Search NASA⌕ Search

SEARCH · Search NASA

Results for “statistical processing methods”

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 325 records · Page 18

Acquisition of quantitative physiological data and computerized image reconstruction using a single scan TV system

Single scan operation of television X-ray fluoroscopic systems allow both analog and digital reconstruction of tomographic sections from single plan images. This type of system combined with a minimum of statistical processing showed excellent capabilities for delineating small changes in differential X-ray attenuation. Patient dose reduction is significant when compared to normal operation or film recording. Flat screen, low light level systems were both rugged and light in weight, making them applicable for a variety of special purposes. Three dimensional information was available from the tomographic methods and the recorded data was sufficient when used with appropriate computer display devices to give representative 3D images.

Baily, N. A.↗

Water Vapor Transport Over the Tropical Oceans During ENSO as Diagnosed from TRMM and SSM/I Data

Traditionally, large-scale water vapor transport [div Q] has been derived directly from circulation statistics in which transport processes are often depicted by mean and eddy motions. Thus detailed and accurate calculations of moisture transport terms over the globe are required. Notably, the lack of systematically spaced conventional measurements of meteorological variables over oceans has hindered understanding of the distribution and transport of water vapor. This motivates the use of indirect calculation methods in which horizontal divergence of water vapor is balanced by the evaporation minus precipitation, assuming the rate of changes of precipitable water and condensates is small over a sufficiently long time period. In order to obtain the water vapor transport, we need evaporation rate minus precipitation (E-P). Focussing on the differences in water vapor transport between El Nino and La Nina periods and their influences on atmospheric circulations, we study January, February, and March of 1998 and 1999 periods which represent El Nino and La Nina respectively. SSM/I-derived precipitation and evaporation rate from SSM/I wind and total precipitable water, in conjunction with NCEP SST and surface air temperature, are used for the calculation of the transport potential function. For the retrieval of evaporation we use a stability-dependent aerodynamic bulk scheme developed by Chou (1993). It was tested against aircraft covariance fluxes measured during cold air outbreaks over the North Atlantic Ocean. Chou et al. (1997) reported that the SSM/I retrieved latent heat flux over the western Pacific warm pool area were found to be comparable with daily mean fluxes of a ship measurements during TOGA/COARE.

Robertson, Franklin R.↗

Exploring the Connection Between Sampling Problems in Bayesian Inference and Statistical Mechanics

The Bayesian and statistical mechanical communities often share the same objective in their work - estimating and integrating probability distribution functions (pdfs) describing stochastic systems, models or processes. Frequently, these pdfs are complex functions of random variables exhibiting multiple, well separated local minima. Conventional strategies for sampling such pdfs are inefficient, sometimes leading to an apparent non-ergodic behavior. Several recently developed techniques for handling this problem have been successfully applied in statistical mechanics. In the multicanonical and Wang-Landau Monte Carlo (MC) methods, the correct pdfs are recovered from uniform sampling of the parameter space by iteratively establishing proper weighting factors connecting these distributions. Trivial generalizations allow for sampling from any chosen pdf. The closely related transition matrix method relies on estimating transition probabilities between different states. All these methods proved to generate estimates of pdfs with high statistical accuracy. In another MC technique, parallel tempering, several random walks, each corresponding to a different value of a parameter (e.g. "temperature"), are generated and occasionally exchanged using the Metropolis criterion. This method can be considered as a statistically correct version of simulated annealing. An alternative approach is to represent the set of independent variables as a Hamiltonian system. Considerab!e progress has been made in understanding how to ensure that the system obeys the equipartition theorem or, equivalently, that coupling between the variables is correctly described. Then a host of techniques developed for dynamical systems can be used. Among them, probably the most powerful is the Adaptive Biasing Force method, in which thermodynamic integration and biased sampling are combined to yield very efficient estimates of pdfs. The third class of methods deals with transitions between states described by rate constants. These problems are isomorphic with chemical kinetics problems. Recently, several efficient techniques for this purpose have been developed based on the approach originally proposed by Gillespie. Although the utility of the techniques mentioned above for Bayesian problems has not been determined, further research along these lines is warranted

Pohorille, Andrew↗

Parameterization of Vertical Cloud Distribution from C3M and MERRA Data Using ML Method

