Search NASA⌕ Search

SEARCH · Search NASA

Results for “Distributed training”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

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

At least 739 records · Page 41

Genetic Control of Flowering in Switchgrass (Final Technical Report)

Lignocellulosic biomass is poised to become a significant source of renewable fuel in the United States, and perennial grasses such as switchgrass are likely to provide a significant amount of this biomass. A decline in biomass yield is typically observed in switchgrass immediately following completion of flowering. In the northern part of the country, flowering occurs in switchgrass around the time that biomass yield peaks, and the approximately two-month growing season for biomass accumulation can be lost due to a flowering event. If flowering time can be delayed about 4-5 weeks, biomass yield can potentially increase 30-50%. To develop a rational strategy for creating improved switchgrass with late flowering, it would be beneficial to have a better understanding of the genes that control flowering time in switchgrass. The proposed research aims to dissect the genetic control of the flowering time of switchgrass. Through genetic, genomic, and bioinformatics approaches, key genes involved in the flowering pathways such as photoperiod and circadian control have been identified. The obtained research information will help further elucidate the molecular mechanisms underlying anthesis phenotypes of switchgrass. The genetic variants related to floral development identified through the project can support a long-term goal of developing late flowering varieties of switchgrass that fully utilize the growing season and achieve high biomass yield. The resources generated from the project have been timely distributed to the community via professional meetings and publications. Graduate students and postdoctoral researchers have been trained through working on the project.

09 BIOMASS FUELS↗

Probabilistic Nanomagnetic Memories for Uncertain and Robust Machine Learning

This project evaluated the use of emerging spintronic memory devices for robust and efficient variational inference schemes. Variational inference (VI) schemes, which constrain the distribution for each weight to be a Gaussian distribution with a mean and standard deviation, are a tractable method for calculating posterior distributions of weights in a Bayesian neural network such that this neural network can also be trained using the powerful backpropagation algorithm. Our project focuses on domain-wall magnetic tunnel junctions (DW-MTJs), a powerful multi-functional spintronic synapse design that can achieve low power switching while also opening the pathway towards repeatable, analog operation using fabricated notches. Our initial efforts to employ DW-MTJs as an all-in-one stochastic synapse with both a mean and standard deviation didn’t end up meeting the quality metrics for hardware-friendly VI. In the future, new device stacks and methods for expressive anisotropy modification may make this idea still possible. However, as a fall back that immediately satisfies our requirements, we invented and detailed how the combination of a DW-MTJ synapse encoding the mean and a probabilistic Bayes-MTJ device, programmed via a ferroelectric or ionically modifiable layer, can robustly and expressively implement VI. This design includes a physics-informed small circuit model, that was scaled up to perform and demonstrate rigorous uncertainty quantification applications, up to and including small convolutional networks on a grayscale image classification task, and larger (Residual) networks implementing multi-channel image classification. Lastly, as these results and ideas all depend upon the idea of an inference application where weights (spintronic memory states) remain non-volatile, the retention of these synapses for the notched case was further interrogated. These investigations revealed and emphasized the importance of both notch geometry and anisotropy modification in order to further enhance the endurance of written spintronic states. In the near future, these results will be mapped to effective predictions for room temperature and elevated operation DW-MTJ memory retention, and experimentally verified when devices become available.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

