Search NASA⌕ Search

SEARCH · Search NASA

Results for “Dynamic State Estimation”

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 1,513 records · Page 84
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Gaussian integral method for void fraction

Here, a novel method, the Gaussian Integral Method (GIM), is presented for calculating void fractions in Computational Fluid Dynamics–Discrete Element Method (CFD-DEM) simulations. GIM is versatile and applicable to various grid types, including structured and unstructured polyhedral meshes, without requiring special boundary treatments. An optimization technique is introduced to make GIM independent of grid resolution and type. The method is validated against experimental data from a fluidized bed, demonstrating that GIM produces realistic simulations closely resembling experimental observations. Additionally, unstructured polyhedral grids using GIM outperform structured grids of equivalent resolution, yielding results more aligned with experimental data. The gradient of the void fraction is computed in the CFD solver and utilized in the DEM solver for precise estimation at particle locations. Overall, GIM provides an effective solution for void fraction calculations in particulate media simulations with complex geometries, enhancing the accuracy and applicability of CFD-DEM simulations for industrial processes.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2026

Meta-analysis of North American Arctic and boreal aboveground biomass datasets: assessing accuracy, dynamics, and similarities

The North American arctic and boreal regions (ABRs) are rapidly warming and experiencing intensifying disturbances. Accurately quantifying aboveground biomass (AGB) is critical for understanding the impacts of these changes on the carbon cycle and for designing climate change mitigation strategies. Several AGB maps have been developed for the North American ABRs, including recent contributions from National Aeronautics and Space Administration’s Arctic-Boreal Vulnerability Experiment (ABoVE) campaign. However, these maps differ widely in training data, methodology, and resulting AGB density estimates. Presently, a comprehensive comparative evaluation is lacking, making it difficult for users to select datasets suited to their research or management needs. Here, in this study, we conducted a comparative analysis of nine AGB density datasets across North American ABRs, specifically for Alaska and Canada. We (1) summarized AGB by ecoregion and Canadian provinces, (2) evaluated their accuracy against field-based measurements, (3) analyzed spatial and temporal similarities among datasets, and (4) assessed their ability to capture disturbance (fire and harvest) impacts on AGB. We found substantial variation in regional and local AGB estimates across datasets, with overall accuracy ranging from R 2 = 0.25–0.62 and Bias% from −47.8% to 69.9% when validated against field plots. Despite these differences, most datasets have comparatively consistent spatial patterns in AGB (r > 0.8 for most cases). In contrast, agreement on the temporal patterns of AGB change is generally low. We found datasets with spatial resolutions ⩽300 m are capable of capturing disturbance impacts on AGB dynamics, though sensitivity varies across products. Our findings and dataset summary provide guidance for selecting appropriate AGB datasets for different applications within our study area. Our analysis also highlights the need to decrease map bias and increase capability to detect temporal change to decrease uncertainty of AGB datasets potentially by using training data which is representative of major plant functional types within the mapped area.

ABoVE↗
DOE OSTI2025

Bridging reaction theory and nuclear structure in $π^±-$ 48 Ca scattering

Here, we extend the pion-nucleus multiple-scattering framework to include detailed second-order rescattering dynamics for nuclei with nonzero isospin. To account for intermediate charge-exchange and nucleon spin-flip effects, we develop a scattering potential that depends on the one- and two-body densities of the target nucleus. We compute one-body densities from coupled-cluster theory and two-body densities within the Hartree-Fock approximation. To estimate theoretical uncertainties, we employ modern nuclear Hamiltonians derived from chiral effective field theory. While the sensitivity to nuclear structure details is mild, second-order corrections are found to be sizable and essential for accurately reproducing differential cross sections measured in 𝜋 ± − 48 Ca elastic scattering within the Δ⁡(1232)-resonance region.

cluster models↗
DOE OSTI2025

Real-time neutron multiplicity and source localization for criticality safety during fuel debris removal

