SEARCH · Search NASA

Results for “Regression approach”

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 127 records · Page 7

All sources NASA NTRS DOE OSTI

NASA NTRS2021

Challenges to Aboveground Biomass Prediction from Waveform Lidar

Accurate accounting of aboveground biomass density (AGBD) is crucial for carbon cycle, biodiversity, and climate change science. The Global Ecosystem Dynamics Investigation (GEDI), which maps global AGBD from waveform lidar, is the first of a new generation of Earth observation missions designed to improve carbon accounting. This paper explores the possibility that lidar waveforms may not be unique to AGBD—that forest stands with different AGBD may produce highly similar waveforms—and we hypothesize that non-uniqueness may contribute to the large uncertainties in AGBD predictions. Our analysis integrates simulated GEDI waveforms from 428 in situ stem maps with output from an individual-based forest gap model, which we use to generate a database of potential forest stands and simulate GEDI waveforms from those stands. We use this database to predict the AGBD of the 428 in situ stem maps via two different methods: a linear regression from waveform metrics, and a waveform-matching approach that accounts for waveform-AGBD non-uniqueness. We find that some in situ waveforms are more unique to AGBD than others, which notably impacts AGBD prediction uncertainty (7–411 Mg ha−1, average of 167 Mg ha−1). We also find that forest structure complexity may influence the non-uniqueness effect; stands with low structural complexity are more unique to AGBD than more mature stands with multiple cohorts and canopy layers. These findings suggest that the non-uniqueness phenomena may be introduced by the measuring characteristics of waveform lidar in combination with how forest structure manifests at small scales, and we discuss how this complexity may complicate uncertainty estimation in AGBD prediction. This analysis suggests a limit to the accuracy and precision of AGBD predictions from lidar waveforms seen in empirical studies, and underscores the need for further exploration of the relationships between lidar remote sensing measurements, forest structure, and AGBD.

GEDI↗

NASA NTRS2023

Accounting for Dose Uncertainty in Dose-Response Curve Estimation Using Hierarchical Bayes Models

As part of their development of the technology for low-boom supersonic flight, the National Aeronautics and Space Administration (NASA) is planning to conduct a set of supersonic aircraft annoyance surveys in select communities in the United States, to measure public perception of the reduced sonic boom. The relationship between a noise exposure level of a supersonic flight event and the probability of an individual being highly annoyed by the event is quantified by a dose-response curve, which is usually based on logistic regression modeling. Unavoidable amounts of estimation error are expected in the upcoming noise measurements, which can bias the results of the logistic regression if ignored. Hence, the development of an estimation approach that accounts for such error when estimating the dose-response curve is imperative. In this paper, an evaluation study is conducted to assess the impact of measurement error on dose-response curve estimation. For this, hierarchical Bayes models using different specifications are fit to data collected by NASA in 2018, as part of a risk-reduction study of supersonic flights affecting Galveston, Texas. It is observed that the estimated dose-response curve appears sensitive to uncertainty in dose measurements, being subject to attenuation bias.

community response↗

NASA NTRS2012

Comparison of Predictive Modeling Methods of Aircraft Landing Speed

Expected increases in air traffic demand have stimulated the development of air traffic control tools intended to assist the air traffic controller in accurately and precisely spacing aircraft landing at congested airports. Such tools will require an accurate landing-speed prediction to increase throughput while decreasing necessary controller interventions for avoiding separation violations. There are many practical challenges to developing an accurate landing-speed model that has acceptable prediction errors. This paper discusses the development of a near-term implementation, using readily available information, to estimate/model final approach speed from the top of the descent phase of flight to the landing runway. As a first approach, all variables found to contribute directly to the landing-speed prediction model are used to build a multi-regression technique of the response surface equation (RSE). Data obtained from operations of a major airlines for a passenger transport aircraft type to the Dallas/Fort Worth International Airport are used to predict the landing speed. The approach was promising because it decreased the standard deviation of the landing-speed error prediction by at least 18% from the standard deviation of the baseline error, depending on the gust condition at the airport. However, when the number of variables is reduced to the most likely obtainable at other major airports, the RSE model shows little improvement over the existing methods. Consequently, a neural network that relies on a nonlinear regression technique is utilized as an alternative modeling approach. For the reduced number of variables cases, the standard deviation of the neural network models errors represent over 5% reduction compared to the RSE model errors, and at least 10% reduction over the baseline predicted landing-speed error standard deviation. Overall, the constructed models predict the landing-speed more accurately and precisely than the current state-of-the-art.

