Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Ensembles”

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 739 records · Page 41

A performance analysis of ensemble averaging for high fidelity turbulence simulations at the strong scaling limit

We analyze the potential performance benefits of estimating expected quantities in large eddy simulations of turbulent flows using true ensembles rather than ergodic time averaging. Multiple realizations of the same flow are simulated in parallel, using slightly perturbed initial conditions to create unique instantaneous evolutions of the flow field. Each realization is then used to calculate statistical quantities. Provided each instance is sufficiently de-correlated, this approach potentially allows considerable reduction in the time to solution beyond the strong scaling limit for a given accuracy. This study focuses on the theory and implementation of the methodology in Nek5000, a massively parallel open-source spectral element code.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

An adaptive knowledge-based data-driven approach for turbulence modeling using ensemble learning technique under complex flow configuration: 3D PWR sub-channel with DNS data

This work describes a new approach to increase the accuracy of Reynolds-averaged Navier–Stokes (RANS) in modeling turbulence flow leveraging the machine learning technique. Traditionally, different turbulence models for Reynolds stress are developed for different flow patterns based on human knowledge. Each turbulence model has a certain application domain and prediction uncertainty. In recent years, with the rapid improvements of machine learning techniques, researchers start to develop an approach to compensate for the prediction discrepancy of traditional turbulence models with statistical models and data. However, the approach has deficiencies in several aspects. For example, the amount of human knowledge introduced to the statistical model couldn’t be controlled, which makes the statistical model learn from a very naïve stage and limits its application. In this work, a new approach is developed to address those deficiencies. Here, the new approach uses the “ensemble learning” technique to control the amount of human knowledge introduced into the statistical model. Therefore, the new approach could be adaptive to the multiple application domains. In conclusion, according to the results of case study, the new approach shows higher accuracy than both traditional turbulence models and the previous machine learning approach.

42 ENGINEERING↗

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↗

Effects of Catalytic Action and Ligand Binding on Conformational Ensembles of Adenylate Kinase

Crystal structures of adenylate kinase (AdK) from Escherichia coli capture two states: an “open” conformation (apo) obtained in the absence of ligands and a “closed” conformation in which ligands are bound. Other AdK crystal structures suggest intermediate conformations that may lie on the transition pathway between these two states. To characterize the transition from open to closed states in solution, X-ray solution scattering data were collected from AdK in the apo form and with progressively increasing concentrations of five different ligands. Scattering data from apo AdK are consistent with scattering predicted from the crystal structure of AdK in the open conformation. In contrast, data from AdK samples saturated with Ap5A do not agree with that calculated from AdK in the closed conformation. Using cluster analysis of available structures, we selected representative structures in five conformational states: open, partially open, intermediate, partially closed, and closed. We used these structures to estimate the relative abundances of these states for each experimental condition. X-ray solution scattering data obtained from AdK with AMP are dominated by scattering from AdK in the open conformation. For AdK in the presence of high concentrations of ATP and ADP, the conformational ensemble shifts to a mixture of partially open and closed states. Even when AdK is saturated with Ap5A, a significant proportion of AdK remains in a partially open conformation. Here, these results are consistent with an induced-fit model in which the transition of AdK from an open state to a closed state is initiated by ATP binding.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

How do modeling decisions affect the spread among hydrologic climate change projections? Exploring a large ensemble of simulations across a diversity of hydroclimates

Methodological choices can have strong effects on projections of climate change impacts on hydrology. In this study, we investigate the ways in which four different steps in the modeling chain influence the spread in projected changes of different aspects of hydrology. To form the basis of these analyses, we constructed an ensemble of 160 simulations from permutations of two Representative Concentration Pathways, 10 global climate models, two downscaling methods, and four hydrologic model implementations. The study is situated in the Pacific Northwest of North America, which has relevance to a diverse, multinational cast of stakeholders. We analyze the effects of each modeling decision on changes in gridded hydrologic variables of snow water equivalent and runoff, as well as streamflow at point locations. Results show that the choice of representative concentration pathway or global climate model is the driving contributor to the spread in annual streamflow volume and timing. On the other hand, hydrologic model implementation explains most of the spread in changes in low flows. Finally, by grouping the results by climate region the results have the potential to be generalized beyond the Pacific Northwest. Future hydrologic impact assessments can use these results to better tailor their modeling efforts.