Clouds play a key role in regulating the hydrological cycle and the Earth's radiative energy budget. However, global climate models (GCMs) with a horizontal grid spacing on the order of 100 km have limitations in representing sub-grid cloud dynamics with spatial scales on the order of 1 km, leading to potential uncertainties in cloud radiative feedback on the global scale. In our research, we will leverage the capabilities of Deep Machine Learning (DML) methods to construct parameterizations of sub-grid volumetric cloud fraction (VCF), which is the frequency of occurrence on a grid volume accumulated in the horizontal and vertical directions. Our investigation delves into the intricate relationship between VCF obtained from the NASA CALIPSO-CloudSat-CERES-MODIS (CCCM) satellite observation data and 3-D MERRA-2 reanalysis meteorological profiling data (e.g., wind, relative humidity, temperature). Through a comprehensive one-year data training utilizing the Sequence to Sequence DML method, we have successfully disentangled the complicated cloud formation dynamics across diverse meteorological conditions through a day-to-day analysis framework. Preliminary findings reveal promising statistical agreements in geographical and vertical distributions and seasonal variations of volumetric cloud fraction between ML prediction and satellite measurements. These results underscore the aptitude of our DML model to discern underlying cloud physical processes and accurately represent sub-grid cloud formation dynamics. Additionally, we have also employed trained neural network to analyze uncertainties arising from errors in meteorological data, further enhancing the robustness of our VCF parameterization.

Shan Zeng↗

A Method for Validating Causal Diagrams of Human Health Risk in Space Flight

The complexity of cause-and-effect relationships between spaceflight hazards and resulting health conditions clouds understanding of the totality of human system risk in space. In response, NASA has introduced Directed Acyclic Graphs (causal diagrams) into the human systems risk management process. These diagrams allow for a common understanding of the mechanisms that lead from unique hazards of spaceflight to the health outcomes important to agencies and astronauts. However, the paucity of available biomedical data from spaceflight creates a need for methods of validating causal models that can accommodate data from spaceflight model analogs. Here we outline one approach utilizing open-access rodent bone datasets from the Ames Life Sciences Data Archive. The properties of directed acyclic graphs themselves can provide an epistemological and statistical framework for validation of a priori causal representations of human system risk in space flight. The assumed causal connections on the graph creates sets of logical implications: variables that – if the causal diagram is correct – should be correlated, as well as sets that should be conditionally independent. By testing these implied correlations and conditional independencies both statistically and heuristically, we can provide evidence for or against specific causal pathways on the causal diagram. In addition to validation of expert-generated causal diagrams, machine learning techniques can learn the most likely structure of a causal diagram from a given dataset. Comparison with and reconciliation between machine-learned causal diagrams and expert-generated diagrams is another technique for challenging assumptions and improving our understanding of causal mechanisms. Accurately representing complex causation is essential to systemic understanding of human health risks in space travel. Having a robust system of validating causal diagrams helps us arrive at more accurate representations of causal systems. This process will be integral to developing the countermeasures necessary for extended exploration of the moon and Mars.

Robert Reynolds↗

Experiments on Nucleation in Different Flow Regimes

The vast majority of metallic engineering materials are solidified from the liquid phase. Understanding the solidification process is essential to control microstructure, which in turn, determines the properties of materials. The genesis of solidification is nucleation, where the first stable solid forms from the liquid phase. Nucleation kinetics determine the degree of undercooling and phase selection. As such, it is important to understand nucleation phenomena in order to control solidification or glass formation in metals and alloys. Early experiments in nucleation kinetics were accomplished by droplet dispersion methods. Dilatometry was used by Turnbull and others, and more recently differential thermal analysis and differential scanning calorimetry have been used for kinetic studies. These techniques have enjoyed success; however, there are difficulties with these experiments. Since materials are dispersed in a medium, the character of the emulsion/metal interface affects the nucleation behavior. Statistics are derived from the large number of particles observed in a single experiment, but dispersions have a finite size distribution which adds to the uncertainty of the kinetic determinations. Even though temperature can be controlled quite well before the onset of nucleation, the release of the latent heat of fusion during nucleation of particles complicates the assumption of isothermality during these experiments. Containerless processing has enabled another approach to the study of nucleation kinetics. With levitation techniques it is possible to undercool one sample to nucleation repeatedly in a controlled manner, such that the statistics of the nucleation process can be derived from multiple experiments on a single sample. The authors have fully developed the analysis of nucleation experiments on single samples following the suggestions of Skripov. The advantage of these experiments is that the samples are directly observable. The nucleation temperature can be measured by noncontact optical pyrometry, the mass of the sample is known, and post processing analysis can be conducted on the sample. The disadvantages are that temperature measurement must have exceptionally high precision, and it is not possible to isolate specific heterogeneous sites as in droplet dispersions.

