Search NASA⌕ Search

SEARCH · Search NASA

Results for “Machine Learning Models”

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 703 records · Page 39

An Advanced Open-Source Platform for Air Quality Analysis, Visualization, and Prediction

Ambient air pollution is the largest environmental health risk factor, leading to several million premature deaths globally per year. The challenge of combating poor air quality is exacerbated by growing urban populations, changing emissions, and a warming climate. While there have been many advances monitoring and modeling of atmospheric composition, reflected in the dramatic increase in archived Earth Observations, there is no single measurement or method that alone can provide an accurate depiction of the entire atmosphere. The rapidly growing collections of observational and modeling data require us to be smarter about what data to include, and how such data is used. In recent years, NASA has invested significantly in advancing the concepts for Analytics Collaborative Framework (ACF) [5] and New Observing Strategies (NOS) [4] to tackle our software infrastructure need for harmonized data management and dynamic acquisition of diverse measurements for on-demand, interactive, multivariate analysis, and access [3]. It is not enough to have a big data, standalone analytics solution; it is critical that we start integrating data from remote sensing, modeling, and in-situ networks in a harmonized manner that enables timely and data-driven decision-making for air quality management. This work presents the design and development of an Air Quality Analytics Collaborative Framework (AQ ACF), as part of NASA’s Advanced Information Systems Technology (AIST) effort, to establish a data, machine-learning, and numerically driven platform for air quality analysis, visualization, and prediction.

Liu, Qian↗

Noise Reduction for Solar-Induced Fluorescence Retrievals Using Machine Learning and Principal Component Analysis: Simulations and Applications to GOME-2 Satellite Retrievals

We use a spectral-based approach that employs principal component analysis along with a relatively shallow artificial neural network (NN) to substantially reduce noise and other artifacts in terrestrial chlorophyll solar-induced fluorescence (SIF) retrievals. SIF is a very small emission at red and far-red wavelengths that is difficult to measure and is highly sensitive to random errors and systematic artifacts. Our approach relies upon an assumption that a trained NN can effectively reconstruct the total SIF signal from a relatively small number of leading principal components of the satellite-observed far-red radiance spectra without using information from the trailing modes that contain most of the random errors. We test the approach with simulated reflectance spectra produced with a full atmospheric and surface radiative transfer model using different observing and geophysical parameters and various noise levels. Resulting noisy and noise-reduced retrieved SIF values are compared with true values to assess performance. We then apply our noise reduction approach to SIF derived from two different satellite spectrometers. For evaluation, since the truth in this case is unknown, we compare SIF retrievals from two independent sensors with each other. We also compare the noise-reduced SIF temporal variations with those from an independent gross primary product (GPP) product that should display similar variations. Results show that our noise reduction approach improves capture of SIF seasonal and inter-annual variability. Our approach should be applicable to many noisy data products derived from spectral measurements. Our methodology does not replace the original retrieval algorithms; rather the original noisy retrievals are needed as the target for the NN training process.

noise reduction, solar-induced fluorescence, SIF, ↗

Pixel Based Model for High Latitude Dust Detection

High Latitude Dust (HLD, ≥ 50°𝑁 𝑎𝑛𝑑 ≥ 40°𝑆 ) load has implications on the energy budget, ocean biodiversity and economy on a regional and global scale. Current methods of dust detection rely on spectral sensitivity at visible (RGB) and infrared wavelengths. The characteristics of HLD vary according to the sediments and sedimentary processes operating on the land surface that are the source of the dust particles. Leveraging machine learning (ML) methods, we propose a new detection method based on convolutional neural network (CNN) using true color images.