Diallo, Ousmane H.↗

NASA NTRS2012

Rainbow Fourier Transform

We present a novel technique for remote sensing of cloud droplet size distributions. Polarized reflectances in the scattering angle range between 135deg and 165deg exhibit a sharply defined rainbow structure, the shape of which is determined mostly by single scattering properties of cloud particles, and therefore, can be modeled using the Mie theory. Fitting the observed rainbow with such a model (computed for a parameterized family of particle size distributions) has been used for cloud droplet size retrievals. We discovered that the relationship between the rainbow structures and the corresponding particle size distributions is deeper than it had been commonly understood. In fact, the Mie theory-derived polarized reflectance as a function of reduced scattering angle (in the rainbow angular range) and the (monodisperse) particle radius appears to be a proxy to a kernel of an integral transform (similar to the sine Fourier transform on the positive semi-axis). This approach, called the rainbow Fourier transform (RFT), allows us to accurately retrieve the shape of the droplet size distribution by the application of the corresponding inverse transform to the observed polarized rainbow. While the basis functions of the proxy-transform are not exactly orthogonal in the finite angular range, this procedure needs to be complemented by a simple regression technique, which removes the retrieval artifacts. This non-parametric approach does not require any a priori knowledge of the droplet size distribution functional shape and is computationally fast (no look-up tables, no fitting, computations are the same as for the forward modeling).

Alexandrov, Mikhail D.↗

NASA NTRS1981

A comparison of stratified versus regression estimators

LANDSAT data acquired over an agricultural area along with ground enumeration of the same area are used to obtain crop acreage estimates which are better (as measured in terms of bias and variance) than can be obtained from either data source alone. Two basic approaches considered within the AgRISTARS program are a stratified crop acreage estimator and a regression estimator. A statement of the problem was mathematically formulated and some theorems were proved which relate to the variance of the two estimators. For a particular set of data, the regression and stratified estimators are compared in terms of certain easily computed parameters.

Takacs, H. C.↗

NASA NTRS1999

A Survey and Synthesis of Solar Cycle Prediction Techniques

A number of techniques for predicting solar activity on a solar cycle time scale are identified, described, and tested with historical data. Some techniques, e.g. regression and curve-fitting, work well as solar activity approaches maximum and provide a complete description of future activity, while others, e.g. geomagnetic precursors, work well near solar minimum but only provide an estimate of the amplitude of the cycle. A synthesis of different techniques is shown to provide a more accurate and useful forecast of solar cycle activity levels. A combination of two uncorrelated geomagnetic precursor techniques provides the most accurate prediction for the amplitude of a solar activity cycle at a time well before activity minimum. This Combined Precursor Method gives a smoothed sunspot number maximum of 154+/-21 at the 95% level of confidence for the next cycle maximum. A mathematical function dependent upon the time of cycle initiation and the cycle amplitude then describes the level of solar activity for the next complete cycle. As the time of cycle maximum approaches a better estimate of the cycle activity is obtained by including the fit between previous activity levels and this function. This Combined Solar Cycle Activity Forecast gives, as of January 1999, a smoothed sunspot maximum of 146+/-20 at the 95% level of confidence for the next cycle maximum. The success of the geomagnetic precursors in predicting future solar activity suggests that solar magnetic phenomena at latitudes above the sunspot activity belts are linked to solar activity which occurs many years later in the lower latitudes.

Hathaway, David H.↗

NASA NTRS2000

Status of Cycle 23 Forecasts

A number of techniques for predicting solar activity on a solar cycle time scale are identified, described, and tested with historical data. Some techniques, e.g,, regression and curve-fitting, work well as solar activity approaches maximum and provide a month- by-month description of future activity, while others, e.g., geomagnetic precursors, work well near solar minimum but provide an estimate only of the amplitude of the cycle. A synthesis of different techniques is shown to provide a more accurate and useful forecast of solar cycle activity levels. A combination of two uncorrelated geomagnetic precursor techniques provides the most accurate prediction for the amplitude of a solar activity cycle at a time well before activity minimum. This precursor method gave a smoothed sunspot number maximum of 154+21 for cycle 23. A mathematical function dependent upon the time of cycle initiation and the cycle amplitude then describes the level of solar activity for the complete cycle. As the time of cycle maximum approaches a better estimate of the cycle activity is obtained by including the fit between recent activity levels and this function. This Combined Solar Cycle Activity Forecast now gives a smoothed sunspot maximum of 140+20 for cycle 23. The success of the geomagnetic precursors in predicting future solar activity suggests that solar magnetic phenomena at latitudes above the sunspot activity belts are linked to solar activity, which occurs many years later in the lower latitudes.

