Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Dynamic 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 271 records · Page 15

Quantifying uncertainty for deep learning based forecasting and flow-reconstruction using neural architecture search ensembles

Classical problems in computational physics such as data-driven forecasting and signal reconstruction from sparse sensors have recently seen an explosion in deep neural network (DNN) based algorithmic approaches. However, most DNN models do not provide uncertainty estimates, which are crucial for establishing the trustworthiness of these techniques in downstream decision making tasks and scenarios. In recent years, ensemble-based methods have achieved significant success for the uncertainty quantification in DNNs on a number of benchmark problems. However, their performance on real-world applications remains under-explored. In this work, we present an automated approach to DNN discovery and demonstrate how this may also be utilized for ensemble-based uncertainty quantification. Specifically, we propose the use of a scalable neural and hyperparameter architecture search for discovering an ensemble of DNN models for complex dynamical systems. We highlight how the proposed method not only discovers high-performing neural network ensembles for our tasks, but also quantifies uncertainty seamlessly. This is achieved by using genetic algorithms and Bayesian optimization for sampling the search space of neural network architectures and hyperparameters. Subsequently, a model selection approach is used to identify candidate models for an ensemble set construction. Afterwards, a variance decomposition approach is used to estimate the uncertainty of the predictions from the ensemble. We demonstrate the feasibility of this framework for two tasks — forecasting from historical data and flow reconstruction from sparse sensors for the sea-surface temperature. In conclusion, we demonstrate superior performance from the ensemble in contrast with individual high-performing models and other benchmarks.

Deep ensembles↗

Using Ensemble Data Assimilation to Estimate Transient Hydrologic Exchange Flow Under Highly Dynamic Flow Conditions

Abstract Quantifying dynamic hydrologic exchange flows (HEFs) within river corridors that experience high‐frequency flow variations caused by dam regulations is important for understanding the biogeochemical processes at the river water and groundwater interfaces. Heat has been widely used as a tracer to infer steady‐state flow velocities through analytical solutions of heat transport defined by the diurnal temperature signals. Under sub‐daily dynamic flow conditions, however, such analytical solutions are not applicable due to the violation of their fundamental assumptions. In this study, we developed a data assimilation‐based approach to estimate the sub‐daily flux under highly dynamic flow conditions using multi‐depth temperature observations at a 5‐min resolution. If the hydraulic gradient is measured, Darcy's law was used to calculate the flux with permeability estimated from temperature responses below the riverbed. Otherwise, flux was estimated directly by assimilating multi‐depth temperature data at 1‐ or 2‐hr time intervals assuming one‐dimensional flow and heat transport governing equation. By comparing estimated fluxes with model‐generated synthetic truth, we demonstrated that both schemes have robust performance in estimating fluxes under highly dynamic flow conditions. This data assimilation‐based flux estimation method was able to capture the vertical sub‐daily fluxes using multi‐depth high‐resolution temperature data alone, even in the presence of multi‐dimensional flow. This approach has been successfully applied to real field temperature data collected at the Hanford site, which experiences highly dynamic HEFs. Our study shows the promise of adopting distributed 1‐D temperature monitoring to capture spatial and temporal exchange dynamics in river corridors at a watershed scale or beyond.

54 ENVIRONMENTAL SCIENCES↗

ENSEMBLE KALMAN FILTERS FOR RELIABILITY ESTIMATION IN PERFUSION INFERENCE

We consider the solution of inverse problems in dynamic contrast–enhanced imaging by means of ensemble Kalman filters. Our quantity of interest is blood perfusion, i.e., blood flow rates in tissue. While existing approaches to compute blood perfusion parameters for given time series of radiological measurements mainly rely on deterministic, deconvolution–based methods, we aim at recovering probabilistic solution information for given noisy measurements. To this end, we model radiological image capturing as a sequential data assimilation process and solve it by an ensemble Kalman filter. Thereby, we recover deterministic results as an ensemble–based mean and are able to compute reliability information such as probabilities for the perfusion to be in a given range. Our target application is the inference of blood perfusion parameters in the human brain. A numerical study shows promising results for artificial measurements generated by a digital perfusion phantom.