Priftis, Georgios↗

High-Performance Computing Optimization for Aladyn – Adaptive Neural Network Molecular Dynamics Mini-Application

This report provides a description and performance evaluation of the optimization techniques for high performance computing (HPC) implementation of the open source Computational Materials mini-application Aladyn (https://github.com/nasa/aladyn). Aladyn is a basic molecular dynamics code written in FORTRAN 2003, which is designed to demonstrate the use of adaptive neural networks (ANNs) in atomistic simulations. The role of ANNs is to efficiently reproduce the very complex energy landscape resulting from the atomic interactions in materials with the accuracy of the more expensive quantum mechanics-based calculations. The ANN is trained on a large set of atomic structures calculated using the density functional theory (DFT) method. While achieving orders of magnitude faster computational performance than DFT, the ANN-based approach was still very computationally demanding compared to the conventional approach of using empirically fitted energy functions. After its initial development, Aladyn was evaluated and optimized by experts at the NASA Advanced Supercomputing (NAS) division to exploit modern supercomputer architectures. The code has been optimized for execution on multicore central processing units (CPUs), including Intel® Skylake microarchitecture, and on graphic accelerators, such as Nvidia® V100 graphic processing units (GPUs), using Open Multi-Processing (OpenMP) and Open Accelerators (OpenACC) programming interfaces. The optimization achieved a speedup of 4.7 times the baseline version on CPU performance and an additional 2.4 times on CPU+GPU performance. Atomistic computer simulations are a fundamental tool in materials research to model material properties form physics-based first principles. Atomic interaction, governed by Quantum Mechanics (QM) require sophisticated and highly computationally demanding mathematical models to calculate [1]. Classical methods use approximate functional forms, empirically fitted through a set of variable parameters to emulate atomic energies as direct functions of atomic coordinates [2]. While empirical potentials are computationally much simpler, allowing simulations of large-scale systems of up to a trillion (1012) atoms [3], they are substantially less accurate compared to quantum calculations and applicable only to very specific atomic configurations or predefined crystallographic phases. A recently suggested approach is to use heuristic machine learning methods [4], such as those based on Adaptive Neural Networks (ANNs) to predict atomic energies, after being trained on a sufficiently large database of QM-calculated structures [5,6]. This approach reduces significantly the computational complexity, allowing for simulations of orders of magnitude larger systems compared to QM-based methods without compromising accuracy. Still, compared to classical methods using empirical energy functions, ANN methods remain two- to three orders of magnitude more computationally demanding. Hence, the computational cost of simulations, together with the need for extensive training of ANNs, still makes the practical implementation of ANN-based methods quite challenging. The purpose of the Aladyn mini-application software [7], available as open source at https://github.com/nasa/aladyn, is to be a testbed for exploring possible optimization strategies to develop highly scalable parallel algorithms for ANN-based atomistic simulations. Aladyn is aimed at utilizing the architecture of the high-end modern highperformance computing (HPC) hardware based on multicore central processing units (CPUs) equipped with graphic processing unit (GPU) accelerators. Specifically, the goal is to optimize the performance on a single HPC compute node, before implementing scaling to multi-node parallelization using message passing interface (MPI). At the same time, the open source code of Aladyn can serve as a training model for students and professors in academia.

Yamakov, Vesselin I.↗

Landslide Likelihood Prediction using Machine Learning Algorithms

The supply of electricity via power plants is criticalto the operation of many critical infrastructure systems in mod-ern society. Natural hazards can disrupt the power supply, causepower outages that can halt economic growth, and impede emer-gency response until power is restored. The proposed work aimsto predict the landslides likelihood in these critical infrastructurelocations in the Northeastern USA using integrated databases ofexplanatory variables and machine learning algorithms. First,data related to landslides are obtained and merged, includingtopographic, soil moisture, and precipitation-related data. Fiveregression algorithms, namely: Random Forest, Extreme Gradi-ent Boosting (XGBoost), K-Nearest Neighbor regression (KNN),Linear Support Vector Regressor (SVR), and Linear regression,are utilized to predict the landslide probability and evaluatedon the dataset. The accuracy of the models is assessed by usingstatistical metrics such as mean absolute error (MAE), meansquared error (MSE), and root mean squared error (RMSE).The study results show that Random Forest outperformed othermodels with the mutual information feature selection method.It achieved an MSE of 0.0011 with mutual information-basedfeature selection and an MSE of 0.00157 without feature selection.KNN regressor outperformed the other models with an MSEof 0.00139 with correlation-based information selection. Theproposed landslide identification model with Random Forestalgorithm shows outstanding robustness and great potential intackling the landslide likelihood prediction by employing MLalgorithms.