Advancing neutron detection and analysis techniques for complex radiation environments is an ongoing focus in nuclear instrumentation and monitoring. This proposal presents research and development of a generalized real-time neutron monitoring and analysis system, applicable to any detector capable of producing time-tagged neutron count data. While the work is demonstrated using the Neutron Multiplication Analysis Detector (NoMAD), a modular 15-tube helium-3 (He-3) array, due to its availability, spatial resolution, and flexible deployment, the methods developed are extensible to other systems, including organic scintillators and fast digital detectors. This research investigates two complementary analytical techniques for real-time characterization of neutron emitting sources: neutron multiplicity estimation based on the Hage-Cifarelli formalism and spatial localization using supervised machine learning applied to spatial count rate patterns. These methods are designed to operate under dynamic, evolving conditions such as fuel debris retrieval or reactor startup, where neutron-emitting material geometries may be partially unknown or changing over time. By integrating statistical neutron emission data with spatial localization, this research aims to develop and evaluate methods for real time neutron monitoring, source characterization, and material verification. Key contributions include implementation of a low-latency data pipeline for continuous neutron multiplicity analysis, development and validation of machine learning models for spatial inference, and experimental evaluation of system performance under variable measurement conditions. The outcomes are intended to support applications in nuclear safeguards, verification, emergency response, and reactor startup.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2025

Energy-efficient, Large-scale Molecular Dynamics Simulations via Hardware- and Algorithm-level Optimization

This work aims to develop a framework for energy-efficient computing that will enable molecular dynamics (MD) simulations of large-scale phenomena with atomic precision and simultaneously remove computational bottlenecks limiting the speed of MD simulations. We seek to implement such an approach through the development of surrogate models for the interatomic force calculation combined with the use of mixed numerical precision formats. For a model system of neutral atoms (only pairwise interactions), significant force calculation efficiency improvements were achieved, without detrimental effects on atomic structures or average energies, using single precision, by developing a surrogate model (deep neural network), and by quantizing this surrogate model. For a model system of charged atoms, the reciprocal-space calculation of electrostatic interactions was identified as the main bottleneck, and the development of a surrogate model should be pursued to achieve an estimated one-order-of-magnitude additional speedup.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2024

Relating Hydro–Mechanical and Elastodynamic Properties of Dynamically Stressed Tensile–Fractured Rock in Relation to Applied Normal Stress, Fracture Aperture, and Contact Area

We exploit nonlinear elastodynamic properties of fractured rock to probe the micro-scale mechanics of fractures and understand the relation between fluid transport and fracture aperture under dynamic stressing. Experiments were conducted on rough, tensile-fractured Westerly granite subject to triaxial stresses. We measure fracture permeability for steady-state fluid flow with deionized water. Pore pressure oscillations are applied at amplitudes ranging from 0.2 to 1 MPa at 1 Hz frequency. During dynamic stressing we transmit ultrasonic signals through the fracture using an array of piezoelectric transducers (PZTs) to monitor evolution of interface properties. We examine the influence of fracture aperture and contact area by conducting measurements at effective normal stresses of 10–20 MPa. Additionally, the evolution of contact area with stress is characterized using pressure sensitive film. These experiments are conducted separately with the same fracture and map contact area at stresses from 9 to 21 MPa. The measurements are a proxy for “true” contact area for the fracture surface and we relate them to elastic properties using the calculated PZT sensor footprints via numerical modeling of Fresnel zones. We compare the elastodynamic response of the fracture using the stress-induced changes in ultrasonic wave velocities for transmitter-receiver pairs to image spatial variations in contact properties. We show that nonlinear elasticity and permeability enhancement decrease with increasing normal stress. Additionally, post-oscillation wave velocity and permeability exhibit quick recoveries toward pre-oscillation values. Estimates of fracture contact area (global and local) demonstrate that the elastodynamic and permeability responses are dominated by fracture topology.

