SEARCH · Engineering Papers

Results for “Generative Neural 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 217 records · Page 12

All sources NASA NTRS DOE OSTI

NASA NTRS1992

Using neural networks and Dyna algorithm for integrated planning, reacting and learning in systems

The traditional AI answer to the decision making problem for a robot is planning. However, planning is usually CPU-time consuming, depending on the availability and accuracy of a world model. The Dyna system generally described in earlier work, uses trial and error to learn a world model which is simultaneously used to plan reactions resulting in optimal action sequences. It is an attempt to integrate planning, reactive, and learning systems. The architecture of Dyna is presented. The different blocks are described. There are three main components of the system. The first is the world model used by the robot for internal world representation. The input of the world model is the current state and the action taken in the current state. The output is the corresponding reward and resulting state. The second module in the system is the policy. The policy observes the current state and outputs the action to be executed by the robot. At the beginning of program execution, the policy is stochastic and through learning progressively becomes deterministic. The policy decides upon an action according to the output of an evaluation function, which is the third module of the system. The evaluation function takes the following as input: the current state of the system, the action taken in that state, the resulting state, and a reward generated by the world which is proportional to the current distance from the goal state. Originally, the work proposed was as follows: (1) to implement a simple 2-D world where a 'robot' is navigating around obstacles, to learn the path to a goal, by using lookup tables; (2) to substitute the world model and Q estimate function Q by neural networks; and (3) to apply the algorithm to a more complex world where the use of a neural network would be fully justified. In this paper, the system design and achieved results will be described. First we implement the world model with a neural network and leave Q implemented as a look up table. Next, we use a lookup table for the world model and implement the Q function with a neural net. Time limitations prevented the combination of these two approaches. The final section discusses the results and gives clues for future work.

Lima, Pedro↗

NASA NTRS1991

Improving classification of crop residues using digital land ownership data and Landsat TM imagery

Plant residue on the surface of cultivated soils in Miami County, Indiana is analyzed in terms of quantity and type with Landsat TM data to generate information for a conservation program for agricultural soil. The Landsat data are enhanced with land-ownership data in a geographic information system to facilitate classification with maximum-likelihood, minimum-distance, and neural-network classifiers. The most effective classifications resulted from the use of the neural network on the enhanced TM data.

Zhuang, Xin↗

NASA NTRS1998

PIMS Data Storage, Access, and Neural Network Processing

The Principal Investigator Microgravity Services (PIMS) project at NASA's Lewis Research Center has supported microgravity science Principal Investigator's (PIs) by processing, analyzing, and storing the acceleration environment data recorded on the NASA Space Shuttles and the Russian Mir space station. The acceleration data recorded in support of the microgravity science investigated on these platforms has been generated in discrete blocks totaling approximately 48 gigabytes for the Orbiter missions and 50 gigabytes for the Mir increments. Based on the anticipated volume of acceleration data resulting from continuous or nearly continuous operations, the International Space Station (ISS) presents a unique set of challenges regarding the storage of and access to microgravity acceleration environment data. This paper presents potential microgravity environment data storage, access, and analysis concepts for the ISS era.

McPherson, Kevin M.↗

NASA NTRS2003

Artificial Neural Network with Hardware Training and Hardware Refresh

A neural network circuit is provided having a plurality of circuits capable of charge storage. Also provided is a plurality of circuits each coupled to at least one of the plurality of charge storage circuits and constructed to generate an output in accordance with a neuron transfer function. Each of a plurality of circuits is coupled to one of the plurality of neuron transfer function circuits and constructed to generate a derivative of the output. A weight update circuit updates the charge storage circuits based upon output from the plurality of transfer function circuits and output from the plurality of derivative circuits. In preferred embodiments, separate training and validation networks share the same set of charge storage circuits and may operate concurrently. The validation network has a separate transfer function circuits each being coupled to the charge storage circuits so as to replicate the training network s coupling of the plurality of charge storage to the plurality of transfer function circuits. The plurality of transfer function circuits may be constructed each having a transconductance amplifier providing differential currents combined to provide an output in accordance with a transfer function. The derivative circuits may have a circuit constructed to generate a biased differential currents combined so as to provide the derivative of the transfer function.