Vasundhara Acharya↗

Predicting Airport Runway Configuration for Decision-Support Using Supervised Learning

One of the most challenging tasks for air traffic controllers is runway configuration management (RCM). It deals with the optimal selection of runways to operate on (for arrivals and departures) based on current and forecast of traffic, surface wind speed, wind direction, other environmental variables, noise constraints, and several other airport-specific factors. In this paper, a methodology using supervised learning is developed to build a predictive model for RCM decision-support from large volumes of historical data. Data from two full years (2018 and 2019) related to current and forecast weather, demand/capacity, etc. is collected, analyzed, and fused together. A variety of supervised learning algorithms are tested for predicting runway configuration and hyperparameter tuning is carried out to select the best performing model. The validation process involves two airports of low (Charlotte Douglas International Airport, CLT) and high (Denver International Airport, DEN) complexity of configuration decision-making. The results show significant promise for the two airports with test accuracy of 93% (CLT) and 73% (DEN). The methodology is scalable and generalizable to other airports across the U.S. National Airspace System.

air traffic management↗

Predicting Airport Runway Configurations for Decision-Support Using Supervised Learning

One of the most challenging tasks for air traffic controllers is runway configuration management (RCM). It deals with the optimal selection of runways to operate on (for arrivals and departures) based on current and forecast of traffic, surface wind speed, wind direction, other environmental variables, noise constraints, and several other airport-specific factors. In this paper, a methodology using supervised learning is developed to build a predictive model for RCM decision-support from large volumes of historical data. Data from two full years (2018 and 2019) related to current and forecast weather, demand/capacity, etc. is collected, analyzed, and fused together. A variety of supervised learning algorithms are tested for predicting runway configuration and hyperparameter tuning is carried out to select the best performing model. The validation process involves two airports of low (Charlotte Douglas International Airport, CLT) and high (Denver International Airport, DEN) complexity of configuration decision-making. The results show significant promise for the two airports with test accuracy of 93% (CLT) and 73% (DEN). The methodology is scalable and generalizable to other airports across the U.S. National Airspace System.

air traffic management↗

Tests characterizing bioprocessor hardware for analytical modeling

The tests outlined in this paper were used to characterize the hardware components of the Salad Machine, a small NASA-developed bioprocessor. The data from these tests are presented, and the methods by which this data can be integrated into system mathematical models are briefly discussed. The subsystems and physical processes discussed include the lighting system, the air loop (condensing heat exchanger and the blower), heat transfer to the surroundings, and leakage. Through this effort it was learned that in the development of a test protocol, care should be taken to order the tests such that environmental parameters, particularly humidity, require as few large adjustments as possible. Sensor calibration and installation take a substantial amount of time, which should be built into the test schedule. Two properties were particularly hard to quantify: the air flow rate and the energy from the lighting system entering into the growth volume. Flow rate can be measured using the appropriate device for the system configuration and airflow. Lighting system radiation level was measured using three methods. The results of these methods varied substantially, putting off conclusive quantification of this value.