58 GEOSCIENCES↗
DOE OSTI2023

Voltage-dependent excitation dynamics in UV-absorbing organic photovoltaics with efficient charge transfer exciton emission

Intermolecular charge-transfer excitons play a central role in determining the performance of organic solar cells as their voltage-dependent formation, dissociation, and recombination dynamics contribute to photocurrent generation, radiative/nonradiative voltage losses, and photovoltaic fill factor. Here, we explore the properties of brightly-emitting wide energy gap (>2 eV) charge transfer excitons by measuring the voltage-dependent photoluminescence, photocurrent, and ultrafast pump–probe transient absorption spectra of organic solar cells employing five UV-absorbing donor molecules that differ only by the length of the oligophenylene or acene group at their core. We find that organic solar cells with a strong correlation between their voltage-dependent photocurrent and charge-transfer exciton photoluminescence have low photovoltaic fill factors as they require voltage to facilitate efficient charge-transfer exciton dissociation. In contrast, solar cells that are efficient can readily generate charges without an applied field and have a separate population of tightly-bound charge-transfer excitons that are responsible for emission. Furthermore, considering that the sum of all excitation loss rates (i.e., recombination and charge extraction) must be equal to the excitation generation rate in the steady state, these voltage-dependent data allow us to solve for the voltage-dependent fate of all excitations in the solar cells and estimate upper and lower bounds for geminate and non-geminate recombination, respectively.

14 SOLAR ENERGY↗
DOE OSTI2024

Unravelling the orbits of cluster galaxy populations according to their dominant gas ionization source

ABSTRACT We investigate the kinematical and dynamical properties of cluster galaxy populations classified according to their dominant source of gas ionization, namely: star-forming (SF) galaxies, optical active galactic nuclei (AGNs), mixed SF plus AGN ionization (transition objects, T), and quiescent (Q) galaxies. We stack 8892 member galaxies from 336 relaxed galaxy clusters to build an ensemble cluster and estimate the observed projected profiles of numerical density and velocity dispersion, $\sigma _P(R)$, of each galaxy population. The MAMPOSSt code and the Jeans equations inversion technique are used to constrain the velocity anisotropy profiles of the galaxy populations in both parametric and non-parametric ways. We find that Q (SF) galaxies display the lowest (highest) typical cluster-centric distances and velocity dispersion values. Transition galaxies are more concentrated and tend to exhibit lower velocity dispersion values than SF galaxies. Galaxies that host an optical AGN are as concentrated as Q galaxies but display velocity dispersion values similar to those of the SF population. MAMPOSSt is able to find equilibrium solutions that successfully recover the observed $\sigma _P(R)$ profile only for the Q, T, and AGN populations. We find that the orbits of all populations are consistent with isotropy in the inner regions, becoming increasingly radial with the distance from the cluster centre. These results suggest that Q galaxies are in equilibrium within their clusters, while SF galaxies have more recently arrived in the cluster environment. Finally, the T and AGN populations appear to be in an intermediate dynamical state between those of the SF and Q populations.

Valk, Greique A. (ORCID:0009000827731299)↗
NASA NTRS2008

Potential Use of a Bayesian Network for Discriminating Flash Type from Future GOES-R Geostationary Lightning Mapper (GLM) data

Continuous monitoring of the ratio of cloud flashes to ground flashes may provide a better understanding of thunderstorm dynamics, intensification, and evolution, and it may be useful in severe weather warning. The National Lighting Detection Network TM (NLDN) senses ground flashes with exceptional detection efficiency and accuracy over most of the continental United States. A proposed Geostationary Lightning Mapper (GLM) aboard the Geostationary Operational Environmental Satellite (GOES-R) will look at the western hemisphere, and among the lightning data products to be made available will be the fundamental optical flash parameters for both cloud and ground flashes: radiance, area, duration, number of optical groups, and number of optical events. Previous studies have demonstrated that the optical flash parameter statistics of ground and cloud lightning, which are observable from space, are significantly different. This study investigates a Bayesian network methodology for discriminating lightning flash type (ground or cloud) using the lightning optical data and ancillary GOES-R data. A Directed Acyclic Graph (DAG) is set up with lightning as a "root" and data observed by GLM as the "leaves." This allows for a direct calculation of the joint probability distribution function for the lighting type and radiance, area, etc. Initially, the conditional probabilities that will be required can be estimated from the Lightning Imaging Sensor (LIS) and the Optical Transient Detector (OTD) together with NLDN data. Directly manipulating the joint distribution will yield the conditional probability that a lightning flash is a ground flash given the evidence, which consists of the observed lightning optical data [and possibly cloud data retrieved from the GOES-R Advanced Baseline Imager (ABI) in a more mature Bayesian network configuration]. Later, actual GLM and NLDN data can be used to refine the estimates of the conditional probabilities used in the model; i.e., the Bayesian network is a learning network. Methods for efficient calculation of the conditional probabilities (e.g., an algorithm using junction trees), finding data conflicts, goodness of fit, and dealing with missing data will also be addressed.