58 GEOSCIENCES↗

Ensemble‐Based Spatially Distributed CLM5 Hydrological Parameter Estimation for the Continental United States

Abstract One of the major challenges in large‐domain hydrological modeling efforts lies in the estimation of spatially distributed hydrological parameters while simultaneously accounting for their associated uncertainties. Addressing this challenge is particularly difficult in ungauged locations. With growing societal demands for large‐scale streamflow projections to inform water resource management and long‐term planning, evaluating and constraining hydrological parameter uncertainty is increasingly vital. This study introduces a hybrid regionalization approach to enhance hydrological predictions of the Community Land Model version 5 (CLM5) across the Continental United States (CONUS), with a total of 50,629 1/8° grid cells. This hybrid method combines the strengths of two existing techniques: parameter regionalization and streamflow signature regionalization. It identifies ensemble behavioral parameters for each 1/8° grid cell across the CONUS domain, tailored to three distinct streamflow signatures focused on low flows, high flows, and annual water balance. Evaluating this hybrid method for 464 CAMELS (Catchment Attributes and Meteorology for Large‐sample Studies) basins demonstrates a significant improvement in CLM5 hydrological predictions, even in challenging arid regions. In CONUS applications, the derived spatially distributed parameter sets capture both spatial continuity and variation of parameters, highlighting their heterogeneous nature within specific regions. Overall, this hybrid regionalization approach offers a promising solution to the complex task of improving hydrological modeling over large domains for important hydrological applications.

CLM5↗

A multi model ensemble reveals net climate benefits from regenerative practices in US Midwest croplands

Process-based cropping systems models (CSMs) are key components of measurement, monitoring, reporting, and verification frameworks of carbon markets, but model-specific differences limit their applicability across diverse pedo-climatic conditions and agronomic practices. Multi-model ensemble (MME) provides an opportunity to better estimate changes in soil organic carbon (SOC) and nitrous oxide (N 2 O) emissions from agronomic practices at scale. We used an MME across 46 million hectares of US Midwest cropland at a resolution of 4-km 2 to assess the aggregate ability of different regenerative practices to sequester SOC and N 2 O emissions compared to their counterfactual dynamic baselines. MME was validated against long-term trials and compared to its constituent CSMs, showing greater accuracy and lower uncertainty. The results show that adopting no-till combined with cover crops increased SOC stocks by 0.36 ± 0.12 Mg ha -1 yr -1 , corresponding to a net regional SOC gain of 16.4 Tg C yr -1 compared to business-as-usual baselines. These benefits are halved when each management is practiced individually, and the SOC gains are only fully realized with low initial carbon stock. By including N 2 O emissions, we can assess the overall climate mitigation potential, specifically, the extent to which carbon sequestration can offset direct N 2 O emissions. The magnitude of this potential varies depending on management practices and geographic location with net climate benefits on average ranging from 0 to 3 Mg CO 2 -eq ha -1 yr -1 . High-resolution MME results allow for robust estimates of climate mitigation, reducing barriers to carbon market participation and supporting regenerative agriculture initiatives at scale.

carbon credits↗

Identifying Optimal Measurement Subspace for the Ensemble Kalman Filter

To reduce the computational load of the ensemble Kalman filter while maintaining its efficacy, an optimization algorithm based on the generalized eigenvalue decomposition method is proposed for identifying the most informative measurement subspace. When the number of measurements is large, the proposed algorithm can be used to make an effective tradeoff between computational complexity and estimation accuracy. This algorithm also can be extended to other Kalman filters for measurement subspace selection.

