Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Ensemble models”

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 973 records · Page 54
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Open-closed 3d gravity as a random ensemble

We investigate an ensemble of boundary CFTs within the framework of a tensor model recently constructed to model 3d quantum gravity. The incorporation of CFT borders introduces new elements to the gravity theory. In particular, it leads to an open-closed extension of Virasoro TQFT, which in the classical limit gives rise to 3d gravity with tensionful end-of-the-world branes. It also provides predictions for off-shell manifolds with bordered asymptotic boundaries, such as the annulus wormhole. As an application, we construct a purely open variant of the tensor model to study a purely open bootstrap problem in the context of CFT triangulation. We also briefly discuss the extension to non-orientable CFTs.

AdS-CFT correspondence↗
DOE OSTI2024

Evolution of highly multimodal Rayleigh–Taylor instabilities

Rayleigh–Taylor (RT) instabilities are important fluid instabilities that arise in inertial confinement fusion (ICF) capsule implosions, and many other contexts. Multi-mode coupling is observed in experiments and plays a substantial role in material mix from RT instabilities. In this work, we study the evolution of highly multimodal perturbations (power law distribution) that approximate those found at manufactured material interfaces. We use simulations of over 2000 different perturbations in the LANL code xRAGE to identify distinct phases in the processes of bubble growth and bubble merger which can be visualized in a 2D phase portrait with clear regimes of mode growth and decay. Our results show that the dynamic evolution of the instability strongly depends on the mode of the perturbations and mode interactions. The merger process accelerates bubble growth. A non-Markovian region and a transition of the instability from: (1) initial exponential growth to (2) linear growth and to (3) quadratic growth and asymptotic behavior, are clearly captured in the phase space. We have developed a quantitative model of bubble growth that reproduces the dynamic behavior of ensembles of perturbations. Implications for ICF capsules designed for robustness against instabilities are discussed.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2026

Active operator learning with predictive uncertainty quantification for partial differential equations

With the increased prevalence of neural operators being used to provide rapid solutions to partial differential equations (PDEs), understanding the accuracy of model predictions and the associated error levels is necessary for deploying reliable surrogate models in scientific applications. Existing uncertainty quantification (UQ) frameworks employ ensembles or Bayesian methods, which can incur substantial computational costs during both training and inference. Here, we propose a lightweight predictive UQ method tailored for Deep operator networks (DeepONets) that also generalizes to other operator networks. Numerical experiments on linear and nonlinear PDEs demonstrate that the framework’s uncertainty estimates are unbiased and provide accurate out-of-distribution uncertainty predictions with a sufficiently large training dataset. Our framework provides fast inference and uncertainty estimates that can efficiently drive outer-loop analyses that would be prohibitively expensive with conventional solvers. We demonstrate how predictive uncertainties can be used in the context of Bayesian optimization and active learning problems to yield improvements in accuracy and data-efficiency for outer-loop optimization procedures. In the active learning setup, we extend the framework to Fourier Neural Operators (FNO) and describe a generalized method for other operator networks. To enable real-time deployment, we introduce an inference strategy based on precomputed trunk outputs and a sparse placement matrix, reducing evaluation time by more than a factor of five. Our method provides a practical route to uncertainty-aware operator learning in time-sensitive settings.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Unusual high-field metal in a Kondo insulator

Strong electronic interactions in condensed-matter systems often lead to unusual quantum phases. One such phase occurs in the Kondo insulator YbB12, the insulating state of which exhibits phenomena that are characteristic of metals, such as magnetic quantum oscillations, a gapless fermionic contribution to heat capacity and itinerant-fermion thermal transport. To understand these phenomena, it is informative to study their evolution as the energy gap of the Kondo insulator state is closed by a large magnetic field. Here we show that clear quantum oscillations are observed in the resulting high-field metallic state in YbB12; this is despite it possessing relatively high resistivity, large effective masses and huge Kadowaki–Woods ratio, a combination that normally precludes quantum oscillations. Both quantum oscillation frequency and cyclotron mass display a strong field dependence. By tracking the Fermi surface area, we conclude that the same quasiparticle band gives rise to quantum oscillations in both insulating and metallic states. These data are understood most simply by using a two-fluid picture in which neutral quasiparticles—contributing little or nothing to charge transport—coexist with charged fermions. Overall, our observations of the complex field-dependent behaviour of the fermion ensemble inhabiting YbB 12 provide strong constraints for existing theoretical models.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Unpolarized transverse momentum dependent parton distributions of the nucleon from lattice QCD

