES/PDF of high-Karlovitz turbulent premixed combustion

Accurate modeling of mixing in the transported probability density function (PDF) method remains a great challenge, especially for turbulent premixed combustion under extreme conditions such as high Karlovitz number Ka. Recently, a power-law based mixing timescale model was developed for the large-eddy simulations (LES)/PDF modeling of high-Ka number turbulent premixed flames. It is found in this work that the power-law mixing timescale model is highly sensitive to the model parameters. It is thus critically needed to develop accurate calibration of these model parameters. The empirical specification of the model parameters developed in Zhang et. al. is found to be inadequate for accurate modeling of the mixing timescale. Here, machine learning is introduced as an attractive alternative in this work for the specification of the model parameters. A high-Ka number DNS jet flame is used as the training and validation of the machine learning models. The choices of the input parameters are discussed and compared for the machine learning models. The effect of differential molecular diffusion on mixing is examined by including the effect of the Lewis number in the training of the machine learning models. The performance of different machine learning algorithms is compared for the specification of the mixing model parameters. Overall, excellent performance of the machine learning models is observed for assisting the mixing modeling. The feasibility, interpretability, applicability, generality, and portability of using machine learning are discussed in general to provide a perspective on applying data-driven machine learning for turbulent combustion modeling studies.

42 ENGINEERING↗
DOE OSTI2021

NO x adsorption with CO and C 2 H 4 on Pd/SSZ-13: Experiments and modeling

A transient monolith model containing microkinetic schemes for NO uptake and release over Pd/SSZ-13 without and with reductants CO or C2H4 is presented. The scheme involves three cationic Pd species (Z – [PdOH] + , Z – Pd 2+ Z – , Z – Pd + ) as the active sites during uptake of NO, CO and C 2 H 4 , and their desorption and conversion at higher temperature. Kinetic parameters are estimated through a combination of density functional theory (DFT) estimates and a fit of uptake, desorption and conversion data. The tuned model is validated at different uptake temperatures, ramp rates, and flowrates. A “degree of uptake control” parameter is defined that helps to identify the step(s) that are uptake controlling. The model helps to interpret the data features and is used to identify operating conditions to meet application-relevant performance metrics, including NO trapping efficiency and NO release temperature. Furthermore, the model demonstrates higher NO uptake on Pd/SSZ-13 compared to Pd/ZSM-5.

42 ENGINEERING↗
DOE OSTI2023

Search for new physics in multijet events with at least one photon and large missing transverse momentum in proton-proton collisions at 13 TeV

A search for new physics in final states consisting of at least one photon, multiple jets, and large missing transverse momentum is presented, using proton-proton collision events at a center-of-mass energy of 13 TeV. The data correspond to an integrated luminosity of 137 fb$^{−1}$, recorded by the CMS experiment at the CERN LHC from 2016 to 2018. The events are divided into mutually exclusive bins characterized by the missing transverse momentum, the number of jets, the number of b-tagged jets, and jets consistent with the presence of hadronically decaying W, Z, or Higgs bosons. The observed data are found to be consistent with the prediction from standard model processes. The results are interpreted in the context of simplified models of pair production of supersymmetric particles via strong and electroweak interactions. Depending on the details of the signal models, gluinos and squarks of masses up to 2.35 and 1.43 TeV, respectively, and electroweakinos of masses up to 1.23 TeV are excluded at 95% confidence level.[graphic not available: see fulltext]

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2022

Search for high-mass resonances decaying to a jet and a Lorentz-boosted resonance in proton-proton collisions at s = 13 TeV

A search is reported for high-mass hadronic resonances that decay to a parton and a Lorentz-boosted resonance, which in turn decays into a pair of partons. The search is based on data collected with the CMS detector at the LHC in proton-proton collisions at $\sqrt{s}$ = 13 TeV, corresponding to an integrated luminosity of 138 fb$^{-1}$. The boosted resonance is reconstructed as a single wide jet with substructure consistent with a two-body decay. The high-mass resonance is thus considered as a dijet system. The jet substructure information and the kinematic properties of cascade resonance decays are exploited to disentangle the signal from the large quantum chromodynamics multijet background. The dijet mass spectrum is analyzed for the presence of new high-mass resonances, and is found to be consistent with the standard model background predictions. Results are interpreted in a warped extra dimension model where the high-mass resonance is a Kaluza-Klein gluon, the boosted resonance is a radion, and the final state partons are all gluons. Limits on the production cross section are set as a function of the Kaluza-Klein gluon and radion masses. These limits exclude at 95% confidence level models with Kaluza-Klein gluon masses in the range from 2.0 to 4.3 TeV and radion masses in the range from 0.20 to 0.74 TeV. By exploring a novel experimental signature, the observed limits on the Kaluza-Klein gluon mass are extended by up to about 1 TeV compared to previous searches.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2020

Caltech Lab Experiments and the Insights They Provide Into Solar Corona Phenomena

A comprehensive overview of two decades of Caltech experiments relevant to solar corona physics is presented. The extent to which the experiments scale to the solar corona, the basic configurations and operation, and the importance of the magnetic force J × B common to all the experiments is discussed. Summaries are given of the various configurations used, the main observations, and interpretations of these observations, including new models developed to provide these interpretations. Topics include observations and explanations for flux rope self-collimation, axial flows along flux ropes, eruption of arched flux ropes, strapping magnetic fields that inhibit eruption, the torus instability, and effects such as X-ray emission of a kink-driven secondary Rayleigh-Taylor instability.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