Gustavino, S.↗

An Artificial Neural Network Approach to Predict Rotor-Airframe Acoustic Waveforms

A surrogate artificial neural network/machine learning model was developed to predict the acoustic interaction for a fixed-pitch rotor in proximity to a downstream cylindrical airframe typical of small Unmanned Aerial System (sUAS) platforms. The model was trained to predict the acoustic waveform under representative hover conditions as a function of rotational speed, airframe proximity, and observer angle. Training data were acquired in an anechoic chamber on both isolated rotors and rotor-airframe configurations. Acoustic amplitude and phase of the revolution-averaged interaction were predicted, which required up to 25 harmonics to capture the impulse event caused by the blade’s approach and departure from the airframe. Prediction performance showed, on average, that the models could estimate the acoustic amplitude and phase over the relevant harmonics for unseen conditions with 86% and 75% accuracy, respectively, enabling a time domain reconstruction of the waveform for the range of geometric and flow parameters tested.

acoustics↗

The universal robot

Our artifacts are getting smarter, and a loose parallel with the evolution of animal intelligence suggests one future course for them. Computerless industrial machinery exhibits the behavioral flexibility of single-celled organisms. Today's best computer-controlled robots are like the simpler invertebrates. A thousand-fold increase in computer power in the next decade should make possible machines with reptile-like sensory and motor competence. Properly configured, such robots could do in the physical world what personal computers now do in the world of data - act on our behalf as literal-minded slaves. Growing computer power over the next half-century will allow this reptile stage to be surpassed, in stages producing robots that learn like mammals, model their world like primates, and eventually reason like humans. Depending on your point of view, humanity will then have produced a worthy successor, or transcended some of its inherited limitations and so transformed itself into something quite new.

Moravec, Hans↗

An Enabling Platform for Achieving Multiscale Multiphysics Analysis of Multiphase Materials

This paper will address current NASA GRC research efforts utilizing the NASA Multiscale Analysis Tool (NASMAT) which address technical gaps in three of the Vision 2040 key discipline areas. NASMAT is a versatile platform for performing computationally efficient multiscale analyses of heterogeneous materials and is available free through the NASA software catalog. It offers the user flexibility to define an arbitrary number of length scales (levels) where a variety of micromechanics theories can be implemented at each level. Micromechanics theories can be selected to balance accuracy and computational efficiency and range from analytical (Mori-Tanaka) to several semi-analytical (method of cells) formulations. The resulting anisotropic, evolving nonlinear, thermomechanical constitutive model can also be coupled with external software and used to perform multiscale analyses of more complex structures. A recent application to model a complex, three-dimensional woven composite, with a particular emphasis placed on multiscale measurements utilized to enhance the quality of the multiscale analysis will be discussed. Since typical NASMAT analyses can be completed on the order of seconds to minutes, a second example will demonstrate the ability to generate large quantities of data useful for sensitivity analysis, uncertainty quantification, or machine learning applications. Current progress on implementing NASMAT within a multiscale digital thread/digital twin framework will also be addressed. These examples will demonstrate that NASMAT is an enabling platform for achieving the goals in Vision 2040.

Multiscale↗

Remote Sensing of Lineage Functional Types for Modeling and Monitoring Biodiversity

Hyperspectral remote sensing has the potential to continuously scale plant function and plant diversity information from landscape to global extents. Numerous studies have indicated that VSWIR (400-2500 nm) reflectance properties of vegetation capture evolutionarily conserved biochemical, structural, and other functional attributes of plant species. Spectral properties conserved in plants provide the opportunity to both 1) aggregate species into lineages with improved classification accuracy and 2) link those lineages directly to plant traits. Full realization of this goal will enable parameterization of Land Surface Models (LSMs) with remotely sensed information, e.g., canopy nitrogen, and better representations of biodiversity and functional diversity in biogeographic studies. In this study, we use hyperspectral AVIRIS data from the 2013 HyspIRI campaign over the Southern Sierra Nevada, California flight box to investigate the potential for incorporating evolutionary thinking into landcover classification. We link the airborne hyperspectral data with vegetation plot data from roughly 1372 surveys and a phylogeny representing 1361 species. We aggregate species into lineages ranging from species level groups down to similar number of Plant Functional Types as often used in LSMs. We assessed the ability of Random Forest and Partial Least Squares Discriminant Analysis to discriminate across these different phylogenetic scales and determine the optimal number of lineages to classify. Although there are some temporal and spatial differences in our training data, our best approaches achieved moderate classification accuracy (Kappa > 0.65). Given an optimal number of lineages, we explored approaches to improve classifications including machine learning and unmixing approaches. This work suggests that lineage-based methods may be a promising way to leverage the huge amounts of data that will come from high resolution and high return interval hyperspectral data planned for the Surface Biology and Geology mission with sparsely sampled existing ground-based ecological data.