Tuan A Duong↗

NASA NTRS2003

Generative Representations for Evolving Families of Designs

Since typical evolutionary design systems encode only a single artifact with each individual, each time the objective changes a new set of individuals must be evolved. When this objective varies in a way that can be parameterized, a more general method is to use a representation in which a single individual encodes an entire class of artifacts. In addition to saving time by preventing the need for multiple evolutionary runs, the evolution of parameter-controlled designs can create families of artifacts with the same style and a reuse of parts between members of the family. In this paper an evolutionary design system is described which uses a generative representation to encode families of designs. Because a generative representation is an algorithmic encoding of a design, its input parameters are a way to control aspects of the design it generates. By evaluating individuals multiple times with different input parameters the evolutionary design system creates individuals in which the input parameter controls specific aspects of a design. This system is demonstrated on two design substrates: neural-networks which solve the 3/5/7-parity problem and three-dimensional tables of varying heights.

Hornby, Gregory S.↗

NASA NTRS2004

Examples of Current and Future Uses of Neural-Net Image Processing for Aerospace Applications

Feed forward artificial neural networks are very convenient for performing correlated interpolation of pairs of complex noisy data sets as well as detecting small changes in image data. Image-to-image, image-to-variable and image-to-index applications have been tested at Glenn. Early demonstration applications are summarized including image-directed alignment of optics, tomography, flow-visualization control of wind-tunnel operations and structural-model-trained neural networks. A practical application is reviewed that employs neural-net detection of structural damage from interference fringe patterns. Both sensor-based and optics-only calibration procedures are available for this technique. These accomplishments have generated the knowledge necessary to suggest some other applications for NASA and Government programs. A tomography application is discussed to support Glenn's Icing Research tomography effort. The self-regularizing capability of a neural net is shown to predict the expected performance of the tomography geometry and to augment fast data processing. Other potential applications involve the quantum technologies. It may be possible to use a neural net as an image-to-image controller of an optical tweezers being used for diagnostics of isolated nano structures. The image-to-image transformation properties also offer the potential for simulating quantum computing. Computer resources are detailed for implementing the black box calibration features of the neural nets.

Decker, Arthur J.↗

NASA NTRS2006

Improving Neural Network Generalization Ability Using Outlier Analysis and Voronoi Tessellation

The data used in this study was obtained from the Sloan Digital Sky Survey (SDSS), which provides astronomers with what is currently the most extensive mapping of the universe, covering 25% of the sky and cataloging the spectral properties (e.g., luminosity, color, surface temperature) of over 100 million celestial objects. Images generated by the SDSS are collected through 5 filters named u, g, r, l, and z that have respective wavelengths of 3540, 4750, 6222, 7632, and 9049 A. By measuring the photometric redshifts of the aforementioned wavelengths of a galaxy, astronomers can ascertain the extent to which galaxy is receding from which the distance to the galaxy can be calculated. Data collected for a small select group of galaxies (approximately 30, 000) contains accurate measurements of the galaxies' redshifts, in addition to measurements of their spectral properties. The above dataset containing both redshift measurements as well as spectral properties of the selected galaxies served as the training set for the purposes of this study; the data set containing only the spectra properties of a separate group of galaxies served as the test set.

Ho, Michelle↗

NASA NTRS2017

Modulation Classification of Satellite Communication Signals Using Cumulants and Neural Networks

National Aeronautics and Space Administration (NASA)'s future communication architecture is evaluating cognitive technologies and increased system intelligence. These technologies are expected to reduce the operational complexity of the network, increase science data return, and reduce interference to self and others. In order to increase situational awareness, signal classification algorithms could be applied to identify users and distinguish sources of interference. A significant amount of previous work has been done in the area of automatic signal classification for military and commercial applications. As a preliminary step, we seek to develop a system with the ability to discern signals typically encountered in satellite communication. Proposed is an automatic modulation classifier which utilizes higher order statistics (cumulants) and an estimate of the signal-to-noise ratio. These features are extracted from baseband symbols and then processed by a neural network for classification. The modulation types considered are phase-shift keying (PSK), amplitude and phase-shift keying (APSK),and quadrature amplitude modulation (QAM). Physical layer properties specific to the Digital Video Broadcasting - Satellite- Second Generation (DVB-S2) standard, such as pilots and variable ring ratios, are also considered. This paper will provide simulation results of a candidate modulation classifier, and performance will be evaluated over a range of signal-to-noise ratios, frequency offsets, and nonlinear amplifier distortions.