Bayuzick, R. J.↗

Data Analysis with Graphical Models: Software Tools

Probabilistic graphical models (directed and undirected Markov fields, and combined in chain graphs) are used widely in expert systems, image processing and other areas as a framework for representing and reasoning with probabilities. They come with corresponding algorithms for performing probabilistic inference. This paper discusses an extension to these models by Spiegelhalter and Gilks, plates, used to graphically model the notion of a sample. This offers a graphical specification language for representing data analysis problems. When combined with general methods for statistical inference, this also offers a unifying framework for prototyping and/or generating data analysis algorithms from graphical specifications. This paper outlines the framework and then presents some basic tools for the task: a graphical version of the Pitman-Koopman Theorem for the exponential family, problem decomposition, and the calculation of exact Bayes factors. Other tools already developed, such as automatic differentiation, Gibbs sampling, and use of the EM algorithm, make this a broad basis for the generation of data analysis software.

Buntine, Wray L.↗

Earth Observing System Covariance Realism

The purpose of covariance realism is to properly size a primary object's covariance in order to add validity to the calculation of the probability of collision. The covariance realism technique in this paper consists of three parts: collection/calculation of definitive state estimates through orbit determination, calculation of covariance realism test statistics at each covariance propagation point, and proper assessment of those test statistics. An empirical cumulative distribution function (ECDF) Goodness-of-Fit (GOF) method is employed to determine if a covariance is properly sized by comparing the empirical distribution of Mahalanobis distance calculations to the hypothesized parent 3-DoF chi-squared distribution. To realistically size a covariance for collision probability calculations, this study uses a state noise compensation algorithm that adds process noise to the definitive epoch covariance to account for uncertainty in the force model. Process noise is added until the GOF tests pass a group significance level threshold. The results of this study indicate that when outliers attributed to persistently high or extreme levels of solar activity are removed, the aforementioned covariance realism compensation method produces a tuned covariance with up to 80 to 90% of the covariance propagation timespan passing (against a 60% minimum passing threshold) the GOF tests-a quite satisfactory and useful result.

Realism↗

Use of observational and model-derived fields and regime model output statistics in mesoscale forecasting

Various empirical and statistical weather-forecasting studies which utilize stratification by weather regime are described. Objective classification was used to determine weather regime in some studies. In other cases the weather pattern was determined on the basis of a parameter representing the physical and dynamical processes relevant to the anticipated mesoscale phenomena, such as low level moisture convergence and convective precipitation, or the Froude number and the occurrence of cold-air damming. For mesoscale phenomena already in existence, new forecasting techniques were developed. The use of cloud models in operational forecasting is discussed. Models to calculate the spatial scales of forcings and resultant response for mesoscale systems are presented. The use of these models to represent the climatologically most prevalent systems, and to perform case-by-case simulations is reviewed. Operational implementation of mesoscale data into weather forecasts, using both actual simulation output and method-output statistics is discussed.

Forbes, G. S.↗

Experiments on Nucleation in Different Flow Regimes

The vast majority of metallic engineering materials are solidified from the liquid phase. Understanding the solidification process is essential to control microstructure, which in turn, determines the properties of materials. The genesis of solidification is nucleation, where the first stable solid forms from the liquid phase. Nucleation kinetics determine the degree of undercooling and phase selection. As such, it is important to understand nucleation phenomena in order to control solidification or glass formation in metals and alloys. Early experiments in nucleation kinetics were accomplished by droplet dispersion methods [1-6]. Dilitometry was used by Turnbull and others, and more recently differential thermal analysis and differential scanning calorimetry have been used for kinetic studies. These techniques have enjoyed success; however, there are difficulties with these experiments. Since materials are dispersed in a medium, the character of the emulsion/metal interface affects the nucleation behavior. Statistics are derived from the large number of particles observed in a single experiment, but dispersions have a finite size distribution which adds to the uncertainty of the kinetic determinations. Even though temperature can be controlled quite well before the onset of nucleation, the release of the latent heat of fusion during nucleation of particles complicates the assumption of isothermality during these experiments. Containerless processing has enabled another approach to the study of nucleation kinetics [7]. With levitation techniques it is possible to undercool one sample to nucleation repeatedly in a controlled manner, such that the statistics of the nucleation process can be derived from multiple experiments on a single sample. The authors have fully developed the analysis of nucleation experiments on single samples following the suggestions of Skripov [8]. The advantage of these experiments is that the samples are directly observable. The nucleation temperature can be measured by noncontact optical pyrometry, the mass of the sample is known, and post processing analysis can be conducted on the sample. The disadvantages are that temperature measurement must have exceptionally high precision, and it is not possible to isolate specific heterogeneous sites as in droplet dispersions.