MAPPING THE GALAXY COLOR–REDSHIFT RELATION: OPTIMAL PHOTOMETRIC REDSHIFT CALIBRATION STRATEGIES FOR COSMOLOGY SURVEYS

Calibrating the photometric redshifts of ≳10{sup 9} galaxies for upcoming weak lensing cosmology experiments is a major challenge for the astrophysics community. The path to obtaining the required spectroscopic redshifts for training and calibration is daunting, given the anticipated depths of the surveys and the difficulty in obtaining secure redshifts for some faint galaxy populations. Here we present an analysis of the problem based on the self-organizing map, a method of mapping the distribution of data in a high-dimensional space and projecting it onto a lower-dimensional representation. We apply this method to existing photometric data from the COSMOS survey selected to approximate the anticipated Euclid weak lensing sample, enabling us to robustly map the empirical distribution of galaxies in the multidimensional color space defined by the expected Euclid filters. Mapping this multicolor distribution lets us determine where—in galaxy color space—redshifts from current spectroscopic surveys exist and where they are systematically missing. Crucially, the method lets us determine whether a spectroscopic training sample is representative of the full photometric space occupied by the galaxies in a survey. We explore optimal sampling techniques and estimate the additional spectroscopy needed to map out the color–redshift relation, finding that sampling the galaxy distribution in color space in a systematic way can efficiently meet the calibration requirements. While the analysis presented here focuses on the Euclid survey, similar analysis can be applied to other surveys facing the same calibration challenge, such as DES, LSST, and WFIRST.

79 ASTRONOMY AND ASTROPHYSICS↗

Electron Beam Pattern Rotation as a Method of Tunable Bunch Train Generation

Transversely modulated electron beams can be formed in photo injectors via microlens array (MLA) UV laser shap- ing technique. Microlenses can be arranged in polygonal lattices, with resulting transverse electron beam modula- tion mimicking the lenses pattern. Conventionally, square MLAs are used for UV laser beam shaping, and generated electron beam patterns form square beamlet arrays. The MLA setup can be placed on a rotational mount, thereby rotating electron beam distribution. In combination with transverse-to-longitudinal emittance exchange (EEX) beam line, it allows to vary beamlets horizontal projection and tune electron bunch train. In this paper, we extend the technique to the case of different MLA lattice arrangements and explore the benefits of its rotational symmetries.

43 PARTICLE ACCELERATORS↗

Adjustment of ablation shapes and subwavelength ripples based on electron dynamics control by designing femtosecond laser pulse trains

A quantum model is proposed to investigate femtosecond laser pulse trains processing of dielectrics by including the plasma model with the consideration of laser particle-wave duality. Central wavelengths (400 nm and 800 nm) strongly impact the surface plasmon field distribution, the coupling field intensity distribution (between the absorbed intensity and the surface plasma), and the distribution of transient localized free electron density in the material. This, in turn, significantly changes the localized transient optical/thermal properties during laser materials processing. The effects of central wavelengths on ablation shapes and subwavelength ripples are discussed. The simulation results show that: (1) ablation shapes and the spacing of subwavelength ripples can be adjusted by localized transient electron dynamics control using femtosecond laser pulse trains; (2) the adjustment of the radii of ablation shapes is stronger than that of the periods of subwavelength ripples.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Playas Grid Reliability and Distributed Energy Research

The future looks bright for solar and renewable energies in the United States. Recent studies claim that by 2050, solar power could supply a third of all electricity demand in the country’s western states. Technology advances, soft policy changes, and increased energy consciousness will all have to happen to achieve this goal. The New Mexico Institute of Mining and Technology (NMT) and the Department of Homeland Security (DHS) perform various training and research activities at this site. Given its unique nature, Playas was chosen to test Micro-Grids and other examples of renewable distributed energy resources (DER). Several proposed distributed energy sources (DERs) were not implemented as planned including the Micro-Grid. However, Micro-Grid design and computer modeling were completed and these results are included in this report. As part of this research, four PV (solar) generating systems were installed with remote Internet based communication and control capabilities. These systems have been integrated into and can interact with the local grid So that (for example) excess power produced by the solar arrays can be exported to the utility grid. Energy efficient LED lighting was installed in several buildings to further reduce consumption of utility-supplied power. By combining reduced lighting costs; lowering HVAC loads; and installing smart PV generating equipment with energy storage (battery banks) these systems can greatly reduce electrical usage drawn from an older rural electrical cooperative (Co-Op) while providing clean dependable power. Several additional tasks under this project involved conducting research to develop methods of producing electricity from organic materials (i.e. biofuels, biomass. etc.), the most successful being the biodiesel reactor. Improvements with Proton Exchange Membranes (PEM) for fuels cells were demonstrated and advances in Dye Sensitized Solar Cells (DSSC) were also shown. The specific goals of the project include: 1) Instrumentation of the power distribution system with distributed energy resources, demand-side control and intelligent homes within the town of Playas, NM, 2) Creation of models (power flow and dynamic) of the Playas power distribution system, 3) Validation of the models through comparison of predicted behavior to data collected from instrumentation, and 4) Utilization of the models and test grid to characterize the impact of new devices and approaches (e.g., distributed generation and load management) on the local distribution system as well as the grid at large. In addition to the above stated objectives, the research also focused on three critical challenges facing renewable distributed energy platforms: 1) hydrogen from biomass, 2) improved catalyst support systems for electrolysis membranes and fuel cell systems, and 3) improved manufacturing methodologies of low cost photovoltaics. The following sections describe activities performed during this project. The various tasks were focused on establishing Playas as a “…theoretical and experimental test bed…” through which components of a modern/smart grid could be characterized. On a broader scale, project efforts were aimed at development of tools and gathering of experience/expertise that would accelerate progress toward implementation of a modern grid.