We present a first lattice QCD calculation of the unpolarized nucleon’s isovector transverse-momentum-dependent parton distribution functions (TMDPDFs), which are essential to predict observables of multiscale, semi-inclusive processes in the standard model. We use a N f = 2 + 1 + 1 MILC ensemble with valence clover fermions on a highly improved staggered quark (HISQ) sea to compute the quark momentum distributions in a large-momentum nucleon on the lattice. The state-of-the-art techniques in renormalization and extrapolation in the correlation distance on the lattice are adopted. The perturbative kernel up to next-to-next-to-leading order is taken into account, and the dependence on the pion mass and the hadron momentum is explored. Our results are qualitatively comparable with phenomenological TMDPDFs, which provide an opportunity to predict high energy scatterings from first principles. Published by the American Physical Society 2024

Astronomy & Astrophysics↗
DOE OSTI2025

Conformational ensembles reveal the origins of serine protease catalysis

Enzymes exist in ensembles of states that encode the energetics underlying their catalysis. Conformational ensembles built from 1231 structures of 17 serine proteases revealed atomic-level changes across their reaction states. By comparing the enzymatic and solution reaction, we identified molecular features that provide catalysis and quantified their energetic contributions to catalysis. Serine proteases precisely position their reactants in destabilized conformers, creating a downhill energetic gradient that selectively favors the motions required for reaction while limiting off-pathway conformational states. The same catalytic features have repeatedly evolved in proteases and additional enzymes across multiple distinct structural folds. In conclusion, our ensemble-function analyses revealed previously unknown catalytic features, provided quantitative models based on simple physical and chemical principles, and identified motifs recurrent in nature that may inspire enzyme design.

Du, Siyuan [Stanford Univ., CA (United States)] (O↗
DOE OSTI2021

Influence of the Indian Ocean Dipole on the large-scale circulation in South America

The influence of each phase of the Indian Ocean Dipole (IOD) on the largescale circulation in South America is investigated using rainfall observations, fully-coupled, large-ensemble, historical simulations (LENS), and forced experiments using the coupled model’s atmospheric component. IOD events often occur when El Niño Southern Oscillation (ENSO), the largest source of interannual variability of precipitation in South America, is active. To distinguish from effects of ENSO, only cases during neutral ENSO conditions are analyzed in LENS and observations. During the positive IOD polarity, a perturbation in the localWalker circulation leads to increased convection over equatorial South America, resulting in wet anomalies in the Amazon basin. This signal is the opposite of what is expected during El Niño events. Tropical convection anomalies in the Indian Ocean also force an extratropical Rossby wave train that reaches subtropical South America. During positive IOD, the moisture flux from the Amazon to central and southeastern Brazil weakens, resulting in a drying of the area associated with the South Atlantic Convergence Zone. Meanwhile, the South Atlantic Subtropical High strengthens, contributing to a drying in southeastern Brazil. During negative IOD, the induced wave train from the Indian Ocean leads to increased moisture transport to the La Plata basin, leading to wet anomalies in the region.

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

A protocol for model intercomparison of impacts of marine cloud brightening climate intervention

