SEARCH · Engineering Papers

Results for “layered networks”

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 667 records · Page 37

All sources NASA NTRS DOE OSTI

DOE OSTI2024

Efficacy of using a dynamic length representation vs. a fixed-length for neuroarchitecture search

Deep learning neuroarchitecture and hyperparameter search are important in finding the best configuration that maximizes learned model accuracy. However, the number of types of layers, their associated hyperparameters, and the myriad of ways to connect layers poses a significant computational challenge in discovering ideal model configurations. Here, we assess two different approaches for neuroarchitecture search for a LeNet style neural network, one that uses a fixed-length approach where there is a preset number of possible layers that can be toggled on or off via mutation, and a variable-length approach where layers can be freely added or removed via special mutation operators. We found that the variable-length implementation trained better models while discovering unusual layer configurations worth further exploration.

Coletti, Mark↗

DOE OSTI2026

CoURAGE KAZR b1 Processing: Corrections, Calibrations, and Processing Report

The U.S. Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) User Facility supports atmospheric and earth system research through a comprehensive network of fixed and mobile observatories. These facilities provide long-term and intensive campaign-based observations of clouds, aerosols, precipitation, radiation, and meteorological state variables. ARM observations are designed to improve the physical understanding and numerical representation of atmospheric processes in earth system models, with particular emphasis on cloud-radiation interactions and precipitation processes. The Coast-Urban-Rural Atmospheric Gradient Experiment (CoURAGE) deploys one of the ARM Mobile Facilities (AMF) to the Mid-Atlantic region surrounding Baltimore, Maryland, for the period 1 December 2024 through 30 November 2025. This deployment focuses on characterizing atmospheric structure, cloud properties, and precipitation processes across strong land-use and surface heterogeneity gradients associated with urban, rural, and coastal (Chesapeake Bay) environments. The CoURAGE deployment complements the Baltimore Social-Environmental Collaborative (BSEC), a DOE Urban Integrated Field Laboratory (UIFL), by providing high-quality atmospheric observations needed to connect urban surface processes, emissions, and meteorology to cloud and precipitation responses. In addition to the central urban site, ancillary observing sites were deployed to rural Maryland northwest of Baltimore and to an island site in Chesapeake Bay. These measurements further complement a long-term atmospheric observatory operated in Beltsville, Maryland, by Howard University in collaboration with the Maryland Department of the Environment. Together, these assets form a four-node regional atmospheric observatory network representing Baltimore and its three primary surrounding environments—urban, rural, and coastal/bay. This coordinated observational strategy enables investigation of spatial gradients in boundary-layer structure, cloud occurrence, precipitation evolution, and aerosol-cloud interactions across complex surface regimes. Within this network, vertically pointing cloud radars play a critical role by providing continuous, high-resolution measurements of cloud and precipitation vertical structure.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2025

An adaptive and stability-promoting layerwise training approach for sparse deep neural network architecture

This work presents a two-stage adaptive framework for progressively developing deep neural network (DNN) architectures that generalize well for a given training data set. In the first stage, a layerwise training approach is adopted where a new layer is added each time and trained independently by freezing parameters in the previous layers. We impose desirable structures on the DNN by employing manifold regularization, sparsity regularization, and physics-informed terms. We introduce a ε – δ – stability-promoting concept as a desirable property for a learning algorithm and show that employing manifold regularization yields a ε – δ stability-promoting algorithm. Further, we also derive the necessary conditions for the trainability of a newly added layer and investigate the training saturation problem. In the second stage of the algorithm (post-processing), a sequence of shallow networks is employed to extract information from the residual produced in the first stage, thereby improving the prediction accuracy. Numerical investigations on prototype regression and classification problems demonstrate that the proposed approach can outperform fully connected DNNs of the same size. Moreover, by equipping the physics-informed neural network (PINN) with the proposed adaptive architecture strategy to solve partial differential equations, we numerically show that adaptive PINNs not only are superior to standard PINNs but also produce interpretable hidden layers with provable stability. As a result, we also apply our architecture design strategy to solve inverse problems governed by elliptic partial differential equations.