Solakiewiz, Richard↗
DOE OSTI2026

Bayesian chain graph models to characterize microbe-environment dynamics

Microbiome data require statistical models that can simultaneously decode microbes' reaction to the environment and interactions among microbes. While a multiresponse linear regression model seems like a straight-forward solution, we argue that treating it as a graphical model is problematic given that the regression coefficient matrix does not encode the conditional dependence structure between response and predictor nodes. This observation is especially important in biological settings when we have prior knowledge on the edges from specific experimental interventions that can only be properly encoded under a conditional dependence model. Here, we propose a chain graph model with two sets of nodes (predictors and responses) whose solution yields a graph with edges that indeed represent conditional dependence, thus agreeing with the experimenter's intuition on the average behavior of nodes under treatment. The solution to our model is sparse via the Bayesian linear regression (LASSO). In addition, we propose an adaptive extension so that different shrinkages can be applied to different edges to incorporate edge-specific prior knowledge. Our model is computationally inexpensive through an efficient Gibbs sampling algorithm and can account for binary, counting, and compositional responses via an appropriate hierarchical structure. We test the performance of our model in a variety of simulated datasets, thereby showing superior performance to state-of-the-art approaches. We further apply our model to human gut and soil microbial compositional datasets, and we highlight that CG-LASSO can estimate biologically meaningful network structures in the data.

compositional data↗
NASA NTRS1996

Simulation of Tropical Biomass Burning

The research objectives that were achieved during the course of our studies include the following: (1) Over the last few years, a model has been developed in the Atmospheric Chemistry and Dynamics Branch at Ames Research Center in collaboration with the Physics Department at San Jose State University. It is referred to as the Global/Regional Atmospheric Chemistry Event Simulator (GRACES). Currently, the GRACES model system combines an atmospheric chemistry and transport model, and a regional mesoscale meteorological model. Therefore this system is suitable for simulating the conditions observed by the tropical observation missions, such as the Pacific Exploratory Mission in the 'Tropics (PEM-Tropics), Study of Ozone and Nitrogen oxides Experiment (SONEX), and other periods. Specifically, the research carried out included the evaluation of the behavior of several components of the MM5 (I.e., Meteorological Model 5, version 2) and the GRACES combined modeling system. We initiated research on (a) the ability of the MM5 model to assimilate downward vertical velocities at least as high as the analyses, (b) the ability of the Graces model to incorporate the vertical velocities from MM5, and (c) other factors related to transport patterns required to transport CO in the observed manner. We carried out improved calculations of the transport of tracers for both NASA airborne missions, SONEX and PEM-Tropics. We also made improved source strength estimates fopr isoprene dust, and similar emissions from the Earths surface. This required the use of newly available databases on the Earth's surface and vegetation. We completed atmospheric chemistry simulations of radicals and nitrogen oxide species. We have greatly improved the handling of cumulonimnbus convection by modifing an existing scheme.