Signal Recognition↗

NASA NTRS2017

Modulation Classification of Satellite Communication Signals Using Cumulants and Neural Networks

Signal Classification↗

NASA NTRS2021

Large-Scale High-Resolution Coastal Mangrove Forests Mapping Across West Africa With Machine Learning Ensemble and Satellite Big Data

Coastal mangrove forests provide important ecosystem goods and services, including carbon sequestration, biodiversity conservation, and hazard mitigation. However, they are being destroyed at an alarming rate by human activities. To characterize mangrove forest changes, evaluate their impacts, and support relevant protection and restoration decision making, accurate and up-to-date mangrove extent mapping at large spatial scales is essential. Available large-scale mangrove extent data products use a single machine learning method commonly with 30 m Landsat imagery, and significant inconsistencies remain among these data products. With huge amounts of satellite data involved and the heterogeneity of land surface characteristics across large geographic areas, finding the most suitable method for large-scale high-resolution mangrove mapping is a challenge. The objective of this study is to evaluate the performance of a machine learning ensemble for mangrove forest mapping at 20 m spatial resolution across West Africa using Sentinel-2 (optical) and Sentinel-1 (radar) imagery. The machine learning ensemble integrates three commonly used machine learning methods in land cover and land use mapping, including Random Forest (RF), Gradient Boosting Machine (GBM), and Neural Network (NN). The cloud-based big geospatial data processing platform Google Earth Engine (GEE) was used for pre-processing Sentinel-2 and Sentinel-1 data. Extensive validation has demonstrated that the machine learning ensemble can generate mangrove extent maps at high accuracies for all study regions in West Africa (92%–99% Producer’s Accuracy, 98%–100% User’s Accuracy, 95%–99% Overall Accuracy). This is the first-time that mangrove extent has been mapped at a 20 m spatial resolution across West Africa. The machine learning ensemble has the potential to be applied to other regions of the world and is therefore capable of producing high-resolution mangrove extent maps at global scales periodically.

coastal environment↗

NASA NTRS1990

Neural networks as a control methodology

While conventional computers must be programmed in a logical fashion by a person who thoroughly understands the task to be performed, the motivation behind neural networks is to develop machines which can train themselves to perform tasks, using available information about desired system behavior and learning from experience. There are three goals of this fellowship program: (1) to evaluate various neural net methods and generate computer software to implement those deemed most promising on a personal computer equipped with Matlab; (2) to evaluate methods currently in the professional literature for system control using neural nets to choose those most applicable to control of flexible structures; and (3) to apply the control strategies chosen in (2) to a computer simulation of a test article, the Control Structures Interaction Suitcase Demonstrator, which is a portable system consisting of a small flexible beam driven by a torque motor and mounted on springs tuned to the first flexible mode of the beam. Results of each are discussed.

Mccullough, Claire L.↗

NASA NTRS2018

Automatic Speech Recognition for Launch Control Center Communication Using Recurrent Neural Networks with Data Augmentation and Custom Language Model

Transcribing voice communications in NASA’s launch control center is important for information utilization. However, automatic speech recognition in this environment is particularly challenging due to the lack of training data, unfamiliar words in acronyms, multiple different speakers and accents, and conversational characteristics of speaking. We used bidirectional deep recurrent neural networks to train and test speech recognition performance. We showed that data augmentation and custom language models can improve speech recognition accuracy. Transcribing communications from the launch control center will help the machine analyze information and accelerate knowledge generation.