Hathaway, D. H.↗

NASA NTRS2012

Formulation and Testing of Paraffin-Based Solid Fuels Containing Energetic Additives for Hybrid Rockets

Many approaches have been considered in an effort to improve the regression rate of solid fuels for hybrid rocket applications. One promising method is to use a fuel with a fast burning rate such as paraffin wax; however, additional performance increases to the fuel regression rate are necessary to make the fuel a viable candidate to replace current launch propulsion systems. The addition of energetic and/or nano-sized particles is one way to increase mass-burning rates of the solid fuels and increase the overall performance of the hybrid rocket motor.1,2 Several paraffin-based fuel grains with various energetic additives (e.g., lithium aluminum hydride (LiAlH4) have been cast in an attempt to improve regression rates. There are two major advantages to introducing LiAlH4 additive into the solid fuel matrix: 1) the increased characteristic velocity, 2) decreased dependency of Isp on oxidizer-to-fuel ratio. The testing and characterization of these solid-fuel grains have shown that continued work is necessary to eliminate unburned/unreacted fuel in downstream sections of the test apparatus.3 Changes to the fuel matrix include higher melting point wax and smaller energetic additive particles. The reduction in particle size through various methods can result in more homogeneous grain structure. The higher melting point wax can serve to reduce the melt-layer thickness, allowing the LiAlH4 particles to react closer to the burning surface, thus increasing the heat feedback rate and fuel regression rate. In addition to the formulation of LiAlH4 and paraffin wax solid-fuel grains, liquid additives of triethylaluminum and diisobutylaluminum hydride will be included in this study. Another promising fuel formulation consideration is to incorporate a small percentage of RDX as an additive to paraffin. A novel casting technique will be used by dissolving RDX in a solvent to crystallize the energetic additive. After dissolving the RDX in a solvent chosen for its compatibility with both paraffin and RDX, the mixture will be combined with the melted paraffin. With the melting point of the paraffin far below the decomposition temperature of the RDX, the solvent will be boiled off, leaving the crystallized RDX embedded in the paraffin. At low percentages of RDX additive and with crystallized RDX surrounded by paraffin, the fuel grains will remain inert, maintaining a key benefit of hybrids in the safety of the solid fuel.

Larson, Daniel B.↗

NASA NTRSDate not supplied

Component Level Regression Testing in a Hierarchical Architecture

The Goddard Earth Observing System (GEOS) is an Earth system model consisting of a large suite of individual model components that can be coupled in a flexible manner to investigate a variety of Earth science issues. Specific GEOS model configurations are composed as a hierarchical collection of components based on the Earth System Modeling Framework (ESMF). Regression testing of GEOS is currently limited to (1) full system tests that are poor at isolating specific defects and (2) a suite of unit tests which have very limited coverage. As part of our approach to improve upon the current testing situation, we have prototyped the capability to perform regression tests on individual GEOS components by leveraging and extending existing checkpoint/restart capabilities. In our implementation, each ESMF component has 3 states: Import (what it needs to run), Export (which it needs to provide to other components), and Internal (the component state proper). By capturing, Import, Export and Internal states for a given component during a ull run of GEOS, a generic driver can then rerun the component offline and compare expected exports with those that have been saved. The hierarchical structure of GEOS introduces an interesting wrinkle when trying to test components that in turn drive interacting child components. To fully isolate a parent component, we use the approach of software mocks, in which the exports of children are also saved during the initial capture run of GEOS. Then when testing the parent component, the children components are replaced by a generic mock component that produces exports from the previously saved data and ensures that that all interdependencies among children components are satisfied.

Thomas Clune↗

NASA NTRSDate not supplied

Component Level Testing in a Hierarchical Architecture

Tom Clune↗

NASA NTRS1975

Comparison of some biased estimation methods (including ordinary subset regression) in the linear model

Ridge, Marquardt's generalized inverse, shrunken, and principal components estimators are discussed in terms of the objectives of point estimation of parameters, estimation of the predictive regression function, and hypothesis testing. It is found that as the normal equations approach singularity, more consideration must be given to estimable functions of the parameters as opposed to estimation of the full parameter vector; that biased estimators all introduce constraints on the parameter space; that adoption of mean squared error as a criterion of goodness should be independent of the degree of singularity; and that ordinary least-squares subset regression is the best overall method.