Hamill, Patrick↗
NASA NTRSDate not supplied

Influence of Heat Flow from the Planetary Interior on Surface Habitability

an primordial and/or radiogenic heat flow from a rocky planet’s interior help maintain a surface environment amenable to life?On rocky Earth-like planets with shallow oceans (≤5 km depth) that are on the outer edge of the habitable zone, the additional energy provided might make the difference between a world supporting regional habitat space and a world uninhabitable by any known life-form. To test this hypothesis, we introduce heat flux from the planetary interior as a forcing to the NASA/GISSROCKE-3D exoplanet GCM, and explore whether it can enhance the habitability of marginal terrestrial rocky planets (e.g.,paleo Earth in a “snowball” climate state).GCM simulations of modern Earth have not included geothermal heat flux as a climate forcing, as it is more than an order of magnitude less than the 2.9 W/m heating caused by anthropogenic greenhouse gases [Flanner, GRL 2009]. Modern Earth measurements in the oceans [Downes et al., GRL 2019] and results from ocean GCMs [Barnes et al., Ocean Model. 2017]suggest that localized higher heat flux can have measurable thermodynamic and dynamic impacts, such as weakening deep ocean stratification and increased poleward ocean heat transports. A similar result with ROCKE-3D would be important, since our snowball Earth simulations without heat flux produce features like ocean stratification that should have hampered the survival of life. We have created a set of spatially variable geothermal heat fluxes at varying horizontal resolutions based on modern Earth heat flow patterns [Davies, Geochem. Geophys. Geosys. 2013] to test the sensitivity of the ROCKE-3D GCM initially in a modern Earth context. Preliminary results show that even modest heat fluxes have a measurable impact on deep ocean temperatures (upto +0.9°C locally) and that this heat can be transported at depth along ocean circulation paths. Estimates of heat flux impacts on vertical ocean mixing and surface ocean conditions require long simulations at the higher spatial resolutions needed to express heat flow at the scale of most ocean ridges. Ongoing simulations are focused on testing the impacts of horizontal resolution and flux scaling on surface environments, especially for “snowball” climate scenarios.

Heat flow↗
DOE OSTI2025

Fine-scale vegetation composition and structure shape spatiotemporal variation in surface albedo across a low Arctic tundra landscape

The unprecedented rate of warming in the Arctic is driving changes in the structure and composition of tundra vegetation. Increases in deciduous tall shrub cover, height, and density are of particular concern, as these changes alter local surface albedo in ways that could amplify effects on the regional surface energy budget (SEB). Despite this importance, significant uncertainties remain in understanding the interplay between fine-scale vegetation patterns and emergent albedo dynamics across space and time. Here, we address these uncertainties by (1) quantifying spatiotemporal variation in surface shortwave albedo and (2) determining the relative influence of fine-scale vegetation composition, structure, and environmental conditions on albedo across a representative low-Arctic tundra landscape on Alaska’s Seward Peninsula. To do this, we synthesized multi-scale, multi-platform remote sensing observations, including a novel Landsat-derived albedo time series, a fine-scale map of Arctic plant functional type (PFT) fractional cover, and airborne LiDAR estimates of canopy height and topography. We show that there are substantial reductions in winter albedo for pixels dominated by tall, woody PFTs (28.13%) relative to pixels dominated by non-woody vegetation, but almost no change in summer albedo (3% increase). Further, we identified a unimodal trend in the relationship between canopy height and the timing of the springtime transition from high (snowy) to low (leafy) albedo (peak at 5.5 m), possibly because of competing ‘snow-fence’ and ‘protrusion’ snow-shrub interactions. To explore the primary drivers of albedo, we constructed a random forest model and found that canopy height and the fractional cover of woody PFTs were as- or more important predictors of winter albedo than topographic features. These findings provide strong evidence for the impacts of local vegetation characteristics on regional surface albedo, highlighting the need for better quantification of snow-shrub interactions to accurately predict the Arctic’s SEB under future environmental change.

