SEARCH · Engineering Papers

Results for “model data volumes”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 901 records · Page 50

All sources NASA NTRS DOE OSTI

DOE OSTI2025

MARIE: A Python-Based Framework for Comprehensive Fuel Recycling Modeling

One of the most pressing challenges to the continued deployment of nuclear energy systems is in the ultimate management and disposition of discharged fuel assemblies. While reprocessing and recovery of valuable materials from UNF assemblies has been considered as part of an overall strategy for minimization of the volume of reactor-based wastes to be managed, the deployment of commercial-scale reprocessing facilities presents an enormous economic challenge. The MARIE software package has been developed as a means of confronting this challenge. Representing components of a generic fuel reprocessing operation as individual physical processes, MARIE is designed as a modular framework intended to allow for analysis and cost-optimization for a hypothetical reprocessing facility while realistically accounting for the physical characteristics of the used fuel source term, such as decay heat, activity, and radiation dose (informing corresponding shielding requirements). Capabilities supported by MARIE include head-end operations such as fuel shearing, voloxidation, and dissolution; generic solvent extraction operations informed by available open-literature data; a suite of unit operations intended to represent electrochemical processing of used fuel assemblies (i.e., oxide reduction, electrorefining, and electrowinning); and finally, accounting for both costs and physical features of discharged waste streams, which can be used to inform follow-on analyses such as the feasibility of deep-borehole disposal of HLW. This paper presents an overview of the MARIE software capabilities, including how individual unit operations are implemented to enable a larger-scale optimization of a hypothetical reprocessing operation on aspects such as cost and recovery of valuable materials.

Skutnik, Steve [ORNL] (ORCID:000000016441135X)↗

DOE OSTI2023

Experimental measurement of the rapid mixing of fuel and air in a multi-element diffusion (Hencken) burner

Flat flame burners are valuable tools for studying complex combustion phenomena like particle synthesis due to the well-controlled and well-behaved initial and boundary conditions. In this study, the methane/air flame temperature and tracer density profiles of a multi-element diffusion burner equipped with a central tube were characterized using krypton (Kr) x-ray fluorescence at the Advanced Photon Source at Argonne National Laboratory. The experiments were performed using a fuel-to-oxygen equivalence ratio of 0.8 and 3.3% Kr by volume was added to the reactant gases. One- and two-dimensional scans of non-combusting (i.e., cold flow) and combusting conditions, both with and without flow through the central tube were performed. Spatially-resolved Kr number density profiles were obtained at different planar locations and heights above the surface of the burner. The results show rapid mixing occurred near the exit-plane of the burner and increased spatial uniformity was achieved from 0.1 mm to 0.5 mm above the surface of the burner. Further, the data also showed the impact of the central tube flow on the main methane/air flame, including evidence of the central tube flow mixing with the main flow. Flame temperatures were calculated from the fluorescence measurements at various heights above the burner surface and were in good agreement with model predictions for a burner stabilized flame. These high-fidelity measurements provide accurate baseline data that are particularly valuable for understanding the effects of fuel and air mixing on downstream phenomena, such as the formation of condensed-phase products.

42 ENGINEERING↗

DOE OSTI2023

Utah FORGE: Optimization of a Plug-and-Perf Stimulation (Fervo Energy)

Information around the plug-and-perf treatment design at Utah FORGE by Fervo Energy. Objective and Purpose: - Develop a multistage hydraulic stimulation approach designed specifically to target the top three factors that control the technical and commercial viability of an EGS system: i) Achieving sufficient injectivity to support high cross-well flow rates ii) Distributing flow evenly across the wellbore and reservoir to maximize heat mining efficiency, ensure sustained heat transfer, and mitigate thermal breakthrough iii) Overcoming the effects of stress heterogeneity, stress shadowing, and variations in natural fracture properties during the stimulation treatment, leading to a more predictable stimulated reservoir volume and offset well placement - The following activities will be performed: i) Design, plan, and execute a multistage plug-and-perf stimulation treatment at a Fervo site with data acquisition and well testing activities aimed at addressing key technical aspects of the issues above ii) Perform data processing and interpretation of field results to translate the results form the Fervo site to a site-specific design at the Utah FORGE site iii) Design, plan, and execute a multistage plug-and-perf stimulation treatment design at the Utah FORGE site Methods and Approach: - Design a detailed data acquisition plan to maximize learning around: i) DFIT testing ii) Petrophysical logging, image logging iii) Permanent DAS/DTS fiber optic monitoring iv) Deep borehole microseismic monitoring v) Shallow borehole induced seismicity monitoring vi) Injection/production testing (RTA analysis, tracer testing) vii) Integrated numerical modeling and production forecasting