09 BIOMASS FUELS↗

Toward a machine-guided approach to energetic material discovery

In this article, we trained a machine learning (ML) model to connect microstructural details of an energetic material formulation to its performance for the purpose of guiding the discovery of new explosive formulations. Our hypothesis was that the algorithm would robustly learn the training data and produce an accurate surrogate model. Specifically, the algorithm learned the relationship between details of the void size distribution (VSD), initiating shock pressure, and the energetic material performance. We used realistic constraints on the VSD and a range of cases were ingested by a physically informed reactive flow model working within a hydrodynamic solver running on high-performance computing resources. The ML algorithm produced a surrogate model that accurately predicted known test points around the parameter space. In addition to the utility of the model and the process used for its development, we noted interesting comparisons between what we, the authors—subject matter experts, would heuristically conclude from the training data and the surrogate model predictions. We detected nuanced details that were missed by the surrogate model; however, these details are not important to an energetic material formulator. We concluded that the algorithm did indeed robustly learn the training data and produce an accurate surrogate model. We further concluded that the surrogate model is a powerful tool to guide the formulator in the absence of subject matter experts and limited-access computing resources.

42 ENGINEERING↗

A meta-learning based distribution system load forecasting model selection framework

This paper presents a meta-learning based, automatic distribution system load forecasting model selection framework. Furthermore, the framework includes the following processes: feature extraction, candidate model preparation and labeling, offline training, and online model recommendation. Using load forecasting needs and data characteristics as input features, multiple metalearners are used to rank the candidate load forecast models based on their forecasting accuracy. Then, a scoring-voting mechanism is proposed to weights recommendations from each meta-leaner and make the final recommendations. Heterogeneous load forecasting tasks with different temporal and technical requirements at different load aggregation levels are set up to train, validate, and test the performance of the proposed framework. Simulation results demonstrate that the performance of the meta-learning based approach is satisfactory in both seen and unseen forecasting tasks.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Detailed Fission Power 2D-Mapping of AFIP-2 Experiment in ATR CFT Position

