SEARCH · Search NASA

Results for “MMD”

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 19 records

All sources NASA NTRS DOE OSTI

DOE OSTI2020

A conservative, consistent, and scalable meshfree mimetic method

Mimetic methods discretize divergence by restricting the Gauss theorem to mesh cells. Because point clouds lack such geometric entities, construction of a compatible meshfree divergence remains a challenge. In this work, we define an abstract Meshfree Mimetic Divergence (MMD) operator on point clouds by contraction of field and virtual face moments. This MMD satisfies a discrete divergence theorem, provides a discrete local conservation principle, and is first-order accurate. We consider two MMD instantiations. The first one assumes a background mesh and uses generalized moving least squares (GMLS) to obtain the necessary field and face moments. This MMD instance is appropriate for settings where a mesh is available but its quality is insufficient for a robust and accurate mesh-based discretization. The second MMD operator retains the GMLS field moments but defines virtual face moments using computationally efficient weighted graph-Laplacian equations. This MMD instance does not require a background grid and is appropriate for applications where mesh generation creates a computational bottleneck. It allows one to trade an expensive mesh generation problem for a scalable algebraic one, without sacrificing compatibility with the divergence operator. We demonstrate the approach by using the MMD operator to obtain a virtual finite-volume discretization of conservation laws on point clouds. Finally, numerical results in the paper confirm the mimetic properties of the method and show that it behaves similarly to standard finite volume methods.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

NASA NTRS2004

A Look from the Inside

NASA's Hubble Space Telescope has given the world amazing images of the distant stars, planets, and galaxies. The cutting-edge imaging technology that enhances the Hubble images also extends its benefits to life here on Earth, from deciphering previously unreadable portions of the Dead Sea Scrolls to improving digital mammographies for advanced cancer detection. This imaging technology is now helping physicians to perform micro-invasive arthroscopic surgery, which is the visual examination of an interior joint such as the knee. In 1997, NASA's Glenn Research Center signed a Space Act Agreement with Micro Medical Devices, Inc. (MMD), a medical device engineering company that licenses miniature surgical devices. At that time, MMD was in the process of developing a new micro-endoscope. This tool enables surgeons to view what is happening inside the body on a screen, eliminating the need for a more invasive diagnostic procedure. The images from the micro-endoscope needed to be extremely clear, a challenge with the tool s small size. The images also needed to be viewed in real time to allow surgeons to see what was happening in the body at that very moment. MMD established Clear Image Technology, LLC, of Elyria, Ohio, to commercialize the miniature endoscope. Clear Image Technology then partnered with Arthrotek, Inc., a sports medicine subsidiary of Biomet, Inc., to introduce the tool to the commercial market as the InnerVue[TM] Diagnostic Scope System. The InnerVue system is designed for use in a diagnostic environment, such as an office or outpatient service, to evaluate conditions within a joint. While the InnerVue scope is primarily being applied to the knee and shoulder, other small joints such as the wrist, elbow, and ankle are being investigated. A physician can use the system alone or in conjunction with magnetic resonance imaging (MRI) to determine the next treatment step for each patient. The procedure can be more accurate than MRI, particularly with diagnosing cartilage and articular surface damage and assessing arthritis severity.

Source record↗

DOE OSTI2026

Unsupervised domain adaptation for radioisotope identification in gamma spectroscopy

Training machine learning models for radioisotope identification using gamma spectroscopy remains an elusive challenge for many practical applications, largely stemming from the difficulty of acquiring and labeling large, diverse experimental datasets. Simulations can mitigate this challenge, but the accuracy of models trained on simulated data can deteriorate substantially when deployed to an out-of-distribution operational environment. In this study, we demonstrate that unsupervised domain adaptation (UDA) can improve the ability of a model trained on synthetic data to generalize to a new testing domain, provided unlabeled data from the target domain are available. Conventional supervised techniques are unable to utilize this data because the absence of isotope labels precludes defining a supervised classification loss. Instead, we first pretrain a spectral classifier using labeled synthetic data and subsequently leverage unlabeled target data to align the learned feature representations between the source and target domains. We compare a range of different UDA techniques, finding that minimizing the maximum mean discrepancy (MMD) between source and target feature vectors yields the most consistent improvement to testing scores. For instance, using a custom transformer-based neural network, we achieved a testing accuracy of $0.904 \pm 0.022$ on an experimental LaBr test set after performing unsupervised feature alignment via MMD minimization, compared to $0.754 \pm 0.014$ before alignment. Overall, our results highlight the potential of using UDA to adapt a radioisotope classifier trained on synthetic data for real-world deployment.