15 GEOTHERMAL ENERGY↗

DOE OSTI2023

Cost and Performance Baseline for Fossil Energy Plants, Volume 5: Natural Gas Electricity Generating Units for Flexible Operation

To address the data needs of energy system designers and to serve as a baseline for research and development, NETL has carried out a study to characterize the flexibility attributes - both performance and cost - of nine common commercial natural gas-fueled electricity generating units. The intermittent output of low-carbon, renewable power generation sources such as wind and solar create challenges to grid stability and reliability. Fossil-fueled power generation technologies are currently used to provide reliable, on-demand power during periods of reduced renewable output. Dispatchable generators must be able to accommodate increasing renewable generation as the nation pursues the Administration’s target of a decarbonized energy sector by 2035. As energy system experts seek to identify least-cost approaches to decarbonization, accurate cost and performance data characterizing dispatchable fossil generators that operate flexibly, at capacity factors that have been declining over time, and are needed to inform models for capacity expansion. Furthermore, these technologies continue to be a significant source of carbon dioxide emissions, providing the impetus for research and development, including the advancement and potential incorporation of carbon capture technologies. This study characterizes the cost and performance of select state-of-the-art natural gas-fueled power generation technologies: reciprocating internal combustion engines (RICE), simple cycle combustion turbines, and natural gas combined cycles (NGCC). An emphasis is placed on flexibility characteristics, such as part-load heat rate, ramp rates, start up times, and start up costs.

03 NATURAL GAS↗

DOE OSTI2020

A Spitzer survey of Deep Drilling Fields to be targeted by the Vera C. Rubin Observatory Legacy Survey of Space and Time

The Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST) will observe several Deep Drilling Fields (DDFs) to a greater depth and with a more rapid cadence than the main survey. In this paper, we describe the ‘DeepDrill’ survey, which used the Spitzer Space Telescope Infrared Array Camera (IRAC) to observe three of the four currently defined DDFs in two bands, centred on 3.6 and 4.5 μm. Furthermore, these observations expand the area that was covered by an earlier set of observations in these three fields by the Spitzer Extragalactic Representative Volume Survey (SERVS). The combined DeepDrill and SERVS data cover the footprints of the LSST DDFs in the Extended Chandra Deep Field–South (ECDFS) field, the ELAIS-S1 field (ES1), and the XMM-Large-Scale Structure Survey field (XMM-LSS). The observations reach an approximate 5σ point-source depth of 2 μJy (corresponding to an AB magnitude of 23.1; sufficient to detect a 10$^{11} \, \mathrm{M}_{\odot}$ galaxy out to z ≈ 5) in each of the two bands over a total area of $\approx 29\,$ deg2. The dual-band catalogues contain a total of 2.35 million sources. In this paper, we describe the observations and data products from the survey, and an overview of the properties of galaxies in the survey. We compare the source counts to predictions from the Shark semi-analytic model of galaxy formation. We also identify a population of sources with extremely red ([3.6]-[4.5] >1.2) colours which we show mostly consists of highly obscured active galactic nuclei.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2020

Comparative Study of Hydrology and Icemelt in Three Nepal River Basins Using the Glacio-Hydrological Degree-Day Model (GDM) and Observations From the Advanced Scatterometer (ASCAT)

An assessment of the water supply and its seasonal and annual changes over the century in the High Mountain Asia (HMA) region is of increasing interest due to its potential impact on one-sixth of the global population. In order to understand the changing hydrology and snow and ice melt, we used remotely sensed Advanced Scatterometer (ASCAT) observations of glacier melt (GM) and a distributed and gridded Glacio-hydrological Degree-day Model (GDM) in three river basins: Tamor, Trishuli and Marsyangdi. The GDM-estimated contribution of snowmelt, icemelt, rainfall and baseflow in river flows is found to be most accurate in the Trishuli River basin, with Nash-Sutcliffe efficiency (NSE) between the estimated and observed discharges of 0.81 and volume differences of -0.5%, and reasonably accurate in the Tamor River basin, with NSE of 0.69 and volume difference of -7.51%. Similarly, NSE of 0.81 and volume difference of 4.64% in Marsyangdi River basin. We find strong similarities in the timing of glacier melting using the GDM and from observations from the ASCAT GM, determining the seasonal start of glacier melting to within 6 days on average. In all basins ASCAT GM observes melting at higher elevations relative to GDM, average of 5,328 m a.s.l. Systematic differences in glacier melting area determined by modeling and satellite observations indicate ASCAT may have suboptimal resolution, view geometry and/or polarimetry for delineating glacier melting at the process-scale in complex topography, especially in the ablation zone. This is the first step in examining the remote sensing products that could potentially be incorporated into hydrologic models to increase the accuracy of the hydrologic flow as well as the ability to estimate river discharge in other basins with limited data.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2023