Chow, Edward↗

NASA NTRSDate not supplied

Snow Depth from AMSR-2 Using Multispectral Satellite Data in an Artificial Neural Network

By using diffusion theory and Monte Carlo lidar radiative transfer simulations, Hu et al. (2022b) has derived snow depth from the first-, second- and third-order moments of the lidar backscattering pathlength distribution. Lu et al. (2022) calculated the snow depth by applying the methods to the satellite ICESat-2 lidar measurements over the Arctic sea ice, as well as land surfaces of Northern Hemisphere. In this paper, an artificial neural network (ANN) algorithm, employing several channels from Advanced Microwave Scanning Radiometer 2 (AMSR-2) and the humidity vertical profiles from Global Modeling and Assimilation Office (GMAO) Goddard Earth Observing System for Instrument Teams (GEOS-IT) product, is trained to determine snow depth identified by time and geolocation matched 2019 ICESat-2 snow-depth data during winter months over the Arctic sea ice. The trained ANN snow-depth was applied to 2018 AMSR-2 clear pixel data, although the algorithms perform reasonably well in thinner clouds. The validation data (different from the training set) of ANN snow depth from AMSR-2 showed a good agreement with time matched and co-located snow-depth values from ICESat-2. The bias was near zero, with mean absolute error (MAE) 0.05 cm and a root-mean-square-error (RMSE) 0.08 cm. Prior applying the trained ANN snow depth to AMSR-2 data, a cloud screening algorithm was developed with a similar approach. A separate ANN cloud mask was trained to determine an AMSR-2 pixel is clear or cloudy with time and geolocation matched 2015 CALIOP Vertical Feature Mask (VFM) over Arctic sea ice. The ANN cloud mask from AMSR-2 under-estimated cloud fraction by 3-6% compared to CALIOP . The additional research is needed to conclusively evaluate the ANN cloud mask accuracy. Finally, this paper will lay the foundation for a sustained long-term snowfall and snow-storm monitoring system. The future Cloud Aerosol LIdar for Global scale Observations of the ocean-Land Atmosphere system (CALIGOLA) mission will provide a means to calculate snow depth from the lidar backscattering pathlength distribution, benefiting from the UV, visible and infrared pulses. With the calculated snow depth as the truth one could develop a machine learning algorithm, as it was done in this paper, using a passive microwave instrument available at that time to generate a wide range of snow depth data, covering extensive spatial areas in the cross-orbit direction.

Neural Network↗

NASA NTRSDate not supplied

Structured Covariance Gaussian Networks for Orion Crew Module Aerodynamic Uncertainty Quantification

In this paper we propose a new approach for nonlinear regression and uncertainty quantification. The method is based on a pair of neural networks which parameterize mean and dense covariance functions of a multivariate Gaussian process, trained together to maximize the log-likelihood of observing the given data. The covariance matrix is made positive definite at every input by construction. We also propose a sampling approach that produces viable surrogate function realizations from the Gaussian process. We call the proposed model a Structured Covariance Gaussian Network (SCGN). We illustrate the use of SCGNs for learning an aerodynamic response surface with built-in uncertainty for the Orion crew module. We find that SCGN provides an efficient and systematic way to learn nonlinear functional relationships and dense covariances. We compare results to a baseline Gaussian process regressor and observe that the SCGN provides comparable uncertainty descriptions with improved scalability to dataset size. The sample functions generated by SCGN are fast to evaluate online and are therefore convenient for use in trajectory simulations. These results suggest that SCGN may be a viable computational method for aerodynamic uncertainty quantification.

machine learning↗

NASA NTRSDate not supplied

Structured Covariance Gaussian Networks for Orion Crew Module Aerodynamic Uncertainty Quantification

machine learning↗

NASA NTRS2019

Developing a Machine-Learning-Based Processing Framework for Twitter and Other Crowdsourced Data