Lalor, Peter W.↗

DOE OSTI2019

A meshfree mimetic divergence operator

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1996

Atomic Oxygen Durability of Second Surface Silver Microsheet Glass Concentrators

Second surface silver microsheet glass concentrators are being developed for potential use in future solar dynamic space power systems. Traditional concentrators are aluminum honeycomb sandwich composites with either aluminum or graphite epoxy face sheets, where a reflective aluminum layer is deposited onto an organic leveling layer on the face sheet. To protect the underlying layers, a SiO2 layer is applied on top of the aluminum reflective layer. These concentrators may be vulnerable to atomic oxygen degradation due to possible atomic oxygen attack of the organic layers at defect sites in the protective and reflective coatings. A second surface microsheet glass concentrator would be inherently more atomic oxygen durable than these first surface concentrators. In addition, a second surface microsheet glass concentrator design provides a smooth optical surface and allows for silver to be used as a reflective layer, which would improve the reflectivity of the concentrator and the performance of the system. A potential threat to the performance of second surface microsheet glass concentrators is atomic oxygen attack of the underlying silver at seams and edges or at micrometeoroid and debris (MMD) impacts sites. Second surface silver microsheet glass concentrator samples were fabricated and tested for atomic oxygen durability. The samples were iteratively exposed to an atomic oxygen environment in a plasma asher. Samples were evaluated for potential degradation at fabrication seams, simulated MMD impact sites, and edges. Optical microscopy was used to evaluate atomic oxygen degradation. Reflectance was obtained for an impacted sample prior to and after atomic oxygen exposure. After an initial atomic oxygen exposure to an effective fluence of approx. 1 x 10(exp 21) atoms/cm(exp 2), oxidation of the silver at defect sites and edges was observed. Exposure to an additional approx. 1 x 10(exp 21) atoms/cm(exp 2) caused no observed increase in oxidation. Oxidation at an impact site caused negligible changes in reflectance. In all cases oxidation was found to be confined to the vicinity of the seams, impact sites, edges or defect sites. Asher to in-space atomic oxygen correlation issues will be addressed.

deGroh, Kim K.↗

NASA NTRS2015

High Ice Water Content at Low Radar Reflectivity near Deep Convection: Part II. Evaluation of Microphysical Pathways in Updraft Parcel Simulations

The aeronautics industry has established that a threat to aircraft is posed by atmospheric conditions of substantial ice water content (IWC) where equivalent radar reflectivity (Ze) does not exceed 20-30 dBZ and supercooled water is not present; these conditions are encountered almost exclusively in the vicinity of deep convection. Part 1 (Fridlind et al., 2015) of this two-part study presents in situ measurements of such conditions sampled by Airbus in three tropical regions, commonly near 11 km and -43 C, and concludes that the measured ice particle size distributions are broadly consistent with past literature with profiling radar measurements of Z(sub e) and mean Doppler velocity obtained within monsoonal deep convection in one of the regions sampled. In all three regions, the Airbus measurements generally indicate variable IWC that often exceeds 2 gm (exp -3) with relatively uniform mass median area-equivalent diameter (MMD(sub eq) of 200-300 micrometers. Here we use a parcel model with size-resolved microphysics to investigate microphysical pathways that could lead to such conditions. Our simulations indicate that homogeneous freezing of water drops produces a much smaller ice MMD(sub eq) than observed, and occurs only in the absence of hydrometeor gravitational collection for the conditions considered. Development of a mass mode of ice aloft that overlaps with the measurements requires a substantial source of small ice particles at temperatures of about -10 C or warmer, which subsequently grow from water vapor. One conceivable source in our simulation framework is Hallett-Mossop ice production; another is abundant concentrations of heterogeneous ice freezing nuclei acting together with copious shattering of water drops upon freezing. Regardless of the production mechanism, the dominant mass modal diameter of vapor-grown ice is reduced as the ice-multiplication source strength increases and as competition for water vapor increases. Both mass and modal diameter are reduced by entrainment and by increasing aerosol concentrations. Weaker updrafts lead to greater mass and larger modal diameters of vapor-grown ice, the opposite of expectations regarding lofting of larger ice particles in stronger updrafts. While stronger updrafts do loft more dense ice particles produced primarily by raindrop freezing, we find that weaker updrafts allow the warm rain process to reduce competition for diffusional growth of the less dense ice expected to persist in convective outflow.