Virtual Volumetric Additive Manufacturing (VirtualVAM)

Abstract Tomographic volumetric additive manufacturing (VAM) produces arbitrary 3D geometries by exposure of a rotating volume of photopolymer resin to tomographically‐patterned illumination. This enables high speed, layer‐less printing of parts from a wide range of photopolymers not amenable to layer‐by‐layer processes. Since the entire geometry is produced at once over the course of a few seconds to minutes, molecular diffusion length scales become significant to the printing process. Understanding these molecular reaction and diffusion processes is imperative for advancing VAM to a usable technology. These processes are experimentally very difficult to monitor and measure. Herein, VirtualVAM ‐ a simulation framework for modeling the tomographic VAM process, is developed and experimentally validated. VirtualVAM simulates reaction, diffusion, and heat generation processes over the course of a print with single‐voxel resolution. From a few experimentally‐determined input parameters and a set of images for projection, VirtualVAM is able to generate a large spatio‐temporal data set for any given tomographic VAM print. Using VirtualVAM, a number of experimentally‐unattainable aspects of the VAM process are investigated such as single‐voxel conversion profiles, effect of molecular oxygen, and stopping time determination. VirtualVAM also enables the optimization of exposure patterns to further improve contrast between in‐part and out‐of‐part delivered dose.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2023

Investigation of K + K – interactions via femtoscopy in Pb-Pb collisions at s N N = 2.76 TeV at the CERN Large Hadron Collider

Femtoscopic correlations of nonidentical charged kaons (K + K – ) are studied in Pb-Pb collisions at a center-of-mass energy per nucleon-nucleon collision $\sqrt{s_{NN}}$=2.76 TeV by ALICE at the CERN Large Hadron Collider. One-dimensional K + K – correlation functions are analyzed in three centrality classes and eight intervals of particle-pair transverse momentum. The Lednický and Luboshitz interaction model used in the K + K – analysis includes the final-state Coulomb interactions between kaons and the final-state interaction through a 0 (980) and f 0 (980) resonances. The mass of f 0 (980) and coupling were extracted from the fit to K + K – correlation functions using the femtoscopic technique. The measured mass and width of the f 0 (980) resonance are consistent with other published measurements. The height of the $\phi$(1020) meson peak present in the K + K – correlation function rapidly decreases with increasing source radius, qualitatively in agreement with an inverse volume dependence. A phenomenological fit to this trend suggests that the $\phi$(1020) meson yield is dominated by particles produced directly from the hadronization of the system. The small fraction subsequently produced by final-state interactions could not be precisely quantified with data presented in this paper and will be assessed in future work.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2023

Analysis of an optical imaging system prototype for autonomously monitoring zooplankton in an aquaculture facility

Traditional approaches to biomonitoring in aquatic systems, such as sample collection, sorting, and identification, require significant time and effort, thereby limiting the spatiotemporal resolution of sample collection. Additionally, collection and preservation of samples for subsequent taxonomic identification and enumeration leads to mortality of organisms. Recent advances in technologies that utilize optical imaging and machine learning have provided new opportunities to expedite biomonitoring and lead to significant cost savings. These technologies can be advantageous to scientists or managers that conduct routine biomonitoring to inform operations, as in the case of aquaculture facilities. The Small Aquatic Organism optical imaging system (SAO) is a high-throughput optical imaging and classification prototype system that relies on computer vision and machine learning (Support Vector Machines, or SVMs) to autonomously identify and enumerate aquatic organisms. The SAO provides a more sustainable method of collecting large volumes of data and has the benefit of being used in situ. In this study, we tested the performance of the SAO in providing comparable results to manual zooplankton community monitoring in ten ponds at an aquaculture facility. We performed a side-by-side study comparing the sampling methods of plankton tow nets, where major zooplankton taxonomic classes were manually identified and enumerated, to sampling with the SAO. Vouchered samples were used to develop a training library for the SAO, where classes consisted of water boatman and zooplankton groups: cladocerans, copepod adults, copepod nauplii, and rotifers. SAO imagery was manually classified and compared with predicted results for validation. Accuracy for the SVM classifier of the SAO was 37.4 %. Convolutional Neural Networks (CNN) and Random Forest classifiers were also applied to SAO imagery and image features for comparison. The best CNN model and our Random Forest model had accuracies of 80.4 % and 46.6 % respectively. Challenges faced included the small size of copepod nauplii and rotifers and the limited resolution of the imaging camera, although there are tradeoffs between imaging resolution and the sample processing rate. Furthermore, our comparison shows that advancement in both optical imaging and ML are needed in order for the SAO prototype to yield comparable results to manual community monitoring in an aquaculture facility.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2024