Engineering↗

Extracting Electric Polarizabilities from Lattice QCD

Charged and neutral, pion and kaon electric polarizabilities are extracted from lattice QCD using an ensemble of anisotropic gauge configurations with dynamical clover fermions. We utilize classical background fields to access the polarizabilities from two-point correlation functions. Uniform background fields are achieved by quantizing the electric field strength with the proper treatment of boundary flux. These external fields, however, are implemented only in the valence quark sector. A novel method to extract charge particle polarizabilities is successfully demonstrated for the first time.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

N-mode coherence in collective neutrino oscillations

We study two-flavor neutrino oscillations in a homogeneous and isotropic ensemble under the influence of neutrino-neutrino interactions. For any density there exist forms of collective oscillations that show self-maintained coherence. They can be classified by a number N of linearly independent functions that describe all neutrino modes as linear superpositions. What is more, the dynamics is equivalent to another ensemble with the same effective density, consisting of N modes with discrete energies E{sub i} with i=1,...,N. We use this equivalence to derive the analytic solution for two-mode (bimodal) coherence, relevant for spectral-split formation in supernova neutrinos.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Spectroscopic properties of inhomogeneously broadened spin ensembles in a cavity

The enhanced collective coupling to weak quantum fields may turn atomic or spin ensembles into an important component in quantum information processing architectures. Inhomogeneous broadening can, however, significantly reduce the coupling and the lifetime of the collective excitation that represent the quantum information. In this paper we show that the width and shape of the inhomogeneous broadening have a striking influence on the dynamics of the cavity-ensemble system and may lead to narrowing of the linewidth of the collective states. We underpin our findings with the examples of a Gaussian and a Lorentzian profile of the inhomogeneity.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A Finite-Time Ensemble Method for Mixed Layer Model Comparison

Here, this work evaluates the fidelity of various upper-ocean turbulence parameterizations subject to realistic monsoon forcing and presents a finite-time ensemble vector (EV) method to better manage the design and numerical principles of these parameterizations. The EV method emphasizes the dynamics of a turbulence closure multimodel ensemble and is applied to evaluate 10 different ocean surface boundary layer (OSBL) parameterizations within a single-column (SC) model against two boundary layer large-eddy simulations (LES). Both LES include realistic surface forcing, but one includes wind-driven shear turbulence only, while the other includes additional Stokes forcing through the wave-average equations that generate Langmuir turbulence. The finite-time EV framework focuses on what constitutes the local behavior of the mixed layer dynamical system and isolates the forcing and ocean state conditions where turbulence parameterizations most disagree. Identifying disagreement provides the potential to evaluate SC models comparatively against the LES. Observations collected during the 2018 monsoon onset in the Bay of Bengal provide a case study to evaluate models under realistic and variable forcing conditions. The case study results highlight two regimes where models disagree 1) during wind-driven deepening of the mixed layer and 2) under strong diurnal forcing.

54 ENVIRONMENTAL SCIENCES↗

Ligand efficacy shifts a nuclear receptor conformational ensemble between transcriptionally active and repressive states

Abstract Nuclear receptors (NRs) are thought to dynamically alternate between transcriptionally active and repressive conformations, which are stabilized upon ligand binding. Most NR ligand series exhibit limited bias, primarily consisting of transcriptionally active agonists or neutral antagonists, but not repressive inverse agonists—a limitation that restricts understanding of the functional NR conformational ensemble. Here, we report a NR ligand series for peroxisome proliferator-activated receptor gamma (PPARγ) that spans a pharmacological spectrum from repression (inverse agonism) to activation (agonism) where subtle structural modifications switch compound activity. While crystal structures provide snapshots of the fully repressive state, NMR spectroscopy and conformation-activity relationship analysis reveals that compounds within the series shift the PPARγ conformational ensemble between transcriptionally active and repressive conformations that are natively populated in the apo/ligand-free ensemble. Our findings reveal a molecular framework for minimal chemical modifications that enhance PPARγ inverse agonism and elucidate their influence on the dynamic PPARγ conformational ensemble.

Science & Technology - Other Topics↗

Interplay Between Membrane Composition and Structural Stability of Membrane-Bound hIAPP

Amyloid aggregates are characteristic of many serious diseases such as Alzheimer’s disease, Parkinson’s, and type 2 diabetes and commonly involve intrinsically disordered proteins (IDPs), those that populate an ensemble of conformations rather than a single folded structure. Human islet amyloid polypeptide (hIAPP or amylin) is an amyloidogenic IDP implicated in pancreatic β-cell death during the pathogenesis of type 2 diabetes. The target of amylin’s toxic activity is thought to be the cell’s lipid membrane, which may also act as a catalyst for aggregation. Since amylin is intrinsically disordered, differing environments can have a large impact on its equilibrium conformational ensemble. We apply atomistic molecular dynamics simulations on multiple systems containing a full-length amylin monomer and a lipid bilayer to study the changes induced by the membrane. Here, we observe stabilized helical conformations structurally similar to those determined by NMR experiments conducted in similar environments. We also find that bilayers of different compositions result in greatly different equilibrium ensembles of amylin. Finally, we discuss how a mixed bilayer containing zwitterionic and anionic lipid headgroups can allow for greater preference toward conformations which are adsorbed below the membrane surface through rearrangement of lipids for more favorable protein–lipid interactions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Statistical theory of cubic Langmuir turbulence

The cubic direct interaction approximation is applied to a truncated (in Fourier space) version of the cubically nonlinear Schroedinger equation model of Langmuir physics. The results are compared (in the three-mode case) to those for an ensemble of numerical solutions of the dynamical equations with 10,000 different sets of Gaussianly distributed initial conditions. In the undriven, undamped case, the statistical theory (but not the ensemble) evolves to a state of thermal equilibrium. In the driven, damped case, the statistical theory appears to evolve to a state close to that corresponding to one of the limit cycles of the dynamical equations.