Reflectivity↗

DOE OSTI2020

Domain adaptation techniques for improved cross-domain study of galaxy mergers

In astronomy, neural networks are often trained on simulated data with the prospect of being applied to real observations. Unfortunately, simply training a deep neural network on images from one domain does not guarantee satisfactory performance on new images from a different domain. The ability to share cross-domain knowledge is the main advantage of modern deep domain adaptation techniques. Here we demonstrate the use of two techniques - Maximum Mean Discrepancy (MMD) and adversarial training with Domain Adversarial Neural Networks (DANN) - for the classification of distant galaxy mergers from the Illustris-1 simulation, where the two domains presented differ only due to inclusion of observational noise. We show how the addition of either MMD or adversarial training greatly improves the performance of the classifier on the target domain when compared to conventional machine learning algorithms, thereby demonstrating great promise for their use in astronomy.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

DOE OSTI2024

SDYN-GANs: Adversarial learning methods for multistep generative models for general order stochastic dynamics

We introduce adversarial learning methods for data-driven generative modeling of dynamics of nth-order stochastic systems. Our approach builds on Generative Adversarial Networks (GANs) with generative model classes based on stable m-step stochastic numerical integrators. From observations of trajectory samples, we introduce methods for learning long-time predictors and stable representations of the dynamics. Our approaches use discriminators based on Maximum Mean Discrepancy (MMD), training protocols using both conditional and marginal distributions, and methods for learning dynamic responses over different time-scales. We show how our approaches can be used for modeling physical systems to learn force-laws, damping coefficients, and noise-related parameters. Our adversarial learning approaches provide methods for obtaining stable generative models for dynamic tasks including long-time prediction and developing simulations for stochastic systems.

• Artificial intelligence (AI) / machine learning ↗

DOE OSTI2021

DeepMerge – II. Building robust deep learning algorithms for merging galaxy identification across domains

In astronomy, neural networks are often trained on simulation data with the prospect of being used on telescope observations. Unfortunately, training a model on simulation data and then applying it to instrument data leads to a substantial and potentially even detrimental decrease in model accuracy on the new target dataset. Simulated and instrument data represent different data domains, and for an algorithm to work in both, domain-invariant learning is necessary. Here we employ domain adaptation techniques— Maximum Mean Discrepancy (MMD) as an additional transfer loss and Domain Adversarial Neural Networks (DANNs)— and demonstrate their viability to extract domain-invariant features within the astronomical context of classifying merging and non-merging galaxies. Additionally, we explore the use of Fisher loss and entropy minimization to enforce better in-domain class discriminability. We show that the addition of each domain adaptation technique improves the performance of a classifier when compared to conventional deep learning algorithms. We demonstrate this on two examples: between two Illustris-1 simulated datasets of distant merging galaxies, and between Illustris-1 simulated data of nearby merging galaxies and observed data from the Sloan Digital Sky Survey. The use of domain adaptation techniques in our experiments leads to an increase of target domain classification accuracy of up to ~20%. With further development, these techniques will allow astronomers to successfully implement neural network models trained on simulation data to efficiently detect and study astrophysical objects in current and future large-scale astronomical surveys.

galaxies: interactions↗

DOE OSTI2019

Generative Adversarial Network for Synthetic Time Series Data Generation in Smart Grids