The NNSA (National Nuclear Security Administration) RERTR (Reduced Enrichment for Research and Test Reactors) program assigned INL (Idaho National Laboratory) the responsibility of developing and demonstrating high uranium density research reactor fuel forms to enable the use of low enriched uranium (LEU) in the research reactors around the world. A series of full-size fuel plate irradiation tests are proposed for the ATR (Advanced Test Reactor). Labeled the AFIP (ATR Full-size-plates In center flux trap Position) experiments, these tests will be conducted in the ATR center flux trap. The AFIP-2 experiment will contain two full size fuel plates fabricated at the Oak Ridge National Laboratory-Y-12 (ORNL). The nominal fuel zone is rectangular in shape having a designed length of 21.5-in (54.6l-cm), width of 1.6-in (4.064-cm), and uniform thickness of 0.014-in (0.03556-cm). This gives a nominal fuel zone volume of 0.482 in3 (7.89 cm3) in each fuel plate. The test holder accommodates two independent test trains. Each test train is designed to hold 2 plates, for a total of 4 plates per test holder. AFIP-2 test plates will be irradiated at a peak surface heat flux of about 350 W/cm2 and discharged at a peak U-235 burn-up of about 70 at.%. Based on limited irradiation testing of the monolithic (U10Mo) fuel form, it is desirable to keep the peak fuel temperature below 250°C; to achieve this, it will be necessary to keep plate heat fluxes below 500 W/cm2. Due to the heavy U-235 loading and width of 1.6-in (4.064 cm), the neutron self-shielding will increase the local-to-average-ratio fission power near the sides of the fuel plates. To assure the AFIP-2 experiment will comply with the ATR safety requirements, a very detailed 2 dimensional (2D) Y-Z fission power profile was evaluated to best predict the fuel plate temperature distribution. The ability to accurately predict fuel plate power and burnup are essential in the AFIP-2 fuel test train design and the irradiated fuel performance evaluation. We obtained the required power and heat generation rates within test train for the thermal analyses. A detailed MCNP Y-Z mini-plate fuel model was developed. The Y-Z model divides each fuel mini-plate into 30 equal intervals in Y and Z directions. The MCNP-calculated results and the detailed Y-Z fission power mapping were used to help design the AFIP-2 fuel test assembly to ensure that the capsule thermal-hydraulic limits will not exceed the ATR safety limit.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Statistical Modeling Approach to Quantitative Analysis of Interobserver Variability in Breast Contouring

Purpose: To develop a new approach for interobserver variability analysis. Methods and Materials: Eight radiation oncologists specializing in breast cancer radiation therapy delineated a patient's left breast “from scratch” and from a template that was generated using deformable image registration. Three of the radiation oncologists had previously received training in Radiation Therapy Oncology Group consensus contouring for breast cancer atlas. The simultaneous truth and performance level estimation algorithm was applied to the 8 contours delineated “from scratch” to produce a group consensus contour. Individual Jaccard scores were fitted to a beta distribution model. We also applied this analysis to 2 or more patients, which were contoured by 9 breast radiation oncologists from 8 institutions. Results: The beta distribution model had a mean of 86.2%, standard deviation (SD) of ±5.9%, a skewness of −0.7, and excess kurtosis of 0.55, exemplifying broad interobserver variability. The 3 RTOG-trained physicians had higher agreement scores than average, indicating that their contours were close to the group consensus contour. One physician had high sensitivity but lower specificity than the others, which implies that this physician tended to contour a structure larger than those of the others. Two other physicians had low sensitivity but specificity similar to the others, which implies that they tended to contour a structure smaller than the others. With this information, they could adjust their contouring practice to be more consistent with others if desired. When contouring from the template, the beta distribution model had a mean of 92.3%, SD ± 3.4%, skewness of −0.79, and excess kurtosis of 0.83, which indicated a much better consistency among individual contours. Similar results were obtained for the analysis of 2 additional patients. Conclusions: The proposed statistical approach was able to measure interobserver variability quantitatively and to identify individuals who tended to contour differently from the others. The information could be useful as feedback to improve contouring consistency.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Convective Transport and Evolution of Clouds and aerosols in the Asian Summer Monsoon Region

The existence of a persistent layer in the lowermost stratosphere in the Asian Monsoon region is well-known and has been observed by satellites and also in situ by balloon-borne sensors. This aerosol layer has been observed to persist every year by both SAGE sensors and by CALIOP, between the tropopause and the lower boundary of the Junge stratospheric sulfate layer. But how does this layer form and why? Apparently, aerosols are lofted by convection form the Indian and Asian continents, into the upper troposphere/lower stratosphere (UT/LS) area in the middle of a persistent anticyclone associate with the Asian summer monsoon. But some mysteries remain. The top of this aerosol layer is well above the altitude of maximum convective outflow, so why is the top of the aerosol layer so high? Are the aerosols coming from China or from India? And finally, is there an observable impact of these aerosol on thin cirrus distributions in the UT/LS? We explore the interesting difference in altitude between the ATAL layer and the convective cloud top height, using observations from CALIOP and from SAGE. In the process of doing this we examine the accuracy of both CALIOP and SAGE cloud/aerosol discrimination, and the impact of various averaging lengths. We compare the apparent "observational tropopause" with observations from COSMIC, CALIOP (clouds), and with modeled interpolation from MERRA-2. We also compare CALIOP and CloudSat daytime cloud top heights for convective clouds with the vertical aerosol distribution, and evaluate convection in this region as represented by MERRA-2. Finally, and maybe most importantly we compare convective cloud top heights and temperatures with diurnal variability during the most recent years by using observations from the geostationary Himawari satellite. How representative are the twice-daily A-Train observations, when compared to the natural diurnal variability of convection? The goal is to establish the origins of these UT/LS aerosols, whether these aerosol distributions can be completely explained by convective transport, and to evaluate convective lofting of aerosol by models. Ultimately, the convective lofting of aerosols as modeled is needed to explain the Asian Monsoon aerosol distributions. The observations contribute much-needed information to enhance the model accuracy.