Sun, G.-Z.↗

Extracting Nucleon Magnetic Moments and Electric Polarizabilities from Lattice QCD in Background Electric Fields

Nucleon properties are investigated in background electric fields. As the magnetic moments of baryons affect their relativistic propagation in constant electric fields, electric polarizabilities cannot be determined without knowledge of magnetic moments. We devise combinations of baryon two-point functions in external electric fields to isolate both observables. Using an ensemble of anisotropic gauge configurations with dynamical clover fermions, we demonstrate how magnetic moments and electric polarizabilities can be determined from lattice QCD simulations in background electric fields. We obtain results for both the neutron and proton. Our study is currently limited to electrically neutral sea quarks.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Decoding Liquid Crystal Oligomer Phase Transitions: Toward Molecularly Engineered Shape Changing Materials

This work details an integrated investigation of liquid crystal (LC) oligomers that combines experiments and molecular dynamics simulations to obtain a detailed understanding of the molecular structure of LC oligomers and the mechanism underlying their phase transition temperatures. We synthesized and characterized a series of LC oligomers prepared from different lengths of methylene spacers in the reactive LC monomers and n-alkylamine chain extenders via the aza -Michael addition reaction. In parallel, we performed isothermal–isobaric ( NPT ) ensemble coarse-grained molecular dynamics (CG-MD) simulation of analogue mesogens that are connected to flexible spacers and extenders at varying temperatures, spacer lengths, and extender lengths. This approach allowed the effect of length in the flexible spacer as well as in the chain extender on the nematic–isotropic transition temperature ( T ni ) to be determined. The results showed that increasing the length of the extender decreases T ni for LC oligomers and amplifies the decrease of T ni in LC oligomers when the spacer length is short. We infer that the combination of spacer and extender changes the shape anisotropy of LC oligomers, changing the packing behavior of constituent mesogens, thus affecting their ability to transition from the isotropic to the nematic phase. Here, the detailed molecular structure–property relationships formulated enable prescribing design rules for LC oligomers geared toward molecularly engineered shape changing materials.

36 MATERIALS SCIENCE↗

Comparison of Dynamically and Statistically Downscaled Seasonal Climate Forecasts for the Cold Season over the United States

This study compares two approaches, dynamical and statistical downscaling, for their potential to improve regional seasonal forecasts for the United States (U.S.) during the cold season. In the MultiRCM Ensemble Downscaling (MRED) project, seven regional climate models (RCMs) are used to dynamically downscale the Climate Forecast System (CFS) seasonal prediction over the conterminous U.S. out to 5 months for the period of 1982–2003. The simulations cover December to April of next year with 10 ensemble members from each RCM with different initial and boundary conditions from the corresponding ensemble members. These dynamically downscaled forecasts are compared with statistically downscaled forecasts produced by two bias correction methods applied to both the CFS and RCM forecasts. Results of the comparison suggest that the RCMs add value in seasonal prediction application, but the improvements largely depend on location, forecast lead time, variables, and skill metrics used for evaluation. Generally, more improvements are found over the Northwest and North Central U.S. for the shorter lead times. The comparison results also suggest a hybrid forecast system that combines both dynamical and statistical downscaling methods have the potential to maximize prediction skill.

Comparison, cold season, downscaled, seasonal, cli↗

Multiscale polar theory of microtubule and motor-protein assemblies

Microtubules and motor proteins are building blocks of self-organized subcellular biological structures such as the mitotic spindle and the centrosomal microtubule array. These same ingredients can form new “bioactive” liquid-crystalline fluids that are intrinsically out of equilibrium and which display complex flows and defect dynamics. It is not yet well understood how microscopic activity, which involves polarity-dependent interactions between motor proteins and microtubules, yields such larger-scale dynamical structures. In our multiscale theory, Brownian dynamics simulations of polar microtubule ensembles driven by cross-linking motors allow us to study microscopic organization and stresses. Polarity sorting and cross-link relaxation emerge as two polar-specific sources of active destabilizing stress. On larger length scales, our continuum Doi-Onsager theory captures the hydrodynamic flows generated by polarity-dependent active stresses. Finally, the results connect local polar structure to flow structures and defect dynamics.