The availability of fine grained time series data is a pre-requisite for research in smart-grids. While data for transmission systems is relatively easily obtainable, issues related to data collection, security and privacy hinder the widespread public availability/accessibility of such datasets at the distribution system level. This has prevented the larger research community from effectively applying sophisticated machine learning algorithms to significantly improve the distribution-level accuracy of predictions and increase the efficiency of grid operations. Synthetic dataset generation has proven to be a promising solution for addressing data availability issues in various domains such as computer vision, natural language processing and medicine. However, its exploration in the smart grid context remains unsatisfactory. Previous works have tried to generate synthetic datasets by modeling the underlying system dynamics: an approach which is difficult, time consuming, error prone and often times infeasible in many problems. In this work, we propose a novel data-driven approach to synthetic dataset generation by utilizing deep generative adversarial networks (GAN) to learn the conditional probability distribution of essential features in the real dataset and generate samples based on the learned distribution. To evaluate our synthetically generated dataset, we measure the maximum mean discrepancy (MMD) between real and synthetic datasets as probability distributions, and show that their sampling distance converges. To further validate our synthetic dataset, we perform common smart grid tasks such as k-means clustering and short-term prediction on both datasets. Experimental results show the efficacy of our synthetic dataset approach: the real and synthetic datasets are indistinguishable by solely examining the output of these tasks.

14 SOLAR ENERGY↗

DOE OSTI2023

Domain Adaptation for Measurements of Strong Gravitational Lenses

Upcoming surveys are predicted to discover galaxy-scale strong lenses on the order of 10\textsuperscript{5}, making deep learning methods necessary in lensing data analysis. Currently, there is insufficient real lensing data to train deep learning algorithms, but the alternative of training only on simulated data results in poor performance on real data. Domain Adaptation may be able to bridge the gap between simulated and real datasets. We utilize domain adaptation for the estimation of Einstein radius ($\Theta_E$) in simulated galaxy-scale gravitational lensing images with different levels of observational realism. We evaluate two domain adaptation techniques - Domain Adversarial Neural Networks (DANN) and Maximum Mean Discrepancy (MMD). We train on a source domain of simulated lenses and apply it to a target domain of lenses simulated to emulate noise conditions in the Dark Energy Survey (DES). We show that both domain adaptation techniques can significantly improve the model performance on the more complex target domain dataset. This work is the first application of domain adaptation for a regression task in strong lensing imaging analysis. Our results show the potential of using domain adaptation to perform analysis of future survey data with a deep neural network trained on simulated data.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

DOE OSTI2023

Domain Adaptive Graph Neural Networks for Constraining Cosmological Parameters Across Multiple Data Sets

Deep learning models have been shown to outperform methods that rely on summary statistics, like the power spectrum, in extracting information from complex cosmological data sets. However, due to differences in the subgrid physics implementation and numerical approximations across different simulation suites, models trained on data from one cosmological simulation show a drop in performance when tested on another. Similarly, models trained on any of the simulations would also likely experience a drop in performance when applied to observational data. Training on data from two different suites of the CAMELS hydrodynamic cosmological simulations, we examine the generalization capabilities of Domain Adaptive Graph Neural Networks (DA-GNNs). By utilizing GNNs, we capitalize on their capacity to capture structured scale-free cosmological information from galaxy distributions. Moreover, by including unsupervised domain adaptation via Maximum Mean Discrepancy (MMD), we enable our models to extract domain-invariant features. We demonstrate that DA-GNN achieves higher accuracy and robustness on cross-dataset tasks. Using data visualizations, we show the effects of domain adaptation on proper latent space data alignment. This shows that DA-GNNs are a promising method for extracting domain-independent cosmological information, a vital step toward robust deep learning for real cosmic survey data.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2023

Bridging the Gap between Cosmological Simulations with Graph Neural Networks and Domain Adaptation