Interpreting strain tensor data to characterize and monitor reservoirs for CO2 storage and other applications

Recent advances in instrumentation have made it feasible to measure the transient strain tensor caused by small changes in fluid volume or pressure in the subsurface and this has opened the door to new opportunities for characterization and monitoring. We have deployed strainmeters and then conducted injection well tests in an underlying reservoir at 530m depth. The resulting data indicated that the horizontal strain at shallow strainmeters (30 to 40m depth) was tensile and the vertical strain was compressive. The radial strain was less than the horizontal strain, and the strain rates decreased from 100 nanostrain/day to roughly 10 n/d over a few days. The signal at two strainmeters at shallow depth were consistent, although the magnitude of the horizontal strains were different reflecting the different radial directions from the well. The signal at a deep strainmeter deployed at reservoir depth was much different, with tensile vertical strains and compressive horizontal strains. These data can be interpreted by inverting poroelastic forward models developed using numerical and analytical methods. The average horizontal strain in the caprock resembles the transient pressure in the underlying reservoir and classic type-curve methods from transient well testing can be used for preliminary interpretations of strain data. We have developed closed-form analytical solutions to a pressurized poroelastic inclusion and inhomogeneity in a half-space. This model is fast and can be inverted to estimate reservoir stiffness and geometry. Numerical models developed using finite element methods allow more details of the subsurface to be included in the inversion, but they require much longer run times and this makes inversion cumbersome using standard methods. We have developed an inversion approach that uses a proxy model created using machine learning to do most of the forward calculations. The proxy model is periodically updated and refined using the finite element model to ensure accuracy. This approach significantly reduces the computational requirements and makes it feasible to use Bayesian inversion with large numerical models. We have shown that the strain tensor in the caprock is sensitive to pressure in the reservoir, boundaries in the reservoir, and pressure in the caprock caused by leaks. These results indicate that coupling strain tensor data with inversion has the potential to help evaluate reservoirs during initial characterization, and to monitor them during the CO2 injection and storage process.

Murdoch, Larry↗

DOE OSTI2026

Object-Based Evaluation of Dynamical and Statistical Downscaled Precipitation Products over CONUS

High-resolution precipitation data, generated through dynamical downscaling (DD) or statistical downscaling (SD) of global climate model output, provide critical information for regional climate assessment and adaptation planning. Most downscaling development and validation have focused on accurate gridscale precipitation construction and ignored the spatial structure of precipitation across model grids and at the event scale. However, many applications, e.g., hydrologic modeling and the analysis using the downscaled precipitation, require a reasonable representation of the spatial structure of precipitation within watersheds. Therefore, a set of standard metrics to evaluate the representation of the spatial structure of individual storms across diverse downscaled precipitation products is desired. To address this need, we conducted an object-based evaluation of precipitation in decades-long DD and SD products over the contiguous United States (CONUS). Specifically, we evaluate their ability to reproduce various features of precipitation objects in the observations: total volume, precipitation area, peak intensity, and spatial structure. Multiple metrics (bias, Perkins score, and nonparametric statistical tests) are used to quantify model performance. Our evaluation reveals notable variations in performance among individual products across different climate zones and seasons, as well as between extreme and nonextreme events. In general, most DD products exhibit balanced performance across the four precipitation object features, while SD products vary more significantly in their performance across products. Based on this comprehensive evaluation, we provide guidance on choosing downscaled products for specific regions, seasons, and precipitation object features. These findings and recommendations can inform precipitation-relevant modeling and analysis over CONUS, guide future downscaling technique developments, and provide actionable information for climate impact assessment and adaptation.