Bayuzick, Robert J.↗

Scatter in Carbon/Silicon Carbide (C/SiC) Composites Quantified

Carbon-fiber-reinforced silicon carbide matrix (C/SiC) composites processed by chemical vapor infiltration are candidate materials for aerospace thermal structures. Carbon fibers can retain properties at very high temperatures, but they are known to have poor oxidation resistance in adverse, high-temperature environments. Nevertheless, the combination of CVI-SiC matrix with higher stiffness and oxidation resistance, the interfacial coating, and additional surface-seal coating provides the necessary protection to the carbon fibers, and makes the material viable for high-temperature space applications operating under harsh environments. Furthermore, C/SiC composites, like other ceramic matrix composites (CMCs), exhibit graceful non-catastrophic failure because of various inherent energy dissipating mechanisms. The material exhibits nonlinearity in deformation even at very low stress levels. This is the result of the severe matrix microcracking present in the as processed composite because of large differences between the coefficients of thermal expansion of the fiber and the matrix. Utilization of these advanced composites in next generation space vehicles will require innovative structural configurations, updated materials, and refined analyses. Structural safety issues for these vehicles are in direct competition with performance and cost. One would have to quantify the uncertainties associated with the design using formal probabilistic methods. Specifically four fundamental aspects on which analyses are based-- (1) loading conditions, (2) material behavior, (3) geometrical configurations, and (4) structural connections between the composite components and baseline structure--are stochastic in nature. A direct way to formally account for uncertainties is to develop probabilistic structural analysis methods where all participating variables are described by appropriate probability density functions. The present work, however, focuses on analyzing the stochastic material behavior of these advanced composites using formal probabilistic analysis methods. Often, some of the desirable property characteristics that allow composites to offer advantages over conventional structural materials (like tailoring of composite properties) and the complexity are in fact responsible for their greater statistical variability and the requirements for more characterization tests. Composite properties are anisotropic as well, having different properties in different directions. This means that characterization of a property such as stiffness--which will vary greatly depending on the orientation of the fiber relative to the direction of the testing--must be repeated for several different directions and loading conditions. The fabrication process for composites also introduces statistical variations in properties and geometry. A composite part is produced in a number of steps, each of which introduces statistical variability. The matrix is usually produced from a combination of raw materials; and the fiber, which has its own set of properties, is often coated or surface treated, introducing yet another source of variability.

Murthy, Pappu L. N.↗

Containerless Processing of Undercooled Melts

The physical mechanisms controlling liquid undercooling were studied for application to solidification processing methods. The undercooling potential of containerless processing was assessed on droplet samples of high melting temperature metals drop tube and drop tower facilities and in a laboratory-scale apparatus. New insight into nucleation and crystal growth will be obtained in undercooled liquids of high melting temperature iron and nickel-base systems. The processing parameters include melt superheat, droplet size and particle statistics and droplet surface coating. The solidification behavior is determined by thermal analysis and by structural and metallographic characterization.

Perepezko, J. H.↗

Improved silicon carbide for advanced heat engines

The development of high strength, high reliability silicon carbide parts with complex shapes suitable for use in advanced heat engines is studied. Injection molding was the forming method selected for the program because it is capable of forming complex parts adaptable for mass production on an economically sound basis. The goals were to reach a Weibull characteristic strength of 550 MPa (80 ksi) and a Weibull modulus of 16 for bars tested in four-point loading. Statistically designed experiments were performed throughout the program and a fluid mixing process employing an attritor mixer was developed. Compositional improvements in the amounts and sources of boron and carbon used and a pressureless sintering cycle were developed which provided samples of about 99 percent of theoretical density. Strengths were found to improve significantly by annealing in air. Strengths in excess of 550 MPa (80 ksi) with Weibull modulus of about 9 were obtained. Further improvements in Weibull modulus to about 16 were realized by proof testing. This is an increase of 86 percent in strength and 100 percent in Weibull modulus over the baseline data generated at the beginning of the program. Molding yields were improved and flaw distributions were observed to follow a Poisson process. Magic angle spinning nuclear magnetic resonance spectra were found to be useful in characterizing the SiC powder and the sintered samples. Turbocharger rotors were molded and examined as an indication of the moldability of the mixes which were developed in this program.