A modeling protocol (defined by a series of climate model simulations with specified model output) is introduced. Studies using these simulations are designed to improve the understanding of climate impacts using a strategy for climate intervention (CI) known as marine cloud brightening (MCB) in specific regions; therefore, the protocol is called MCB-REG (where REG stands for region). The model simulations are not intended to assess consequences of a realistic MCB deployment intended to achieve specific climate targets but instead to expose responses to interventions in six regions with pervasive cloud systems that are often considered candidates for such a deployment. A calibration step involving simulations with fixed sea surface temperatures (SSTs) is first used to identify a common forcing, and then coupled simulations with forcing in individual regions and combinations of regions are used to examine climate impacts. Synthetic estimates constructed by superposing responses from simulations with forcing in individual regions are considered a means of approximating the climate impacts produced when MCB interventions are introduced in multiple regions. A few results comparing simulations from three modern climate models (CESM2, E3SMv2, and UKESM1) are used to illustrate the similarities and differences between model behavior and the utility of estimates of MCB climate responses that were synthesized by summing responses introduced in individual regions. Cloud responses to aerosol injections differ substantially between models (CESM2 clouds appear much more susceptible to aerosol emissions than the other models), but patterns in precipitation and surface temperature responses were similar when forcing is imposed with similar amplitudes in the same regions. A previously identified La Niña-like response to forcing introduced in the Southeast Pacific is evident in this study, but the amplitude of the response was shown to markedly differ across the three models. Other common response patterns were also found and are discussed. Forcing in the Southeast Atlantic consistently (across all three models) produces weaker global cooling than that in other regions, and the Southeast Pacific and South Pacific show the strongest cooling. This indicates that the efficiency of a given intervention depends on not only the susceptibility of the clouds to aerosol perturbations, but also the strength of the underlying radiative feedbacks and ocean responses operating within each region. These responses were generally robust across models, but more studies and an examination of responses with ensembles would be beneficial.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Reweighting Monte Carlo predictions and automated fragmentation variations in Pythia 8

This work reports on a method for uncertainty estimation in simulated collider-event predictions. The method is based on a Monte Carlo-veto algorithm, and extends previous work on uncertainty estimates in parton showers by including uncertainty estimates for the Lund string-fragmentation model. This method is advantageous from the perspective of simulation costs: a single ensemble of generated events can be reinterpreted as though it was obtained using a different set of input parameters, where each event now is accompanied with a corresponding weight. This allows for a robust exploration of the uncertainties arising from the choice of input model parameters, without the need to rerun full simulation pipelines for each input parameter choice. Such explorations are important when determining the sensitivities of precision physics measurements. Accompanying code is available at https://gitlab.com/uchep/mlhad-weights-validation.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

EFIT-Prime: Probabilistic and physics-constrained reduced-order neural network model for equilibrium reconstruction in DIII-D

We introduce EFIT-Prime, a novel machine learning surrogate model for EFIT (Equilibrium FIT) that integrates probabilistic and physics-informed methodologies to overcome typical limitations associated with deterministic and ad hoc neural network architectures. EFIT-Prime utilizes a neural architecture search-based deep ensemble for robust uncertainty quantification, providing scalable and efficient neural architectures that comprehensively quantify both data and model uncertainties. Physically informed by the Grad–Shafranov equation, EFIT-Prime applies a constraint on the current density J tor and a smoothness constraint on the first derivative of the poloidal flux, ensuring physically plausible solutions. Furthermore, the spatial location of the diagnostics is explicitly incorporated in the inputs to account for their spatial correlation. Extensive evaluations demonstrate EFIT-Prime's accuracy and robustness across diverse scenarios, most notably showing good generalization on negative-triangularity discharges that were excluded from training. Timing studies indicate an ensemble inference time of 15 ms for predicting a new equilibrium, offering the possibility of plasma control in real-time, if the model is optimized for speed.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Quantitative evaluations of subtropical westerly jet simulations over East Asia based on multiple CMIP5 and CMIP6 GCMs