Hyperspectral↗

An Artificial Neural Network Approach to Predict Rotor-Airframe Acoustic Waveforms

A surrogate artificial neural network/machine learning model was developed to predict the acoustic interaction for a fixed-pitch rotor in proximity to a downstream cylindrical airframe typical of small Unmanned Aerial System (sUAS) platforms. The model was trained to predict the acoustic waveform under representative hover conditions as a function of rotational speed, airframe proximity, and observer angle. Training data were acquired in an anechoic chamber on both isolated rotors and rotor-airframe configurations. Acoustic amplitude and phase of the revolution-averaged interaction were predicted, which required up to 25 harmonics to capture the impulse event caused by the blade’s approach and departure from the airframe. Prediction performance showed, on average, that the models could estimate the acoustic amplitude and phase over the relevant harmonics for unseen conditions with 86% and 75% accuracy, respectively, enabling a time domain reconstruction of the waveform for the range of geometric and flow parameters tested.

acoustics↗

Detection of CH 4 Hotspots from NASA’s GEOS Composition Analysis System

Recent research indicates that 8 to 12% of the global oil and gas production methane emissions could be attributed to ultra-emitters, which result in high concentration ‘hotspots’ near point sources. Identifying these emissions in near real time provides useful information to the policy makers and private industry, who are working to reduce their impact. To meet this need, scientists are increasingly analyzing satellite data from the TROPOspheric Monitoring Instrument (TROPOMI) instrument aboard ESA’s Sentinel 5-Precursor mission. While direct analysis of TROPOMI level 2 swath data has been successful in identifying some large emission events, identifying hotpots is challenging because of the imaging noise due to a variety of artifacts and limits in daily coverage. Here we explore possible methodologies to detect methane hotspots using a new, gap-filled, and temporally continuous methane product from NASA’s Goddard Earth Observing System (GEOS) Constituent Data Assimilation System (CoDAS), which assimilates column averaged methane mole fractions from the TROPOMI with capabilities to assimilate other remote sensing measurements. The CoDAS has been expanded from a heritage of stratospheric composition and carbon dioxide assimilation allowing for the support of regional modeling, validation with non-coincident operations, and merging variety of datasets. The current work mainly explores Observing System Simulation Experiments with methane GEOS simulations without assimilation to prepare the groundwork for further experiments with assimilated TROPOMI. First, known hotspots based on the known inventory are identified to demonstrate the capability of the system to point out emission hotspots. In the next step, a variety of machine learning techniques such as Self-Organizing Maps and Deep Learning are explored to automate detection of the plumes. Finally, a few approaches to quantify emissions from the identified hotspots are presented and are evaluated against the inventory. The effort is directed toward a future evaluation of the CoDAS based methane monitoring system’s ability to successfully detect and quantify hotspots.

Nikolay Balashov↗

A Convolutional Neural Network for Enhancement of Multi-Scale Localization in Granular Metallic Representative Unit Cells

A convolutional neural network was used to enhance the localization of strain and stress for a generalized method of cells model of a metallic microstructure. Enhanced shear strains, measured in terms of the linear regression coefficients as a function of ground truth strains, were improved from inaccurate and uncorrelated (slope=0.003, Rsq=0.000) to accurate and well correlated (slope=0.890, Rsq=0.882) relative to ground truth (slope=1.0, Rsq=1.0). In applying the convolutional neural network, a convolutional stride of 1.0 (padding=’same’) was only modestly effective while strides of 2 or 3 were more effective yet at higher cost. Additional convolutional layers were generally more expensive than additional dense layers, often with limited benefit. The accuracy of enhanced localized shear strains and stress is expected to yield benefits for damage progression models, especially in the context of hierarchical multi-scale methods where the generalized method of cells is applied at the intermediate scale.