Downscaling↗

DOE OSTI2026

Object-Based Evaluation of Dynamical and Statistical Downscaled Precipitation Products over CONUS

Environmental sciences↗

DOE OSTI2021

Evaluating causal‐based feature selection for fuel property prediction models

Abstract In‐silico screening of novel biofuel molecules based on chemical and fuel properties is a critical first step in the biofuel evaluation process due to the significant volumes of samples required for experimental testing, the destructive nature of engine tests, and the costs associated with bench‐scale synthesis of novel fuels. Predictive models are limited by training sets of few existing measurements, often containing similar classes of molecules that represent just a subset of the potential molecular fuel space. Software tools can be used to generate every possible molecular descriptor for use as input features, but most of these features are largely irrelevant and training models on datasets with higher dimensionality than size tends to yield poor predictive performance. Feature selection has been shown to improve machine learning models, but correlation‐based feature selection fails to provide scientific insight into the underlying mechanisms that determine structure–property relationships. The implementation of causal discovery in feature selection could potentially inform the biofuel design process while also improving model prediction accuracy and robustness to new data. In this study, we investigate the benefits causal‐based feature selection might have on both model performance and identification of key molecular substructures. We found that causal‐based feature selection performed on par with alternative filtration methods, and that a structural causal model provides valuable scientific insights into the relationships between molecular substructures and fuel properties.

Nguyen, Bernard↗

DOE OSTI2022

Battery Energy Storage Scenario Analyses Using the Lithium-Ion Battery Resource Assessment (LIBRA) Model

Meeting aggressive carbon emission goals will entail widespread deployment of renewable sources of electricity. Because these sources are variable, there is a need to develop scalable energy storage technologies. The U.S. Department of Energy is supporting efforts to increase U.S. manufacturing and recycling capabilities for LIBs and to decrease costs of stationary storage batteries. Many factors influence the domestic manufacturing and cost of stationary storage batteries, including availability of critical raw materials (lithium, cobalt, and nickel), competition from various demand sectors (consumer electronics, vehicles, and battery energy storage), resource recovery (recycling), government policies, and learning in the industry, among other factors. Understanding how these factors interact and identifying synergies and bottlenecks is important for developing effective strategies for the LIB stationary energy storage system. We developed the Lithium-Ion Battery Resource Analysis (LIBRA) model as a tool to help stakeholders better understand the following types of questions: What are the roles of R&D, industrial learning, and scaling of demand in lowering the cost of battery energy storage system production? How do the intersections between the EV and stationary storage sectors affect the battery supply chain? For various stationary storage and EV penetration scenarios, what volumes of critical materials might be required and what role can resource recovery play? What does expected demand for both EVs and stationary storage portend for mineral resources and overall mineral scarcity? The LIBRA model is developed using a System dynamics (SD) modeling approach to represent interactions across the segments of the battery materials supply chain. System dynamics models can capture the complex interactions and feedback between the various system components that influence supply and demand. The LIBRA model is comprised of several interacting modules that represent specific portions of the LIB supply chain. The model tracks the buildout of the domestic LIB industry over time (2020 - 2050) and in the context of competing demands for raw materials, recycling, and markets for LIBs. The LIBRA model represents major systemic feedback loops and delays across the supply chain. This report provides a complete documentation for the LIBRA model, including model assumptions, data, scenario analysis results, and sensitivity analysis of the model's input space.

25 ENERGY STORAGE↗

DOE OSTI2022

Potassium carbonate decomposition modeling within rotating detonation engines for direct power extraction applications

We report that recently, there has been a significant interest in detonation-based combustion systems, such as rotating detonation engines (RDE), due to potential performance advantages in propulsion and energy applications. This includes improvements in thermodynamic efficiency, inherently high gas velocities, and the possibility of achieving an increase in total pressure (i.e. pressure gain) through the combustion process. Due to approximating a constant volume combustion process, these devices also tend to exhibit extremely high local gas temperatures relative to a comparable constant pressure combustor. These advantages overlap well with desirable performance characteristics of direct power extraction (DPE) technologies, such as a magnetohydrodynamic (MHD) generator. Typically, in DPE systems hot combustion products are seeded with an easily ionizable material such as potassium carbonate (K 2 CO 3 ) in order to boost the electrical conductivity. However, due to the short gas residence times within an RDE, it was unclear whether forming an electrically conductive combustion plasma would be feasible for integration with a downstream MHD generator. A model is presented which describes the heating, decomposition, and ionization of solid K 2 CO 3 particles and aqueous solutions of K 2 CO 3 in water, for a given initial particle diameter. This model was combined with available computational fluid dynamics (CFD) data for an oxygen-methane RDE in a one-way coupled Eulerian-Lagrangian framework to predict particle trajectories and the corresponding heating, decomposition, and ionization histories. Electrical conductivities were computed using a previously developed model, and a method was proposed to determine an equivalent average electrical conductivity. Results show that particle sizes below ~30 µm are able to fully decompose before reaching the exit of the RDE. While the one-way coupled nature of the simulations precluded rigorous evaluation of the effects of seed material on detonability, a substantial temperature reduction is expected at the detonation wave front due to heating and decomposition. A preliminary comparison is presented between the RDE and an equivalent constant pressure adiabatic combustor, showing a potential performance advantage for the RDE.