Sidik, S. M.↗

NASA NTRSDate not supplied

Uncertainty Propagation in Pre-Flight Prediction of Unmanned Aerial Vehicle Separation Violation

Current forecasts on the future of aeronautics suggest an in- creasing number of unmanned aerial vehicles entering the low- altitude airspace in the next decades (FAA, 2018; Kopardekar et al., 2016). Small vehicles for package delivery as well as larger vehicles for urban air mobility will change the airspace drastically, increasing density of operations both in time, i.e. high number of take-off and landings per unit time, and in space, operating in dense urban environment. This scenario poses challenges to the current approach to air traffic control, and large efforts from academia, industry and regulatory bodies are dedicated to the development of new traffic management strategies that leverage higher computing and simulating capabilities available today. In this paper, we propose a simple look-ahead approach to predict potential minimum separation violations at the strategic level, that is before vehicles start flying, depending on the predefined 4D trajectories and uncertainty affecting the wind acting along those routes. The wind field is extracted from the NOAA North America Mesoscale Forecast System and interpolated using Gaussian process regression, while uncertainty affecting the expected cruise airspeed is propagated through error intervals. The approach allows the prediction of aircraft separation as a function of time, highlighting potential safety violations that would go undetected if uncertainty affecting the expected 4D trajectories is not considered. The paper will also discuss issues related to accuracy and scalability of the approach to multiple vehicle operations.

Trajectory Prediction↗

NASA NTRSDate not supplied

A Novel Machine Learning Method for Surface PM2.5 Estimations from Geostationary Satellites

Particulate matter (PM) with a diameter of less or equal to 2.5 μm, known as PM , affects human health as it penetrates the respiratory system. The Environmental Protection Agency (EPA) measures the atmospheric concentration of PM using air quality monitors stationed throughout the Continental United States (CONUS). Such measurements are points on a spatial domain and therefore, might not be representative of the air quality at nearby areas considering that the composition of the atmosphere is highly variable from place to place. Satellite based AOD permits a spatially uniform means of estimating PM and new geostationary satellites provide high temporal and spatial resolution estimation of AOD. However, the concentration of PM is non-linearly dependent on other atmospheric parameters that include relative humidity, temperature, and height of the planetary boundary layer. This information may be estimated at similar spatial and temporal resolutions as AOD from numerical modeling such as from the National Oceanic and Atmospheric Administration’s (NOAA) High Resolution Rapid Refresh (HRRR) model which resolves near real-time atmospheric conditions over the CONUS. The estimation of PM concentration is a multi-parametric problem that considers the effect of temporal dependencies among the different parameters. Deep learning approaches are appropriate for such complex estimation problems as they intrinsically capture relations among multiple non-linear parameters. This study compares deep-learning methods to traditional regression analysis to demonstrate the capabilities of these methods in predicting PM2.5 concentrations. Additionally, a novel ensemble learning approach is employed to identify scientific processes that could further improve the estimation of PM concentration. Utilizing Long Short-Term Memory (LSTM) neural networks, which are suitable for multivariate time series estimation problems as they are capable of learning long-term dependencies, individual models are created for each EPA station and trained on the aforementioned dataset collocated over each station. Individual station models are merged if the model's performance is improved by reducing the root mean squared error (RMSE) metric. This ensemble training method ultimately reduces the RMSE value. Evaluation of these results provide insights into physical processes and related observable parameters that may contribute to PM concentrations. Identified parameters evaluated to be statistically different between the merged and unmerged models are expected to improve overall performance. These new parameters are then utilized for reevaluation of the deep learning methods with an extreme gradient boosting model with an RMSE of 5.5 providing the best results.

George Priftis↗

NASA NTRS1983

Advanced microwave soil moisture studies

Comparisons of low level L-band brightness temperature (TB) and thermal infrared (TIR) data as well as the following data sets: soil map and land cover data; direct soil moisture measurement; and a computer generated contour map were statistically evaluated using regression analysis and linear discriminant analysis. Regression analysis of footprint data shows that statistical groupings of ground variables (soil features and land cover) hold promise for qualitative assessment of soil moisture and for reducing variance within the sampling space. Dry conditions appear to be more conductive to producing meaningful statistics than wet conditions. Regression analysis using field averaged TB and TIR data did not approach the higher sq R values obtained using within-field variations. The linear discriminant analysis indicates some capacity to distinguish categories with the results being somewhat better on a field basis than a footprint basis.