Whalen, Thomas J.↗

A Ground Flash Fraction Retrieval Algorithm for GLM

A Bayesian inversion method is introduced for retrieving the fraction of ground flashes in a set of N lightning observed by a satellite lightning imager (such as the Geostationary Lightning Mapper, GLM). An exponential model is applied as a physically reasonable constraint to describe the measured lightning optical parameter distributions. Population statistics (i.e., the mean and variance) are invoked to add additional constraints to the retrieval process. The Maximum A Posteriori (MAP) solution is employed. The approach is tested by performing simulated retrievals, and retrieval error statistics are provided. The approach is feasible for N greater than 2000, and retrieval errors decrease as N is increased.

Koshak, William J.↗

Use of Design of Experiments in Determining Neural Network Architectures for Loss of Control Detection

We describe empirical methods for selecting a neural network architecture to implement belief state inference on generic commercial transport aircraft. We highlight a case study on the planning, execution, and analysis of a set of experiments to determine the configurations of a conditional variational autoencoder (CVAE). Our main contribution is the application of a structured method that can be used for machine learning in many aerospace applications. This method optimizes the structure and training parameters of a neural network for belief state inference, using Design of Experiments (DOE) statistical methodologies. The motivation for this specific DOE analysis was to identify the appropriate hyperparameters for measuring the CVAE reconstruction probability and latent space, such that the measurements can be used to infer qualitative state changes for the aircraft. We demonstrate that this process yields information about a trained neural network’s utility for this specific application, along with a quantifiable range of certainty. We execute 84 experiments using loss-of-control flight maneuver data from the NASA T-2 aircraft, demonstrating that this empirical process allows us to construct cheap and simple models with specific attributes amenable to belief state inference in aerospace applications.

Loss of Control↗

Use of Design of Experiments in Determining Neural Network Architectures for Loss of Control Detection

We describe empirical methods for selecting a neural network architecture to implement belief state inference on generic commercial transport aircraft. We highlight a case study on the planning, execution, and analysis of a set of experiments to determine the configurations of a conditional variational autoencoder (CVAE). Our main contribution is the application of a structured method that can be used for machine learning in many aerospace applications. This method optimizes the structure and training parameters of a neural network for belief state inference, using Design of Experiments (DOE) statistical methodologies. The motivation for this specific DOE analysis was to identify the appropriate hyperparameters for measuring the CVAE reconstruction probability and latent space, such that the measurements can be used to infer qualitative state changes for the aircraft. We demonstrate that this process yields information about a trained neural network’s utility for this specific application, along with a quantifiable range of certainty. We execute 84 experiments using loss-of-control flight maneuver data from the NASA T 2 aircraft, demonstrating that this empirical process allows us to construct cheap and simple models with specific attributes amenable to belief state inference in aerospace applications.

Loss of Control↗

Use of Design of Experiments in Determining Neural Network Architectures for Loss of Control Detection

Abstract—We describe empirical methods for selecting a neural network architecture to implement belief state inference on generic commercial transport aircraft. We highlight a case study on the planning, execution, and analysis of a set of experiments to determine the configurations of a conditional variational autoencoder (CVAE). Our main contribution is the application of a structured method that can be used for machine learning in many aerospace applications. This method optimizes the structure and training parameters of a neural network for belief state inference, using Design of Experiments (DOE) statistical methodologies. The motivation for this specific DOE analysis was to identify the appropriate hyperparameters for measuring the CVAE reconstruction probability and latent space, such that the measurements can be used to infer qualitative state changes for the aircraft. We demonstrate that this process yields information about a trained neural network’s utility for this specific application, along with a quantifiable range of certainty. We execute 84 experiments using loss-of-control flight maneuver data from the NASA T-2 aircraft, demonstrating that this empirical process allows us to construct cheap and simple models with specific attributes amenable to belief state inference in aerospace applications.

neural networks↗

Predicting Time Series Outputs and Time-to-Failure for an Aircraft Controller Using Bayesian Modeling

Safety of unmanned aerial systems (UAS) is paramount, but the large number of dynamically changing controller parameters makes it hard to determine if the system is currently stable, and the time before loss of control if not. We propose a hierarchical statistical model using Treed Gaussian Processes to predict (i) whether a flight will be stable (success) or become unstable (failure), (ii) the time-to-failure if unstable, and (iii) time series outputs for flight variables. We first classify the current flight input into success or failure types, and then use separate models for each class to predict the time-to-failure and time series outputs. As different inputs may cause failures at different times, we have to model variable length output curves. We use a basis representation for curves and learn the mappings from input to basis coefficients. We demonstrate the effectiveness of our prediction methods on a NASA neuro-adaptive flight control system.

Statistics↗