42 ENGINEERING↗

DOE OSTI2026

Machine Learning Framework for Characterizing Processing–Structure Relationship in Block Copolymer Thin Films

The morphology of block copolymers (BCPs) critically influences material properties and applications. This work introduces a machine learning (ML)-enabled, high-throughput framework for analyzing grazing incidence small-angle X-ray scattering (GISAXS) data and atomic force microscopy (AFM) images to characterize BCP thin film morphology. A convolutional neural network was trained to classify AFM images by surface features, achieving 97% testing accuracy. Classified images were then analyzed to extract 2D grain size measurements from the samples in a high-throughput manner. ML models were trained to predict domain orientation based on processing parameters such as solvent ratio, additive type, and additive ratio. GISAXS-based properties were predicted with strong performances (R 2 > 0.75), while AFM-based property predictions were less accurate (R 2 < 0.60), likely due to the localized nature of AFM measurements compared to the bulk information captured by GISAXS. Beyond model performance, interpretability was addressed using SHapley Additive exPlanations (SHAP). SHAP analysis revealed that the additive ratio had the largest impact on morphological predictions, where additive provides the BCP chains with increased volume to rearrange into thermodynamically favorable morphologies. This interpretability helps validate model predictions and offers insight into parameter importance. Altogether, the presented framework combining high-throughput characterization and interpretable ML offers an approach to exploring and optimizing BCP thin film morphology across a broad processing landscape.

36 MATERIALS SCIENCE↗

DOE OSTI2020

Decreasing wind speed extrapolation error via domain-specific feature extraction and selection

Abstract. Model uncertainty is a significant challenge in the wind energy industry and can lead to mischaracterization of millions of dollars' worth of wind resources. Machine learning methods, notably deep artificial neural networks (ANNs), are capable of modeling turbulent and chaotic systems and offer a promising tool to produce high-accuracy wind speed forecasts and extrapolations. This paper uses data collected by profiling Doppler lidars over three field campaigns to investigate the efficacy of using ANNs for wind speed vertical extrapolation in a variety of terrains, and it quantifies the role of domain knowledge in ANN extrapolation accuracy. A series of 11 meteorological parameters (features) are used as ANN inputs, and the resulting output accuracy is compared with that of both standard log-law and power-law extrapolations. It is found that extracted nondimensional inputs, namely turbulence intensity, current wind speed, and previous wind speed, are the features that most reliably improve the ANN's accuracy, providing up to a 65 % and 52 % increase in extrapolation accuracy over log-law and power-law predictions, respectively. The volume of input data is also deemed important for achieving robust results. One test case is analyzed in depth using dimensional and nondimensional features, showing that the feature nondimensionalization drastically improves network accuracy and robustness for sparsely sampled atmospheric cases.

17 WIND ENERGY↗

DOE OSTI2022

GFINNs: GENERIC formalism informed neural networks for deterministic and stochastic dynamical systems

Here we propose the GENERIC formalism informed neural networks (GFINNs) that obey the symmetric degeneracy conditions of the GENERIC formalism. GFINNs comprise two modules, each of which contains two components. We model each component using a neural network whose architecture is designed to satisfy the required conditions. The component-wise architecture design provides flexible ways of leveraging available physics information into neural networks. We prove theoretically that GFINNs are sufficiently expressive to learn the underlying equations, hence establishing the universal approximation theorem. We demonstrate the performance of GFINNs in three simulation problems: gas containers exchanging heat and volume, thermoelastic double pendulum and the Langevin dynamics. In all the examples, GFINNs outperform existing methods, hence demonstrating good accuracy in predictions for both deterministic and stochastic systems.

97 MATHEMATICS AND COMPUTING↗