Dalsted, K. J.↗

NASA NTRS2019

Using Deep Learning to Automate Inference of Meteoroid Pre-Entry Properties

Properly assessing the asteroid threat depends on the knowledge of asteroid pre-entry parameters, such as size, velocity, mass, density, and strength. Although a vast number of possible bodies to study exist, such characterization of asteroid populations is currently limited by substantial costs associated with space rendezvous missions and rare meteorite findings. As asteroids fragment, ablate, and decelerate in the atmosphere, they emit light detectable by ground-based and space-borne instruments. Earth’s atmosphere, thus, becomes an accessible laboratory that enables impactor risk assessments by facilitating inference of the pre-entry parameters. These asteroid pre-entry conditions are typically deduced by modeling the entry and breakup physics that best reproduce the observed light or energy deposition curve. However, this process requires extensive manual trial-and-error of uncertain modeling parameters. Automating meteor modeling and inference would improve property distributions used in risk assessments and enable population characterization as more light curves become more readily available through the presence of space assets and ground-based camera networks. We previously developed a genetic algorithm to automate meteor modeling by using the fragment-cloud model (FCM) to search for the values of the FCM input parameters (e.g., diameter) that generate energy deposition profiles that match the observed one. Now, we apply deep learning to infer asteroid diameter, velocity, and density from observed energy deposition curves. We trained and tested our neural network models with synthetic energy deposition curves modeled using the FCM rubble pile implementation. We present an application of a 1D convolutional neural network and compare its performance to other attempted regressors and machine learning techniques, such as a fully connected neural network and Random Forest regression, to demonstrate its capabilities. We validate our model weights and approach using the Chelyabinsk, Tagish Lake, Benešov, Košice, and Lost City meteors.

Tarano, Ana Maria↗

NASA NTRS1995

Hybrid fuel formulation and technology development

The objective was to develop an improved hybrid fuel with higher regression rate, a regression rate expression exponent close to 0.5, lower cost, and higher density. The approach was to formulate candidate fuels based on promising concepts, perform thermomechanical analyses to select the most promising candidates, develop laboratory processes to fabricate fuel grains as needed, fabricate fuel grains and test in a small lab-scale motor, select the best candidate, and then scale up and validate performance in a 2500 lbf scale, 11-inch diameter motor. The characteristics of a high performance fuel have been verified in 11-inch motor testing. The advanced fuel exhibits a 15% increase in density over an all hydrocarbon formulation accompanied by a 50% increase in regression rate, which when multiplied by the increase in density yields a 70% increase in fuel mass flow rate; has a significantly lower oxidizer-to-fuel (O/F) ratio requirement at 1.5; has a significantly decreased axial regression rate variation making for more uniform propellant flow throughout motor operation; is very clean burning; extinguishes cleanly and quickly; and burns with a high combustion efficiency.

Dean, D. L.↗

NASA NTRS1982

Least-squares sequential parameter and state estimation for large space structures

This paper presents the formulation of simultaneous state and parameter estimation problems for flexible structures in terms of least-squares minimization problems. The approach combines an on-line order determination algorithm, with least-squares algorithms for finding estimates of modal approximation functions, modal amplitudes, and modal parameters. The approach combines previous results on separable nonlinear least squares estimation with a regression analysis formulation of the state estimation problem. The technique makes use of sequential Householder transformations. This allows for sequential accumulation of matrices required during the identification process. The technique is used to identify the modal prameters of a flexible beam.

Thau, F. E.↗

NASA NTRS1998

Deterministic Multiaxial Creep and Creep Rupture Enhancements for CARES/Creep Integrated Design Code

High temperature and long duration applications of monolithic ceramics can place their failure mode in the creep rupture regime. A previous model advanced by the authors described a methodology by which the creep rupture life of a loaded component can be predicted. That model was based on the life fraction damage accumulation rule in association with the modified Monkman-Grant creep rupture criterion. However, that model did not take into account the deteriorating state of the material due to creep damage (e.g., cavitation) as time elapsed. In addition, the material creep parameters used in that life prediction methodology, were based on uniaxial creep curves displaying primary and secondary creep behavior, with no tertiary regime. The objective of this paper is to present a creep life prediction methodology based on a modified form of the Kachanov-Rabotnov continuum damage mechanics (CDM) theory. In this theory, the uniaxial creep rate is described in terms of sum, temperature, time, and the current state of material damage. This scalar damage state parameter is basically an abstract measure of the current state of material damage due to creep deformation. The damage rate is assumed to vary with stress, temperature, time, and the current state of damage itself. Multiaxial creep and creep rupture formulations of the CDM approach are presented in this paper. Parameter estimation methodologies based on nonlinear regression analysis are also described for both, isothermal constant stress states and anisothermal variable stress conditions This creep life prediction methodology was preliminarily added to the integrated design code CARES/Creep (Ceramics Analysis and Reliability Evaluation of Structures/Creep), which is a postprocessor program to commercially available finite element analysis (FEA) packages. Two examples, showing comparisons between experimental and predicted creep lives of ceramic specimens, are used to demonstrate the viability of Ns methodology and the CARES/Creep program.

Jadaan, Osama M.↗