42 ENGINEERING↗

DOE OSTI2020

Microscopy Studies of the Catalyst Inks for PEM Fuel Cells

The microstructure of the catalyst layer in proton exchange membrane fuel cells (PEMFC) is one of the key factors that determine fuel cell performance. The ink preparation, including dispersion solvent, mixing time, and mixing power, are among the less reported parameters that would influence the ionomer distribution on the catalyst surface and the continuity of the carbon network. In this work, novel characterization approaches have been used to investigate the effect of the Pt/C catalyst ink preparation conditions on the morphology and structures of the catalyst layer. Microscopy analysis revealed that longer mixing time led to better performing electrode structures than shorter mixing time. The longer mixing time enabled improved continuity of the ionomer network and high porosity in the cathode layer that contributes to improved proton conductivity and mass transport. This has been reflected in performance and durability tests, where the electrode made from a 5-day catalyst ink displayed improved performance compared to the one made from a 3-day ink. Durability studies showed 26% and 43% loss of the initial mass activity for 5 days and 3 days mixing, respectively. Finally, in addition, the catalyst layer prepared with ethylene glycol as the dispersion solvent showed better durability than water/1-propanol based solution.

25 ENERGY STORAGE↗

DOE OSTI2023

A deep learning upscaling framework: Reactive transport and mineral precipitation in fracture-matrix systems

Pore-scale modeling has limited applicability at large scales due to its high computational cost. One common approach to upscale pore-scale models is the use of effective medium theories, which homogenize small-scale features in a porous structure and characterize the medium by macroscale properties (e.g., permeability) and equations (e.g., Darcy’s law). However, there are classes of physical processes for which effective medium approximations may become inaccurate, e.g., mineral precipitation and clogging during reactive transport. We have developed a deep learning upscaling framework, in which pore-scale modeling is directly employed in macroscale systems, without relying on effective medium approximations. The upscaling framework is first developed for general multiscale systems and then applied to modeling reactive transport with mineral precipitation in the altered layer in fracture-matrix structures. Solute transport from the fractures to the matrix is modeled as a wall boundary condition for the fractures, which, in turn, is predicted by recurrent neural networks using the concentration histories at the fracture-matrix boundary. Specifically, we consider a meter-scale fracture network embedded in sandstones, where the smallest feature is at the micron scale. Here the proposed framework allows us to span five orders of magnitude in length scales by capturing mineral precipitation in the altered layer of the rock matrix at the pore scale across the entire meter-scale fracture network.

42 ENGINEERING↗

NASA NTRS2000

Experimental Verification of Electric Drive Technologies Based on Artificial Intelligence Tools

In this report, a fully integrated prototype of a flight servo control system is successfully developed and implemented using brushless dc motors. The control system is developed by the fuzzy logic theory, and implemented with a multilayer neural network. First, a neural network-based architecture is introduced for fuzzy logic control. The characteristic rules and their membership functions of fuzzy systems are represented as the processing nodes in the neural network structure. The network structure and the parameter learning are performed simultaneously and online in the fuzzy-neural network system. The structure learning is based on the partition of input space. The parameter learning is based on the supervised gradient decent method, using a delta adaptation law. Using experimental setup, the performance of the proposed control system is evaluated under various operating conditions. Test results are presented and discussed in the report. The proposed learning control system has several advantages, namely, simple structure and learning capability, robustness and high tracking performance and few nodes at hidden layers. In comparison with the PI controller, the proposed fuzzy-neural network system can yield a better dynamic performance with shorter settling time, and without overshoot. Experimental results have shown that the proposed control system is adaptive and robust in responding to a wide range of operating conditions. In summary, the goal of this study is to design and implement-advanced servosystems to actuate control surfaces for flight vehicles, namely, aircraft and helicopters, missiles and interceptors, and mini- and micro-air vehicles.

Rubaai, Ahmed↗

NASA NTRS1981

Dynamic study of the upper Sao Francisco River and the Tres Marias reservoir using MSS/LANDSAT images

The use of LANDSAT multispectral ban scanner imagery to verify the relationship between the behavior of the Tres Marias reservoir and the dynamics of the Sao Francisco River supply basin is described. The dispersion of suspended sediments and their concentration in the surface layers of the water are considered. A five year survey of the region during both dry and rainy seasons was performed. The drainage network was analyzed based on the patterns of dessication, water rises and soil use in the supply basin. Surface layers of the reservoir were tabulated as a function of the levels of gray in the imagery. In situ observations of water depth and reflectance were performed. Ground truth and LANDSAT data were correlated to determine the factors affecting the dynamics of the supply basin.

Dejesusparada, N.↗

NASA NTRS1983

Networking DEC and IBM computers

Local Area Networking of DEC and IBM computers within the structure of the ISO-OSI Seven Layer Reference Model at a raw signaling speed of 1 Mops or greater are discussed. After an introduction to the ISO-OSI Reference Model nd the IEEE-802 Draft Standard for Local Area Networks (LANs), there follows a detailed discussion and comparison of the products available from a variety of manufactures to perform this networking task. A summary of these products is presented in a table.