As a salient feature of the Asian monsoon system, the East Asian subtropical westerly jet (EASWJ) exerts significant impacts on weather and climate changes in China and even throughout East Asia. In this paper, we applied a new self-adaptive algorithm to detect the EASWJ, identify its boundaries, and then represent its characteristics by defining three indices: the intensity index, meridional displacement index, and width index. Compared to the reanalysis data, we carried out a comprehensive, objective, and quantitative EASWJ evaluation using historical experiments from multiple global climate models (GCMs) in the Coupled Model Intercomparison Project Phases 5 and 6 (CMIP5 and CMIP6). The results show that the multimodel ensemble mean (MME) of both CMIP5 and CMIP6 can simulate the characteristics of winter EASWJ well. While for the other three seasons, the MME of both phase models underestimate the 200-hPa zonal wind (U200) strength in the jet coverage area and overestimate the U200 outside the jet area, such simulation results weaken the meridional shear of the wind field. The EASWJ simulations from the CMIP5 GCMs had no consistent intensity or location characteristic tendencies, and most CMIP5 GCMs tended to simulate relatively wide-coverage jets. In contrast, most CMIP6 GCMs are inclined to simulate significantly weaker, wider, and more-northward jets. Compared to the predecessors in CMIP5, about half of CMIP6 GCMs significantly minimized the jet intensity bias, but remarkable errors were still observed in their jet location and coverage representations. Furthermore, the comparative analysis performed by classifying models based on their evaluated simulation results suggested that the simulated performance of the meridional temperature gradient was important for capturing the EASWJ characteristics. Further in-depth study of the causes of model differences is warranted to improve simulation results.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Global phase diagram of D-wave superconductivity in the square-lattice t-J model

The Hubbard and closely related t-J models are exciting platforms for unconventional superconductivity (SC). Through state-of-the-art density matrix renormalization group calculations using the grand canonical ensemble, we address open issues regarding the ground-state phase diagram of the extended t-J model on a square lattice in the parameter regime relevant to cuprate superconductors. On large 8-leg cylinders, we demonstrate that the pure t-J model with only nearest-neighbor hoppings and superexchange interactions, for a wide range of doping ( δ = 0.1 to 0.2 ), hosts robust d-wave superconductivity possibly coexisting with weak unidirectional pair density wave. Furthermore, a small next nearest neighbor hopping t 2 suppresses pair and charge density waves, resulting in a uniform d-wave SC phase in both electron- and hole-doped cuprate model systems. Our work validates the t-J model as a proper minimum model for the emergence of superconductivity in cuprate superconductors.

Chen, Feng (ORCID:000000018196030X)↗
DOE OSTI2023

Neural network ensembles and uncertainty estimation for predictions of inelastic mechanical deformation using a finite element method-neural network approach

The finite element method (FEM) is widely used to simulate a variety of physics phenomena. Approaches that integrate FEM with neural networks (NNs) are typically leveraged as an alternative to conducting expensive FEM simulations in order to reduce the computational cost without significantly sacrificing accuracy. However, these methods can produce biased predictions that deviate from those obtained with FEM, since these hybrid FEM-NN approaches rely on approximations trained using physically relevant quantities. In this work, an uncertainty estimation framework is introduced that leverages ensembles of Bayesian neural networks to produce diverse sets of predictions using a hybrid FEM-NN approach that approximates internal forces on a deforming solid body. The uncertainty estimator developed herein reliably infers upper bounds of bias/variance in the predictions for a wide range of interpolation and extrapolation cases using a three-element FEM-NN model of a bar undergoing plastic deformation. This proposed framework offers a powerful tool for assessing the reliability of physics-based surrogate models by establishing uncertainty estimates for predictions spanning a wide range of possible load cases.

42 ENGINEERING↗
DOE OSTI2023

An Ensemble Score Filter for Tracking High-Dimensional Nonlinear Dynamical Systems

We propose an ensemble score filter (EnSF) for solving high-dimensional nonlinear filtering problems with superior accuracy. A major drawback of existing filtering methods, e.g., particle filters or ensemble Kalman filters, is the low accuracy in handling high-dimensional and highly nonlinear problems. EnSF attacks this challenge by exploiting the score-based diffusion model, defined in a pseudo-temporal domain, to characterizing the evolution of the filtering density. EnSF stores the information of the recursively updated filtering density function in the score function, in stead of storing the information in a set of finite Monte Carlo samples (used in particle filters and ensemble Kalman filters). Unlike existing diffusion models that train neural networks to approximate the score function, we develop a training-free score estimation that uses mini-batch-based Monte Carlo estimator to directly approximate the score function at any pseudo-spatial-temporal location, which provides sufficient accuracy in solving high-dimensional nonlinear problems as well as saves tremendous amount of time spent on training neural networks. Another essential aspect of EnSF is its analytical update step, gradually incorporating data information into the score function, which is crucial in mitigating the degeneracy issue faced when dealing with very high-dimensional nonlinear filtering problems. High-dimensional Lorenz systems are used to demonstrate the performance of our method. EnSF provides surprisingly impressive performance in reliably tracking extremely high-dimensional Lorenz systems (up to 1,000,000 dimension) with highly nonlinear observation processes, which is a well-known challenging problem for existing filtering methods.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Robust and scalable uncertainty estimation with conformal prediction for machine-learned interatomic potentials