Deep learning models have been shown to outperform methods that rely on summary statistics, like the power spectrum, in extracting information from complex cosmological data sets. However, due to differences in the subgrid physics implementation and numerical approximations across different simulation suites, models trained on data from one cosmological simulation show a drop in performance when tested on another. Similarly, models trained on any of the simulations would also likely experience a drop in performance when applied to observational data. Training on data from two different suites of the CAMELS hydrodynamic cosmological simulations, we examine the generalization capabilities of Domain Adaptive Graph Neural Networks (DA-GNNs). By utilizing GNNs, we capitalize on their capacity to capture structured scale-free cosmological information from galaxy distributions. Moreover, by including unsupervised domain adaptation via Maximum Mean Discrepancy (MMD), we enable our models to extract domain-invariant features. We demonstrate that DA-GNN achieves higher accuracy and robustness on cross dataset tasks (up to 28% better relative error and up to almost an order of magnitude better χ 2 ). Using data visualizations, we show the effects of domain adaptation on proper latent space data alignment. This shows that DA-GNNs are a promising method for extracting domain-independent cosmological information, a vital step toward robust deep learning for real cosmic survey data.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1972

Design guide for low cost standardized payloads, volume 1

Concept point designs of low cost and refurbishable spacecraft, subsystems, and modules revealed payload program savings up to 50 percent. The general relationship of payload approaches to program costs; cost reductions from low cost standardized payloads; cost effective application of payload reliability, MMD, repair, and refurbishment; and implementation of standardization for future spacecraft are discussed. Shuttle interfaces and support equipment for future payloads are also considered

Source record↗

NASA NTRS1999

An Object-Oriented Collection of Minimum Degree Algorithms: Design, Implementation, and Experiences

The multiple minimum degree (MMD) algorithm and its variants have enjoyed 20+ years of research and progress in generating fill-reducing orderings for sparse, symmetric positive definite matrices. Although conceptually simple, efficient implementations of these algorithms are deceptively complex and highly specialized. In this case study, we present an object-oriented library that implements several recent minimum degree-like algorithms. We discuss how object-oriented design forces us to decompose these algorithms in a different manner than earlier codes and demonstrate how this impacts the flexibility and efficiency of our C++ implementation. We compare the performance of our code against other implementations in C or Fortran.

Kumfert, Gary↗

NASA NTRS2019

Comparisons of Cloud In-Situ Microphysical Properties of Deep Convective Clouds to Appendix D/P, Using Data from the HAIC-HIWC and HIWC RADAR I Flight Campaigns

In-situ cloud data from three international flight campaigns is compared to the new Federal Aviation Administration (FAA) Title 14 Code of Federal Regulations Part 33 Appendix D mixed-phase/glaciated environmental envelope (1), and the corresponding identical European Aviation Safety Agency (EASA) CS-25 Appendix P envelope (2) (hereafter "Appendix D/P"). Appendix D/P consists of a temperature-altitude envelope, a 99th percentile total water content (TWC99) envelope at the 17.4 Nm distance scale, a distance factor to estimate TWC99 values at other distance scales, ice crystal median mass diameter (MMD), and recommended liquid water content (LWC) levels in mixed-phase icing conditions. The dataset is from 45 flight missions in 3 tropical locations, with 472 runs in approximately 115 clouds, providing about 29,600 Nm of in-cloud data in deep convection over four targeted temperature intervals: 10, 30, 40, and 50 +5 C. The measurements span altitueds from about 17,000' to 39,000'. The comparisons will serve as a basis for regulatory and industry assessment of the efficacy of Appendix D/P.

Strapp, J. Walter↗

NASA NTRS2020

VEG-04: The Effects of Light Quality on Mizuna Mustard Growth, Nutritional Composition, and Organoleptic Acceptability for a Space Diet