Mish, W. H.↗

NASA NTRS1993

Application of neural network to remote sensing of soil moisture using theoretical polarimetric backscattering coefficients

This paper investigates the potential application of neural network to inversion of soil moisture using polarimetric remote sensing data. The neural network used for the inversion of soil parameters is multi-layer perceptron trained with the back-propagation algorithm. The training data include the polarimetric backscattering coefficients obtained from theoretical surface scattering models together with an assumed nominal range of soil parameters which are comprised of the soil permittivity and surface roughness parameters. Soil permittivity is calculated from the soil moisture and the assumed soil texture based on an empirical formula at C-, L-, and P-bands. The rough surface parameters for the soil surface, which is described by the Gaussian random process, are the root-mean-square (rms) height and correlation length. For the rough surface scattering, small perturbation method is used for the L-band frequency, and Kirchhoff approximation is used for the C-band frequency to obtain the corresponding backscattering coefficients. During the training, the backscattering coefficients are the inputs to the neural net and the output from the net are compared with the desired soil parameters to adjust the interconnecting weights. The process is repeated for each input-output data entry and then for the entire training data until convergence is reached. After training, the backscattering coefficients are applied to the trained neural net to retrieve the soil parameters which are compared with the desired soil parameters to verify the effectiveness of this technique. Several cases are examined. First, for simplicity, the correlation length and rms height of the soil surface are fixed while soil moisture is varied. Soil moisture obtained using the neural networks with either L-band or C-band backscattering coefficients for the HH and VV polarizations as inputs is in good agreement with the desired soil moisture. The neural net output matches the desired output for the soil moisture range of 16 to 60 percent for the C-band case. The next case investigated is to vary both soil moisture and rms height while keeping the correlation length fixed. For this case, C-band backscattering coefficients are not sufficient for retrieving two parameters because the Kirchhoff approximation gives the same HH and VV backscattering coefficients. Therefore, the backscattering coefficients at two different frequency bands are necessary to find both the soil moisture and rms height. Finally, the neural nets are also applied to simultaneously invert soil moisture, rms height, and correlation length. Overall, the soil moisture retrieved from the neural network agrees very well with the desired soil moisture. This suggests that the neural network shows potential for retrieval of soil parameters from remote sensing data.

Wang, L.↗

DOE OSTI2026

Make the Fastest Faster: Importance Mask Synthesis for Interactive Volume Visualization using Reconstruction Neural Networks

Visualizing a large-scale volumetric dataset with high resolution is challenging due to the substantial computational time and space complexity. Recent deep learning-based image inpainting methods significantly improve rendering latency by reconstructing a high-resolution image for visualization in constant time on GPU from a partially rendered image where only a portion of pixels go through the expensive rendering pipeline. However, existing solutions need to render every pixel of either a predefined regular sampling pattern or an irregular sample pattern predicted from a low-resolution image rendering. Both methods require a significant amount of expensive pixel-level rendering. In this work, we provide Importance Mask Learning (IML) and Synthesis (IMS) networks, which are the first attempts to directly synthesize important regions of the regular sampling pattern from the user’s view parameters, to further minimize the number of pixels to render by jointly considering the dataset, user behavior, and the downstream reconstruction neural network. Our solution is a unified framework to handle various types of inpainting methods through the proposed differentiable compaction/decompaction layers. Experiments show our method can further improve the overall rendering latency of state-of-the-art volume visualization methods using reconstruction neural network for free when rendering scientific volumetric datasets. Our method can also directly optimize the off-the-shelf pre-trained reconstruction neural networks without elongated retraining.

Large-scale data↗

DOE OSTI2020

The Brachypodium ENCODE Project – from sequence to function: Predicting physiological responses in grasses to facilitate engineering of biofuel crops

The overarching goal of this proposal was to gain deep knowledge of chromatin dynamics and gene networks related to stress responses in grasses. To accomplish this goal our Brachypodium ENCODE project (1) coupled drought-focused physiological analysis with multiple layers of ENCODE ‘omics’ profiling; (2) obtained fundamental information about the Brachypodium transcriptome, epigenome, chromatin dynamics, and gene regulatory networks; and (3) elucidated the regulatory dynamics of gene networks that can be manipulated to increase energy grass crop productivity.