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↗

PySFD: comprehensive molecular insights from significant feature differences detected among many simulated ensembles

Motivation Many modeling analyses of molecular dynamics (MD) simulations are based on a definition of states that can be (groups of) clusters of simulation frames in a feature space composed of molecular coordinates. With increasing dimension of this feature space (due to the increasing size or complexity of a simulated molecule), it becomes very difficult to cluster the underlying MD data and estimate a statistically robust model. To mitigate this “curse of dimensionality”, one can reduce the feature space, e.g., with principal component or time-lagged independent component analysis transformations, focusing the analysis on the most important modes of transitions. In practice, however, all these reduction strategies may neglect important molecular details that are susceptible to experimental verification. Results To recover such molecular details, I have developed PySFD (Significant Feature Differences analyzer for Python), a multi-processing software package that efficiently selects significantly different features of any user-defined feature type among potentially many different simulated state ensembles, such as meta-stable states of a Markov State Model (MSM). Applying PySFD on MSMs of an aggregate of 300 microseconds MD simulations recently performed on the major histocompatibility complex class II (MHCII) protein, I demonstrate how this toolkit can extract and visualize valuable mechanistic information from big MD simulation data, e.g., in form of networks of dynamic interaction changes connecting functionally relevant sites of a protein complex.

Biochemistry & Molecular Biology↗

Analytic and numerical vibronic spectra from quasi-classical trajectory ensembles

The truncated Wigner approximation to quantum dynamics in phase space is explored in the context of computing vibronic line shapes for monomer linear optical spectra. We consider multiple model potential forms including a shifted harmonic oscillator with both equal and unequal frequencies on the ground and excited state potentials as well as a shifted Morse potential model. For the equal-frequency shifted harmonic oscillator model, we derive an analytic expression for the exact vibronic line shape that emphasizes the importance of using a quantum mechanical distribution of phase space initial conditions. For the unequal-frequency shifted harmonic oscillator model, we are no longer able to obtain an exact expression for the vibronic line shape in terms of independent deterministic classical trajectories. We show how one can rigorously account for corrections to the truncated Wigner approximation through nonlinear responses of the line shape function to momentum fluctuations along a classical trajectory and demonstrate the qualitative improvement in the resulting spectrum when the leading-order quantum correction is included. Finally, we numerically simulate absorption spectra of a highly anharmonic shifted Morse potential model. We find that, while finite quantization and the dissociation limit are captured with reasonable accuracy, there is a qualitative breakdown of the quasi-classical trajectory ensemble’s ability to describe the vibronic line shape when the relative shift in Morse potentials becomes large. The work presented here provides clarity on the origin of unphysical negative features known to contaminate absorption spectra computed with quasi-classical trajectory ensembles.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Models and Algorithms for Equilibrium Analysis of Mixed-Material Nucleic Acid Systems

Dynamic programming algorithms within the NUPACK software suite enable analysis of equilibrium base-pairing properties for complex and test tube ensembles containing arbitrary numbers of interacting nucleic acid strands. Currently, calculations are limited to single-material systems that are either all-RNA or all-DNA. Here, to enable analysis of mixed-material systems that are critical for modern applications in vitro, in situ, and in vivo, we develop physical models and dynamic programming algorithms that allow the material of the system to be specified at nucleotide resolution. Free energy parameter sets are constructed for both RNA/DNA and RNA/2'OMe-RNA mixed-material systems by combining available empirical mixed-material parameters with single-material parameter sets to enable treatment of the full complex and test tube ensembles. New dynamic programming recursions account for the material of each nucleotide throughout the recursive process. For a complex with N nucleotides, the mixed-material dynamic programming algorithms maintain the O(N 3 ) time complexity of the single-material algorithms, enabling efficient calculation of diverse physical quantities over complex and test tube ensembles (e.g., complex partition function, equilibrium complex concentrations, equilibrium base-pairing probabilities, minimum free energy secondary structure(s), and Boltzmann-sampled secondary structures) at a cost increase of roughly 2.0-3.5×. The results of existing single-material algorithms are exactly reproduced when applying the new mixed-material algorithms to single-material systems. Accuracy is significantly enhanced using mixed-material models and algorithms to predict RNA/DNA and RNA/2'OMe-RNA duplex melting temperatures from the experimental literature as well as RNA/DNA melt profiles from new experiments. In conclusion, mixed-material analyses can be performed online using the NUPACK web app (www.nupack.org) or locally using the NUPACK Python module.

2′OMe-RNA↗