Machine Learning↗

Regression Analysis of Top of Descent Location for Idle-thrust Descents

In this paper, multiple regression analysis is used to model the top of descent (TOD) location of user-preferred descent trajectories computed by the flight management system (FMS) on over 1000 commercial flights into Melbourne, Australia. The independent variables cruise altitude, final altitude, cruise Mach, descent speed, wind, and engine type were also recorded or computed post-operations. Both first-order and second-order models are considered, where cross-validation, hypothesis testing, and additional analysis are used to compare models. This identifies the models that should give the smallest errors if used to predict TOD location for new data in the future. A model that is linear in TOD altitude, final altitude, descent speed, and wind gives an estimated standard deviation of 3.9 nmi for TOD location given the trajec- tory parameters, which means about 80% of predictions would have error less than 5 nmi in absolute value. This accuracy is better than demonstrated by other ground automation predictions using kinetic models. Furthermore, this approach would enable online learning of the model. Additional data or further knowl- edge of algorithms is necessary to conclude definitively that no second-order terms are appropriate. Possible applications of the linear model are described, including enabling arriving aircraft to fly optimized descents computed by the FMS even in congested airspace. In particular, a model for TOD location that is linear in the independent variables would enable decision support tool human-machine interfaces for which a kinetic approach would be computationally too slow.

trajectory prediction↗

Application of Machine Learning Techniques to Aviation Operations: Promises and Challenges

There is an increasing interest in applying methods based on Machine Learning Techniques (MLT) to problems in aviation operations. The current interest is based on developments in Cloud Computing, the availability of open software and the success of MLT in automation, consumer behavior and finance involving large database. This paper reviews the current-state-of-the art in applying MLT to aviation operations, its promises and challenges. Historically aviation operations have been analyzed using physics-based models and provide information for making operational decisions. This paper compares the methodology used in and issues to be addressed in applying either model-driven or data-driven methods. Some aviation operations involving many decision makers, multiple objectives, poor or unavailable physics-based models and a rich historical database are prime candidates for data-driven methods. The application of MLT to aviation operations falls into three categories: (a) based on the lack of a physics-based model, MLT is the favored approach, (b) marginal difference between regression methods using physics-based models and MLT and (c) better results using a blend of physics-based methods combined with MLT. As always, the best approach depends on the task, the physical understanding of the problem and the quality and quantity of the available data.

Sridhar, Banavar↗

An Ensemble of Bayesian Neural Networks for Exoplanetary Atmospheric Retrieval

Machine learning (ML) is now used in many areas of astrophysics, from detecting exoplanets in Kepler transit signals to removing telescope systematics. Recent work demonstrated the potential of using ML algorithms for atmospheric retrieval by implementing a random forest (RF) to perform retrievals in seconds that are consistent with the traditional, computationally expensive nested-sampling retrieval method. We expand upon their approach by presenting a new ML model, plan-net, based on an ensemble of Bayesian neural networks (BNNs) that yields more accurate inferences than the RF for the same data set of synthetic transmission spectra. We demonstrate that an ensemble provides greater accuracy and more robust uncertainties than a single model. In addition to being the first to use BNNs for atmospheric retrieval, we also introduce a new loss function for BNNs that learns correlations between the model outputs. Importantly, we show that designing ML models to explicitly incorporate domain-specific knowledge both improves performance and provides additional insight by inferring the covariance of the retrieved atmospheric parameters. We apply plan-net to the Hubble Space Telescope Wide Field Camera 3 transmission spectrum for WASP-12b and retrieve an isothermal temperature and water abundance consistent with the literature. We highlight that our method is flexible and can be expanded to higher resolution spectra and a larger number of atmospheric parameters.

Adam D. Cobb↗