09 BIOMASS FUELS↗

DOE OSTI2022

Machine Learning-Assisted High-Temperature Reservoir Thermal Energy Storage Optimization: Numerical Modeling and Machine Learning Input and Output Files

This data set includes the numerical modeling input files and output files used to synthesize data, and the reduced-order machine learning models trained from the synthesized data for reservoir thermal energy storage site identification. In this study, a machine-learning-assisted computational framework is presented to identify High-Temperature Reservoir Thermal Energy Storage (HT-RTES) site with optimal performance metrics by combining physics-based simulation with stochastic hydrogeologic formation and thermal energy storage operation parameters, artificial neural network regression of the simulation data, and genetic algorithm-enabled multi-objective optimization. A doublet well configuration with a layered (aquitard-aquifer-aquitard) generic reservoir is simulated for cases of continuous operation and seasonal-cycle operation scenarios. Neural network-based surrogate models are developed for the two scenarios and applied to generate the Pareto fronts of the HT-RTES performance for four potential HT-RTES sites. The developed Pareto optimal solutions indicate the performance of HT-RTES is operation-scenario (i.e., fluid cycle) and reservoir-site dependent, and the performance metrics have competing effects for a given site and a given fluid cycle. The developed neural network models can be applied to identify suitable sites for HT-RTES, and the proposed framework sheds light on the design of resilient HT-RTES systems. All the simulations and the neural network model were done by Idaho National Laboratory. A detailed description of the work was reported in publication linked below.

15 GEOTHERMAL ENERGY↗

NASA NTRS1997

Neural Network Prediction of Failure of Damaged Composite Pressure Vessels from Strain Field Data Acquired by a Computer Vision Method

This effort used a new and novel method of acquiring strains called Sub-pixel Digital Video Image Correlation (SDVIC) on impact damaged Kevlar/epoxy filament wound pressure vessels during a proof test. To predict the burst pressure, the hoop strain field distribution around the impact location from three vessels was used to train a neural network. The network was then tested on additional pressure vessels. Several variations on the network were tried. The best results were obtained using a single hidden layer. SDVIC is a fill-field non-contact computer vision technique which provides in-plane deformation and strain data over a load differential. This method was used to determine hoop and axial displacements, hoop and axial linear strains, the in-plane shear strains and rotations in the regions surrounding impact sites in filament wound pressure vessels (FWPV) during proof loading by internal pressurization. The relationship between these deformation measurement values and the remaining life of the pressure vessels, however, requires a complex theoretical model or numerical simulation. Both of these techniques are time consuming and complicated. Previous results using neural network methods had been successful in predicting the burst pressure for graphite/epoxy pressure vessels based upon acoustic emission (AE) measurements in similar tests. The neural network associates the character of the AE amplitude distribution, which depends upon the extent of impact damage, with the burst pressure. Similarly, higher amounts of impact damage are theorized to cause a higher amount of strain concentration in the damage effected zone at a given pressure and result in lower burst pressures. This relationship suggests that a neural network might be able to find an empirical relationship between the SDVIC strain field data and the burst pressure, analogous to the AE method, with greater speed and simplicity than theoretical or finite element modeling. The process of testing SDVIC neural network analysis and some encouraging preliminary results are presented in this paper. Details are given concerning the processing of SDVIC output data such that it may be used as back propagation neural network (BPNN) input data. The software written to perform this processing and the BPNN algorithm are also discussed. It will be shown that, with limited training, test results indicate an average error in burst pressure prediction of approximately six percent,

Russell, Samuel S.↗

DOE OSTI2021

Intrusion Detection for Cybersecurity of Power Electronics Dominated Grids: Inverters PQ Set-Points Manipulation

This work provides cybersecurity analytics for a high photovoltaic (PV) penetrated distribution network, representing future power electronics dominated grid (PEDG). Firstly, the impact of active and reactive power (PQ) set-points manipulation on the network is studied. Then, an intrusion detection system (IDS) is developed for identifying the potentially compromised PV inverters in the network. The proposed IDS is based on defining the normal, safe, and abnormal operation regions of the PV inverters from point of view of the steady state voltage stability of the network. These three-operation regions are identified by utilizing active power, reactive power, and voltage (PQV) limits of each specific grid-following inverter in the network. Each grid-following inverter's PQV contour includes the information of network topology, inverter ratings, and inverter controller. The developed PQ limits are integrated into the centralized secondary control layer for realization of the proposed IDS. Furthermore, the proposed secondary control layer is capable of providing remedial actions during an anomaly event to enhance the grid resiliency. The theoretical analyses are verified by several attack scenarios for a network of grid-following inverters.