99 GENERAL AND MISCELLANEOUS↗

Significant impact of forcing uncertainty in a large ensemble of climate model simulations

Forcing due to solar and volcanic variability, on the natural side, and greenhouse gas and aerosol emissions, on the anthropogenic side, are the main inputs to climate models. Reliable climate model simulations of past and future climate change depend crucially upon them. Here we analyze large ensembles of simulations using a comprehensive Earth System Model to quantify uncertainties in global climate change attributable to differences in prescribed forcings. The different forcings considered here are those used in the two most recent phases of the Coupled Model Intercomparison Project (CMIP), namely CMIP5 and CMIP6. We show significant differences in simulated global surface air temperature due to volcanic aerosol forcing in the second half of the 19th century and in the early 21st century. The latter arise from small-to-moderate eruptions incorporated in CMIP6 simulations but not in CMIP5 simulations. We also find significant differences in global surface air temperature and Arctic sea ice area due to anthropogenic aerosol forcing in the second half of the 20th century and early 21st century. These differences are as large as those obtained in different versions of an Earth System Model employing identical forcings. In simulations from 2015 to 2100, we find significant differences in the rates of projected global warming arising from CMIP5 and CMIP6 concentration pathways that differ slightly but are equivalent in terms of their nominal radiative forcing levels in 2100. Our results highlight the influence of assumptions about natural and anthropogenic aerosol loadings on carbon budgets, the likelihood of meeting Paris targets, and the equivalence of future forcing scenarios.

54 ENVIRONMENTAL SCIENCES↗

Kinetic description of site ensembles on catalytic surfaces

Significance The kinetic assessment of catalytic reactions is essential to the design and application of thermochemical processes useful in petrochemical synthesis, energy conversion, and environmental remediation. We demonstrate that the ubiquitously used Langmuir–Hinshelwood formalism is incomplete and fails to correctly describe even simple catalytic reactions (e.g., A + A → A 2 ). An accurate kinetic description requires explicit rate equations for multisite configurations (e.g., A*–A* pairs) to account for the nonrandom clustering/isolation of surface species. We rigorously describe clustering/isolation phenomena by derivation of coverage-dependent stoichiometric coefficients, which capture the distinct influence each elementary step exerts on each catalytic site ensemble—thereby overcoming limitations of the mean-field (i.e., Hinshelwood) assumption in describing reactions of mobile Langmuirian adsorbates on catalytic surfaces.

36 MATERIALS SCIENCE↗

Reduced-order autodifferentiable ensemble Kalman filters

This paper introduces a computational framework to reconstruct and forecast a partially observed state that evolves according to an unknown or expensive-to-simulate dynamical system. Our reduced-order autodifferentiable ensemble Kalman filters (ROAD-EnKFs) learn a latent low-dimensional surrogate model for the dynamics and a decoder that maps from the latent space to the state space. The learned dynamics and decoder are then used within an EnKF to reconstruct and forecast the state. Numerical experiments show that if the state dynamics exhibit a hidden low-dimensional structure, ROAD-EnKFs achieve higher accuracy at lower computational cost compared to existing methods. If such structure is not expressed in the latent state dynamics, ROAD-EnKFs achieve similar accuracy at lower cost, making them a promising approach for surrogate state reconstruction and forecasting.

Mathematics↗

Using ensembles and distillation to optimize the deployment of deep learning models for the classification of electronic cancer pathology reports

One of the goals of the Surveillance, Epidemiology, and End Results (SEER) program is to estimate incidence, prevalence, and mortality of all cancers. To that end, cancer registries across the country maintain a massive database of cancer pathology reports which contain rich information to understand cancer trends. However, these reports are stored in the form of unstructured text, and human annotators are required to read and extract relevant information. In this article, we show that existing deep learning models for automating information extraction from cancer pathology reports can be significantly improved by using ensemble model distillation. We found that by training multiple predictive models and transferring their knowledge to a single, low-resource model, we can reduce the number of highly confident wrong predictions. Our results show that our implemented methods could save 1000s of manual annotation hours.