clouds↗

Enhancing synchrotron radiation micro-CT images using deep learning: an application of Noise2Inverse on bone imaging

In bone-imaging research, in situ synchrotron radiation micro-computed tomography (SRµCT) mechanical tests are used to investigate the mechanical properties of bone in relation to its microstructure. Low-dose computed tomography (CT) is used to preserve bone's mechanical properties from radiation damage, though it increases noise. To reduce this noise, the self-supervised deep learning method Noise2Inverse was used on low-dose SRµCT images where segmentation using traditional thresholding techniques was not possible. Simulated-dose datasets were created by sampling projection data at full, one-half, one-third, one-fourth and one-sixth frequencies of an in situ SRµCT mechanical test. After convolutional neural networks were trained, Noise2Inverse performance on all dose simulations was assessed visually and by analyzing bone microstructural features. Visually, high image quality was recovered for each simulated dose. Lacunae volume, lacunae aspect ratio and mineralization distributions shifted slightly in full, one-half and one-third dose network results, but were distorted in one-fourth and one-sixth dose network results. Following this, new models were trained using a larger dataset to determine differences between full dose and one-third dose simulations. Significant changes were found for all parameters of bone microstructure, indicating that a separate validation scan may be necessary to apply this technique for microstructure quantification. Noise present during data acquisition from the testing setup was determined to be the primary source of concern for Noise2Inverse viability. While these limitations exist, incorporating dose calculations and optimal imaging parameters enables self-supervised deep learning methods such as Noise2Inverse to be integrated into existing experiments to decrease radiation dose.

Obata, Yoshihiro (ORCID:0000000303659129)↗

Naturalistic Decision Making For Power System Operators

Abstract: Motivation -- As indicated by the Blackout of 2003, the North American interconnected electric system is vulnerable to cascading outages and widespread blackouts. Investigations of large scale outages often attribute the causes to the three T’s: Trees, Training and Tools. A systematic approach has been developed to document and understand the mental processes that an expert power system operator uses when making critical decisions. The approach has been developed and refined as part of a capability demonstration of a high-fidelity real-time power system simulator under normal and emergency conditions. To examine naturalistic decision making (NDM) processes, transcripts of operator-to-operator conversations are analyzed to reveal and assess NDM-based performance criteria. Findings/Design -- The results of the study indicate that we can map the Situation Awareness Level of the operators at each point in the scenario. We can also identify clearly what mental models and mental simulations are being performed at different points in the scenario. As a result of this research we expect that we can identify improved training methods and improved analytical and visualization tools for power system operators. Originality/Value -- The research applies for the first time, the concepts of Recognition Primed Decision Making, Situation Awareness Levels and Cognitive Task Analysis to training of electric power system operators. Take away message -- The NDM approach provides an ideal framework for systematic training management and mitigation to accelerate learning in team-based training scenarios with high-fidelity power grid simulators.