Growing fresh, nutritious, palatable produce for crew consumption during spaceflight may provide health-promoting, bioavailable nutrients and enhance the astronaut dietary experience as we move toward longer-duration missions. Tending plants may also serve as a countermeasure for crew psychological stresses associated with spaceflight. However, requirements to support consistent growth of a variety of high quality, nutritious crops under spaceflight environmental conditions remain unclear. This study explores the potential to grow crops for consumption on the International Space Station (ISS) using the Veggie vegetable-production system. VEG-04A and B were two flight tests conducted in 2019 with the leafy green crop mizuna mustard. Mizuna was grown in two Veggie chambers simultaneously, with the chambers set to different red-to-blue light formulations; one Veggie was programmed as "red-rich" and the second as "blue-rich." Light quality is known to impact plant growth, nutrition, microbiology, and organoleptic characteristics on Earth, and the Veggie flight tests examined how these impacts might differ in microgravity. VEG-04A, a 35-day growth test with a single harvest, was initiated in June and harvested in July 2019. At harvest, the astronauts froze half of the edible plant tissue to return to Earth and weighed the remaining half using the Mass Measurement Device (MMD). Weighed samples were then cleaned with produce-sanitizing wipes, and consenting crew members participated in organoleptic evaluation of the fresh produce. The remaining sanitized produce was available for crew consumption as desired. Frozen flight samples were returned at the end of August for microbial and chemical analyses to assess food safety and nutritional quality. No pathogens were detected on VEG-04A flight or ground control samples. On average, bacterial and fungal counts were significantly lower on ground control samples than flight samples. VEG-04B, a 56-day test with multiple harvests from the same plants, assessed sustained productivity. VEG-04B was initiated in October 2019 with three harvests at four, six, and eight weeks after initiation. Challenges with the watering program occurred early during VEG-04A, and several plants failed to survive in both the flight and ground control operations. Thus, prior to VEG-04B, an extra test was conducted to tailor water timing and volumes. This test determined that mizuna grew best if the wicks inside the plant pillow were allowed to dry after plants germinated, reducing persistent water around the stem. The wicks changed from being a conduit for water out of the plant pillow to being a conduit for air into the root zone. This test allowed a fine tuning of methods for VEG-04B. It is our hope that these tests on ISS will help mitigate the risk of an inadequate food supply for long-duration missions by adding fresh vegetables to the crew diet. This research was co-funded by the Human Research Program and Space Biology (MTL#1075) in the ILSRA 2015 NRA call.

Massa, Gioia D.↗

NASA NTRSDate not supplied

VEG-04 Pick-and-Eat Crop Production and Human Research on the International Space Station

Growing fresh, nutritious, palatable produce for crew consumption during spaceflight may provide health-promoting, bioavailable nutrients and enhance the astronaut dietary experience as we move toward longer-duration missions. Tending plants may also serve as a countermeasure for crew psychological stresses associated with spaceflight. However, requirements to support consistent growth of a variety of nutritious crops under spaceflight environmental conditions remain unclear. This study explores the potential to grow crops for consumption on the International Space Station (ISS) using the Veggie vegetable-production system. VEG-04A and B were two flight tests conducted in 2019 with the leafy green crop Mizuna mustard. Mizuna was grown in two Veggie chambers simultaneously, with the chambers set to different red-to-blue light formulations; one Veggie was programmed as “red-rich” and the second as “blue-rich.” Light quality is known to impact plant growth, nutrition, microbiology, and organoleptic characteristics on Earth, and the Veggie flight tests examined how these impacts might differ in microgravity. VEG-04A was a 35-day growth test with a single harvest. VEG-04B, a 58-day test with harvests at four, six, and eight weeks from the same plants, assessed sustained productivity. Challenges with the watering program occurred early during VEG-04A, and several plants failed to survive in both the flight and ground control operations. Thus, prior to VEG-04B, an extra test was conducted to tailor water timing and volumes. This test allowed a fine tuning of methods for VEG-04B, and generally plant survival was better in that experiment. At each harvest, the astronauts froze half of the edible plant tissue to return to Earth and weighed the remaining half using the ISS Mass Measurement Device (MMD). Weighed samples were then cleaned with produce-sanitizing wipes, and consenting crew members participated in organoleptic evaluation of the fresh produce. Organoleptic evaluations were conducted on the Mizuna crops grown in both light treatments, and data from these tests are compared to ground data collected at JSC. The remaining sanitized produce was available for crew consumption as desired. Frozen flight samples were returned for microbial and chemical analyses to assess food safety and nutritional quality. Microbial assessments included culturing and identifying aerobic bacteria, yeasts, and molds, and screening for specific human pathogens. Chemical nutrient analyses included assessing elements, antioxidants, and phenolics in plant tissues. Crew members involved in plant operations completed behavioral health surveys pre-flight, in-flight associated with plant growth and harvest activities, and post-flight. Surveys captured the amount of time spent on different plant-related operations, enjoyment of the different activities, engagement with Veggie, experience interacting with Veggie, and sensory stimulation associated with growing plants in Veggie.

Veggie↗