60 APPLIED LIFE SCIENCES↗

From deep TLS validation to ensembles of atomic models built from elemental motions. Addenda and corrigendum

Researcher feedback has indicated that in Urzhumtsev et al. [(2015) Acta Cryst. D 71 , 1668–1683] clarification of key parts of the algorithm for interpretation of TLS matrices in terms of elemental atomic motions and corresponding ensembles of atomic models is required. Also, it has been brought to the attention of the authors that the incorrect PDB code was reported for one of test models. Lastly, these issues are addressed in this article.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Application of Ensemble Kalman Filter in Power System State Tracking and Sensitivity

Ensemble Kalman Filter (EnKF) is proposed to track dynamic states of generators. The algorithm of EnKF and its application to generator state tracking are presented in detail. The accuracy and sensitivity of the method are analyzed with respect to initial state errors, measurement noise, unknown fault locations, time steps and parameter errors. It is demonstrated through simulation studies that even with some errors in the parameters, the developed EnKF can effectively track generator dynamic states using disturbance data.

dynamic state tracking↗

libEnsemble: A Library to Coordinate the Concurrent Evaluation of Dynamic Ensembles of Calculations

Almost all applications stop scaling at some point; those that don't are seldom performant when considering time to solution on anything but aspirational/unicorn resources. Recognizing these tradeoffs as well as greater user functionality in a near-term exascale computing era, we present libEnsemble, a library aimed at particular scalability- and capability-stretching uses. libEnsemble enables running concurrent instances of an application in dynamically allocated ensembles through an extensible Python library. Here, we highlight the structure, execution, and capabilities of the library on leading pre-exascale environments as well as advanced capabilities for exascale environments and beyond.

97 MATHEMATICS AND COMPUTING↗

A Novel Active Optimization Approach for Rapid and Efficient Design Space Exploration Using Ensemble Machine Learning

In this work, a novel design optimization technique based on active learning, which involves dynamic exploration and exploitation of the design space of interest using an ensemble of machine learning algorithms, is presented. In this approach, a hybrid methodology incorporating an explorative weak learner (regularized basis function model) that fits high-level information about the response surface and an exploitative strong learner (based on committee machine) that fits finer details around promising regions identified by the weak learner is employed. For each design iteration, an aristocratic approach is used to select a set of nominees, where points that meet a threshold merit value as predicted by the weak learner are selected for evaluation. In addition to these points, the global optimum as predicted by the strong learner is also evaluated to enable rapid convergence to the actual global optimum once the most promising region has been identified by the optimizer. Additionally, this methodology is first tested by applying it to the optimization of a two-dimensional multi-modal surface and, subsequently, to a complex internal combustion (IC) engine combustion optimization case with nine control parameters related to fuel injection, initial thermodynamic conditions, and in-cylinder flow. It is found that the new approach significantly lowers the number of function evaluations that are needed to reach the optimum design configuration (by up to 80%) when compared to conventional optimization techniques, such as particle swarm and genetic algorithm-based optimization techniques.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Multilevel ensemble Kalman filtering

This study embeds a multilevel Monte Carlo sampling strategy into the Monte Carlo step of the ensemble Kalman filter (EnKF) in the setting of finite dimensional signal evolution and noisy discrete-time observations. The signal dynamics is assumed to be governed by a stochastic differential equation (SDE), and a hierarchy of time grids is introduced for multilevel numerical integration of that SDE. Finally, the resulting multilevel EnKF is proved to asymptotically outperform EnKF in terms of computational cost versus approximation accuracy. The theoretical results are illustrated numerically.

ltering↗

Addressing Interdependency in a Multimodel Ensemble by Interpolation of Model Properties

The diverse set of Earth system models used to conduct the CMIP5 ensemble can partly sample the uncertainties in future climate projections. However, combining those projections is complicated by the fact that models developed by different groups share ideas and code and therefore biases. The authors propose a method for combining model results into single or multivariate distributions that are more robust to the inclusion of models with a large degree of interdependency. This study uses a multivariate metric of present-day climatology to assess both model performance and similarity in two recent model intercomparisons, CMIP3 and CMIP5. Model characteristics can be interpolated and then resampled in a space defined by independent climate properties. A form of weighting can be applied by sampling more densely in the region of the space close to the projected observations, thus taking into account both model performance and interdependence. Furthermore, the choice of the sampling distribution’s parameters is a subjective decision that should reflect the researcher’s prior assumptions as to the acceptability of different model errors.

58 GEOSCIENCES↗