Arctic↗
DOE OSTI2024

Photosynthetic and stomatal response, and leaf traits, of two pine species, Talladega National Forest, Alabama, 2023

Photosynthetic response to irradiance (AQ), CO2 (ACi) and stomatal response to irradiance measurements on two pine species common to the southeast United States, long leaf pine, Pinus palustris Mill., and loblolly pine, Pinus taeda L.. Data were collected in June 2023 within the Oakmulgee District of the Talladega National Forest in west-central Alabama at the NEON site (TALL). Measurements were made with four LI-COR LI-6800 and one LI-COR LI-6400XT gas exchange systems. ACi and AQ curves enable the estimation of photosynthetic parameters (e.g., Vcmax, Jmax, Rd, aQY), while stomatal response curves enable the estimation of stomatal slope and intercept parameters. Foliar trait data (needle dimensions, leaf mass per area, leaf nitrogen concentration) is provided for all gas exchange samples. The data package files include multiple raw and processed data files (csv), metadata files (csv), the complete instrument output for all measurements in .zip format, and a pdf describing the experimental protocol.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1993

Description and availability of airborne Doppler radar data

An airborne, forward-looking, pulse, Doppler radar has been developed in conjunction with the joint FAA/NASA Wind Shear Program. This radar represents a first in an emerging technology. The radar was developed to assess the applicability of an airborne radar to detect low altitude hazardous wind shears for civil aviation applications. Such a radar must be capable of looking down into the ground clutter environment and extracting wind estimates from relatively low reflectivity weather targets. These weather targets often have reflectivities several orders of magnitude lower than the surrounding ground clutter. The NASA radar design incorporates numerous technological and engineering achievements in order to accomplish this task. The basic R/T unit evolved from a standard Collins 708 weather radar, which supports specific pulse widths of 1-7 microns and Pulse Repetition Frequencies (PRF) of less than 1-10 kHz. It was modified to allow for the output of the first IF signal, which fed a NASA developed receiver/detector subsystem. The NASA receiver incorporated a distributed, high-speed digital attenuator, producing a range bin to range bin automatic gain control system with 65 dB of dynamic range. Using group speed information supplied by the aircraft's navigation system, the radar signal is frequency demodulated back to base band (zero Doppler relative to stationary ground). The In-phase & Quadrature-phase (I/Q) components of the measured voltage signal are then digitized by a 12-bit A-D converter (producing an additional 36 dB of dynamic range). The raw I/Q signal for each range bin is then recorded (along with the current radar & aircraft state parameters) by a high-speed Kodak tape recorder.

Harrah, S. D.↗
NASA NTRS1996

Electronic Characterization of Defects in Narrow Gap Semiconductors-Comparison of Electronic Energy Levels and Formation Energies in Mercury Cadmium Telluride, Mercury Zinc Telluride, and Mercury Zinc Selenide

We have used a Green's function technique to calculate the energy levels and formation energy of deep defects in the narrow gap semiconductors mercury cadmium telluride (MCT), mercury zinc telluride (MZT) and mercury zinc selenide (MZS). The formation energy is calculated from the difference between the total energy with an impurity cluster and the total energy for the perfect crystal. Substitutional (including antisite), interstitial (self and foreign), and vacancy deep defects are considered. Relaxation effects are calculated (with molecular dynamics). By use of a pseudopotential, we generalize the ideal vacancy model so as to be able to consider relaxation for vacancies. Different charge states are considered and the charged state energy shift (as computed by a modified Haldane-Anderson model) can be twice that due to relaxation. Different charged states for vacancies were not calculated to have much effect on the formation energy. For all cases we find deep defects in the energy gap only for cation site s-like orbitals or anion site p-like orbitals, and for the substitutional case only the latter are appreciably effected by relaxation. For most cases for MCT, MZT, MZS, we consider x (the concentration of Cd or Zn) in the range appropriate for a band gap of 0.1 eV. For defect energy levels, the absolute accuracy of our results is limited, but the precision is good, and hence chemical trends are accurately predicted. For the same reason, defect formation energies are more accurately predicted than energy level position. We attempt, in Appendix B, to calculate vacancy formation energies using relatively simple chemical bonding ideas due to Harrison. However, these results are only marginally accurate for estimating vacancy binding energies. Appendix C lists all written reports and publications produced for the grant. We include abstracts and a complete paper that summarizes our work which is not yet available.