Abstract Uncertainty quantification (UQ) is important to machine learning (ML) force fields to assess the level of confidence during prediction, as ML models are not inherently physical and can therefore yield catastrophically incorrect predictions. Established a-posteriori UQ methods, including ensemble methods, the dropout method, the delta method, and various heuristic distance metrics, have limitations such as being computationally challenging for large models due to model re-training. In addition, the uncertainty estimates are often not rigorously calibrated. In this work, we propose combining the distribution-free UQ method, known as conformal prediction (CP), with the distances in the neural network’s latent space to estimate the uncertainty of energies predicted by neural network force fields. We evaluate this method (CP+latent) along with other UQ methods on two essential aspects, calibration, and sharpness, and find this method to be both calibrated and sharp under the assumption of independent and identically-distributed (i.i.d.) data. We show that the method is relatively insensitive to hyperparameters selected, and test the limitations of the method when the i.i.d. assumption is violated. Finally, we demonstrate that this method can be readily applied to trained neural network force fields with traditional and graph neural network architectures to obtain estimates of uncertainty with low computational costs on a training dataset of 1 million images to showcase its scalability and portability. Incorporating the CP method with latent distances offers a calibrated, sharp and efficient strategy to estimate the uncertainty of neural network force fields. In addition, the CP approach can also function as a promising strategy for calibrating uncertainty estimated by other approaches.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2023

An Ensemble of Neural Networks for Moist Physics Processes, Its Generalizability and Stable Integration

Abstract With the recent advances in data science, machine learning has been increasingly applied to convection and cloud parameterizations in global climate models (GCMs). This study extends the work of Han et al. (2020, https://doi.org/10.1029/2020MS002076 ) and uses an ensemble of 32‐layer deep convolutional residual neural networks, referred to as ResCu‐en, to emulate convection and cloud processes simulated by a superparameterized GCM, SPCAM. ResCu‐en predicts GCM grid‐scale temperature and moisture tendencies, and cloud liquid and ice water contents from moist physics processes. The surface rainfall is derived from the column‐integrated moisture tendency. The prediction uncertainty inherent in deep learning algorithms in emulating the moist physics is reduced by ensemble averaging. Results in 1‐year independent offline validation show that ResCu‐en has high prediction accuracy for all output variables, both in the current climate and in a warmer climate with +4K sea surface temperature. The analysis of different neural net configurations shows that the success to generalize in a warmer climate is attributed to convective memory and the 1‐dimensional convolution layers incorporated into ResCu‐en. We further implement a member of ResCu‐en into CAM5 with real world geography and run the neural‐network‐enabled CAM5 (NCAM) for 5 years without encountering any numerical integration instability. The simulation generally captures the global distribution of the mean precipitation, with a better simulation of precipitation intensity and diurnal cycle. However, there are large biases in temperature and moisture in high latitudes. These results highlight the importance of convective memory and demonstrate the potential for machine learning to enhance climate modeling.

Meteorology & Atmospheric Sciences↗
DOE OSTI2022

Groundwater Flow and Contaminant Transport Model of Corrective Action Units 101 and 102: Central and Western Pahute Mesa, Nye County, Nevada (Rev. 1)

This document presents the groundwater flow and transport models along with simulation forecasts needed for the purpose of developing the ensemble of contaminant boundaries (CBs) for the Central and Western Pahute Mesa (PM) corrective action units (CAUs) 101 and 102. The PM CAUs are areas of the Nevada National Security Site (NNSS) that were historically used for underground nuclear testing.

54 ENVIRONMENTAL SCIENCES↗