41 EE - Solar Energy Technologies Office (EE-4S)↗

DOE OSTI2022

Machine-learning-assisted high-temperature reservoir thermal energy storage optimization

High-temperature reservoir thermal energy storage (HT-RTES) has the potential to become an indispensable component in achieving the goal of the net-zero carbon economy, given its capability to balance the intermittent nature of renewable energy generation. In this study, a machine-learning-assisted computational framework is presented to co-optimize the performance metrics of HT-RTES by combining physics-based simulation with stochastic hydrogeologic formation and thermal energy storage operation parameters, artificial neural network regression of the simulation data, and genetic algorithm-enabled multi-objective optimization. A doublet well configuration with a layered (aquitard-aquifer-aquitard) generic reservoir is simulated for cases of continuous operation and seasonal-cycle operation scenarios. Further, neural network-based surrogate models are developed for the two scenarios and applied to generate the Pareto fronts of the HT-RTES performance for four potential HT-RTES sites. The developed Pareto optimal solutions indicate the performance of HT-RTES is operation-scenario (i.e., fluid cycle) and reservoir-site dependent, and the performance metrics have competing effects for a given site and a given fluid cycle. The developed neural network models can be applied to identify suitable sites for HT-RTES, and the proposed framework sheds light on the design of resilient HT-RTES systems.

15 GEOTHERMAL ENERGY↗

NASA NTRS1992

Characterization of low thermal conductivity PAN-based carbon fibers

The microstructure and surface chemistry of eight low thermal conductivity (LTC) PAN-based carbon fibers were determined and compared with PAN-based fibers heat treated to higher temperatures. Based on wide-angle x ray diffraction, the LTC PAN fibers all appear to have a similar turbostratic structure with large 002 d-spacings, small crystallite sizes, and moderate preferred orientation. Limited small-angle x ray scattering (SAXS) results indicate that, with the exception of LTC fibers made by BASF, the LTC fibers do not have well developed pores. Transmission electron microscopy shows that the texture of the two LTC PAN-based fibers studied (Amoco T350/23X and /25X) consists of multiple sets of parallel, wavy, bent layers that interweave with each other forming a complex three dimensional network oriented randomly around the fiber axis. X ray photoelectron spectroscopy (XPS) analysis finds correlations between heat treated temperatures and the surface composition chemistry of the carbon fiber samples.

Katzman, Howard A.↗

DOE OSTI2021

Rotational Equivariance for Object Classification Using xView

With the recent addition of large, curated and labeled data sets to the remote sensing discipline, deep learning models have largely surpassed the performance of classical techniques. These deep models, typically Convolutional Neural Networks, are invariant to translation through the use of successive convolution layers which are themselves equivariant to translation. Further, the combination of multiple convolution and pooling layers means that in practice, the model is also approximately invariant to translation. However, until recently these models could only approach rotational invariance through data augmentation. Here we propose using a new model formulation which achieves rotational equaivariance without data augmentation for overhead imagery classification. We utilize the popular xView data set to compare the rotational equivariance formalization against a regular CNN and CNN with rotational data augmentation for the task of image classification.

Bynum, Lucius EJ↗

DOE OSTI2023

Variations in boundary layer stability across Antarctica: a comparison between coastal and interior sites

The range of boundary layer stability profiles, from the surface to 500 m a.g.l. (above ground level), present in radiosonde observations from two continental-interior (South Pole Station and Dome Concordia Station) and three coastal (McMurdo Station, Georg von Neumayer Station III, and Syowa Station) Antarctic sites, is examined using the self-organizing maps (SOMs) neural network algorithm. A wide range of potential temperature profiles is revealed, from shallow boundary layers with strong near-surface stability to deeper boundary layers with weaker or near-neutral stability, as well as profiles with weaker near-surface stability and enhanced stability aloft, above the boundary layer. Boundary layer regimes were defined based on the range of profiles revealed by the SOM analysis; 20 boundary layer regimes were identified to account for differences in stability near the surface as well as above the boundary layer. Strong, very strong, or extremely strong stability, with vertical potential temperature gradients of 5 to in excess of 30 K per 100 m, occurred more than 80 % of the time at South Pole and Dome Concordia in the winter. Weaker stability was found in the winter at the coastal sites, with moderate and strong stability (vertical potential temperature gradients of 1.75 to 15 K per 100 m) occurring 70 % to 85 % of the time. Even in the summer, moderate and strong stability is found across all five sites, either immediately near the surface or aloft, just above the boundary layer. While the mean boundary layer height at the continental-interior sites was found to be approximately 50 m, the mean boundary layer height at the coastal sites was deeper, around 110 m. Further, a commonly described two-stability-regime system in the Arctic associated with clear or cloudy conditions was applied to the 20 boundary layer regimes identified in this study to understand if the two-regime behavior is also observed in the Antarctic. It was found that moderate and strong stability occur more often with clear- than cloudy-sky conditions, but weaker stability regimes occur almost equally for clear and cloudy conditions.

54 ENVIRONMENTAL SCIENCES↗