Crowdsourced data streams such as Twitter and other social media are important sources of real-time and historical global information for Earth science applications. At the NASA Goddard Earth Sciences Data and Information Services Center (GES DISC), we have been exploring the Twitter data stream for its potential in augmenting the validation program of NASA's Global Precipitation Measurement (GPM) mission. To realize this potential, we need to increase the information density and enhance the quality of filtered precipitation tweets. We have implemented various components of a machine learning (ML)-based processing infrastructure for crowdsourced data that outputs, in this instance, useful and usable information derived from precipitation tweets. We have test enriched the Twitter stream with higher quality active tweets from those knowingly contributing to our effort and from existing crowdsourced programs (e.g., mPING, CoCoRaHS). We have experimented with various algorithms for processing tweets, including NaÃ ve Bayes, Convolutional Neural Network (CNN), Hierarchical Attention Network (HAN), and semi-supervised learning (with tri-training). Our current work focuses on (1) automated review of Earth science-related publications to determine relationships between discipline research needs and ML algorithms; (2) investigating Sequential Generative Adversarial Network (SeqGAN) for processing precipitation tweets for anomaly detection; and (3) managing crowdsourced data in a way that is compatible with existing NASA satellite data archives and using the data for ML applications. Key results include (1) network visualization of NLP-processed publications in various Earth science disciplines; (2) difference between GPM-linked, generated tweets and collected actual tweets that is small for GPM-determined light to moderate rain cases and high for GPM-determined heavy rain cases; and (3) identification of MongoDB for storing raw tweets and Zarr format for gridded tweets (compatible with GPM data). Our results have taken us a step closer to an operational ML-based tweet processing infrastructure and have already demonstrated that tweet-derived precipitation information is potentially useful for validation of Earth science satellite data.

Teng, William↗

NASA NTRS1997

Applying Space Technology to Enhance Control of an Artificial Arm

At the present time, myoelectric prostheses perform only one function of the hand: open and close with the thumb, index and middle finger coming together to grasp various shaped objects. To better understand the limitations of the current single-function prostheses and the needs of the individuals who use them, The Institute for Rehabilitation and Research (TIRR), sponsored by the National Institutes of Health (August 1992 - November 1994), surveyed approximately 2500 individuals with upper limb loss. When asked to identify specific features of their current electric prosthesis that needed improvement, the survey respondents overwhelmingly identified the lack of wrist and finger movement as well as poor control capability. Simply building a mechanism with individual finger and wrist motion is not enough. Individuals with upper limb loss tend to reject prostheses that require continuous visual monitoring and concentration to control. Robotics researchers at NASA's Johnson Space Center (JSC) and Rice University have made substantial progress in myoelectric teleoperation. A myoelectric teleoperation system translates signals generated by an able-bodied robot operator's muscles during hand motions into commands that drive a robot's hand through identical motions. Farry's early work in myoelectric teleoperation used variations over time in the myoelectric spectrum as inputs to neural networks to discriminate grasp types and thumb motions. The resulting schemes yielded up to 93% correct classification on thumb motions. More recently, Fernandez achieved 100% correct non-realtime classification of thumb abduction, extension, and flexion on the same myoelectric data. Fernandez used genetic programming to develop functions that discriminate between thumb motions using myoelectric signal parameters. Genetic programming (GP) is an evolutionary programming method where the computer can modify the discriminating functions' form to improve its performance, not just adjust numerical coefficients or weights. Although the function development may require much computational time and many training cases, the resulting discrimination functions can run in realtime on modest computers. These results suggest that myoelectric signals might be a feasible teleoperation medium, allowing an operator to use his or her own hand and arm as a master to intuitively control an anthropomorphic robot in a remote location such as outer space.

Atkins, Diane↗

NASA NTRS2005

Creation and testing of an artificial neural network based carbonate detector for Mars rovers

We have developed an artificial neural network (ANN) based carbonate detector capable of running on current and future rover hardware. The detector can identify calcite in visible/NIR (350-2500 nm) spectra of both laboratory specimens covered by ferric dust and rocks in Mars analogue field environments. The ANN was trained using the Backpropagation algorithm with sigmoid activation neurons. For the training dataset, we chose nine carbonate and eight non-carbonate representative mineral spectra from the USGS spectral library. Using these spectra as seeds, we generated 10,000 variants with up to 2% Gaussian noise in each reflectance measurement. We cross-validated several ANN architectures, training on 9,900 spectra and testing on the remaining 100. The best performing ANN correctly detected, with perfect accuracy, the presence (or absence) of carbonate in spectral data taken on field samples from the Mojave desert and clean, pure marbles from CT. Sensitivity experiments with JSC Mars-1 simulant dust suggest the carbonate detector would perform well in aeolian Martian environments.

neural networks↗