Patterson, James D.↗
DOE OSTI2023

Efficient Subset Simulation using Hamiltonian Neural Network enhanced Markov Chain Monte Carlo Methods

The Monte Carlo method delivers an unbiased estimate of the probability of failure. However, the variance of the estimate depends on the number of evaluated samples. This number must be very large for estimations of a low probability of failure. If the evaluation of each sample is computationally expensive, the crude Monte Carlo simulation strategy is impracticable. Therefore, subset simulations are used to reduce the required number of evaluations. Subset simulations require a Markov Chain Monte Carlo sampler, such as the random walk Metropolis-Hastings algorithm. The algorithm, however, struggles with sampling in low-probability regions, especially if they are narrow. As a consequence, advanced Markov Chain Monte Carlo simulations have been developed. In particular, the Hamiltonian Monte Carlo method explores the target distribution rapidly. Driven by the idea of Hamiltonian dynamics, this sampler provides a non-random walk through the target distribution. The incorporation of subset simulation and Hamiltonian Monte Carlo methods has shown promising results for reliability analysis. One downside of the Hamiltonian Monte Carlo method is that gradient evaluations are computationally expensive, especially when dealing with high-dimensional problems and evaluating long trajectories. We show that integrating Hamiltonian neural networks in Hamiltonian Monte Carlo simulations significantly speeds up the sampling task. Furthermore, the enhancement of adaptive trajectory length within the Hamiltonian Monte Carlo results in the efficient proposal of the following states. Based on this recent enhancement, we provide a fast sampling strategy for subset simulations using Hamiltonian neural networks to replace the evaluation of the gradient and significantly speed up the Hamiltonian Monte Carlo simulation.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Machine learning-based ethylene concentration estimation, real-time optimization and feedback control of an experimental electrochemical reactor

With the increase in electricity supply from clean energy sources, electrochemical reduction of carbon dioxide (CO 2 ) has received increasing attention as an alternative source of carbon-based fuels. As CO 2 reduction is becoming a stronger alternative for the clean production of chemicals, the need to model, optimize and control the electrochemical reduction of the CO 2 process becomes inevitable. However, on one hand, a first-principles model to represent the electrochemical CO 2 reduction has not been fully developed yet because of the complexity of its reaction mechanism, which makes it challenging to define a precise state-space model for the control system. On the other hand, the unavailability of efficient concentration measurement sensors continues to challenge our ability to develop feedback control systems. Gas chromatography (GC) is the most common equipment for monitoring the gas product composition, but it requires a period of time to analyze the sample, which means that GC can provide only delayed measurements during the operation. Moreover, the electrochemical CO 2 reduction process is catalyzed by a fast-deactivating copper catalyst and undergoes a selectivity shift from the product-of-interest at the later stages of experiments, which can pose a challenge for conventional control methods. To this end, machine learning (ML) techniques provide a potential approach to overcome those difficulties due to their demonstrated ability to capture the dynamic behavior of a chemical process from data. Motivated by the above considerations, we propose a machine learning-based modeling methodology that integrates support vector regression and first-principles modeling to capture the dynamic behavior of an experimental electrochemical reactor; this model, together with limited gas chromatography measurements, is employed to predict the evolution of gas-phase ethylene concentration. The model prediction is directly used in a proportional-integral (PI) controller that manipulates the applied potential to regulate the gas-phase ethylene concentration at energy-optimal set-point values computed by a real-time process optimizer (RTO). Specifically, the RTO calculates the operation set-point by solving an optimization problem to maximize the economic benefit of the reactor. Finally, suitable compensation methods are introduced to further account for the experimental uncertainties and handle catalyst deactivation. The proposed modeling, optimization, and control approaches are the first demonstration of active control for a CO 2 electrolyzer and contribute to the automation and scale-up efforts for electrified manufacturing of fuels and chemicals starting from CO 2 .

42 ENGINEERING↗