24 POWER TRANSMISSION AND DISTRIBUTION↗

A 3D Active Learning Application for NeMO-Net, the NASA Neural Multi-Modal Observation and Training Network for Global Coral Reef Assessment

NeMO-Net, the NASA neural multi-modal observation and training network for global coral reef assessment, is an open-source deep convolutional neural network and interactive active learning training software aiming to accurately assess the present and past dynamics of coral reef ecosystems through determination of percent living cover and morphology as well as mapping of spatial distribution. We present an interactive video game prototype for tablet and mobile devices where users interactively label morphology classifications over mm-scale 3D coral reef imagery captured using fluid lensing to create a dataset that will be used to train NeMO-Nets convolutional neural network. The application currently allows for users to classify preselected regions of coral in the Pacific and will be expanded to include additional regions captured using our NASA FluidCam instrument, presently the highest-resolution remote sensing benthic imaging technology capable of removing ocean wave distortion, as well as lower-resolution airborne remote sensing data from the ongoing NASA CORAL campaign. Active learning applications present a novel methodology for efficiently training large-scale Neural Networks wherein variances in identification can be rapidly mitigated against control data. NeMO-Net periodically checks users input against pre-classified coral imagery to gauge their accuracy and utilize in-game mechanics to provide classification training. Users actively communicate with a server and are requested to classify areas of coral for which other users had conflicting classifications and contribute their input to a larger database for ranking. In partnering with Mission Blue and IUCN, NeMO-Net leverages an international consortium of subject matter experts to classify areas of confusion identified by NeMO-Net and generate additional labels crucial for identifying decision boundary locations in coral reef assessment.

NeMO=NET↗

CIRCLEZ : Reliable photometric redshifts for active galactic nuclei computed solely using photometry from Legacy Survey Imaging for DESI

Photometric redshifts for galaxies hosting an accreting supermassive black hole in their center, known as active galactic nuclei (AGNs), are notoriously challenging. At present, they are most optimally computed via spectral energy distribution (SED) fittings, assuming that deep photometry for many wavelengths is available. However, for AGNs detected from all-sky surveys, the photometry is limited and provided by a range of instruments and studies. This makes the task of homogenizing the data challenging, presenting a dramatic drawback for the millions of AGNs that wide surveys such as SRG/eROSITA are poised to detect. This work aims to compute reliable photometric redshifts for X-ray-detected AGNs using only one dataset that covers a large area: the tenth data release of the Imaging Legacy Survey (LS10) for DESI. LS10 provides deep grizW1-W4 forced photometry within various apertures over the footprint of the eROSITA-DE survey, which avoids issues related to the cross-calibration of surveys. We present the results from CIRCLEZ, a machine-learning algorithm based on a fully connected neural network. CIRCLEZ is built on a training sample of 14 000 X-ray-detected AGNs and utilizes multi-aperture photometry, mapping the light distribution of the sources. The accuracy (σNMAD) and the fraction of outliers (η) reached in a test sample of 2913 AGNs are equal to 0.067 and 11.6%, respectively. The results are comparable to (or even better than) what was previously obtained for the same field, but with much less effort in this instance. We further tested the stability of the results by computing the photometric redshifts for the sources detected in CSC2 and Chandra-COSMOS Legacy, reaching a comparable accuracy as in eFEDS when limiting the magnitude of the counterparts to the depth of LS10. The method can be applied to fainter samples of AGNs using deeper optical data from future surveys (for example, LSST, Euclid), granting LS10-like information on the light distribution beyond the morphological type. Along with this paper, we have released an updated version of the photometric redshifts (including errors and probability distribution functions) for eROSITA/eFEDS.

79 ASTRONOMY AND ASTROPHYSICS↗

Variability Extraction and Synthesis via Multi-Resolution Analysis using Distribution Transformer High-Speed Power Data

A library of load variability classes is created to produce scalable synthetic data sets using historical high-speed raw data. These data are collected from distribution monitoring units connected at the secondary side of a distribution transformer. Because of the irregular patterns and large volume of historical high-speed data sets, the utilization of current load characterization and modeling techniques are challenging. Multi-resolution analysis techniques are applied to extract the necessary components and eliminate the unnecessary components from the historical high-speed raw data to create the library of classes, which are then utilized to create new synthetic load data sets. A validation is performed to ensure that the synthesized data sets contain the same variability characteristics as the training data sets. The synthesized data sets are intended to be utilized in quasi-static time-series studies for distribution system planning studies on a granular scale, such as detailed PV interconnection studies.

24 POWER TRANSMISSION AND DISTRIBUTION↗