Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “analysis and statistical 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 937 records · Page 52

Determination of proton PDF uncertainties with Markov chain Monte Carlo

We present an analysis of parton distribution functions (PDFs) of the proton using Markov chain Monte Carlo (MCMC) methods. The MCMC approach naturally implements Bayes’ theorem and, thus, provides a means to directly sample the underlying probability distribution—in this case, the probability distribution of the PDF parameters. This allows for a straightforward propagation of the resulting uncertainties into any PDF-dependent observable, preserving their simple probabilistic interpretation. In our analysis we include a broad set of deep inelastic scattering data from HERA, BCDMS and NMC experiments along with the Drell-Yan, 𝑊 and 𝑍 boson data from LHC and Tevatron experiments, which combined with theoretical calculations at next-to-next-to-leading order in QCD allow for realistic determination of PDFs. The main focus of this analysis is to explore alternative methods for PDF uncertainty estimation that are more firmly grounded in statistical principles. We show that the flexibility of the Bayes framework, allowing one, e.g., to account for non-Gaussianity or inconsistencies of datasets, is crucial to extract realistic uncertainties when such assumptions are not fulfilled. We also demonstrate that MCMC allows one to determine the Δ⁢𝜒 2 value corresponding to a given confidence level in the sample, which can, in turn, be used as a statistically well-founded tolerance criterion used in the Hessian method, thus addressing one of its main long-standing drawbacks.

Risse, Peter Clemens [Universität Münster (Germany↗

TU-AB-303-05: Clinical Guidelines for Determining When An Adaptive Replan May Be Warranted for Head and Neck Patients

Purpose: Tools are now available to perform daily dose assessment in radiotherapy, however, guidance is lacking as to when to replan to limit increase in normal tissue dose. This work performs statistical analysis to provide guidance for when adaptive replanning may be necessary for head/neck (HN) patients. Methods: Planning CT and daily kVCBCT images for 50 HN patients treated with VMAT were retrospectively evaluated. Twelve of 50 patients were replanned due to anatomical changes noted over their RT course. Daily dose assessment was performed to calculate the variation between the planned and delivered dose for the 38 patients not replanned and the patients replanned using their delivered plan. In addition, for the replanned patients, the dose that would have been delivered if the plan was not modified was also quantified. Deviations in dose were analyzed before and after replanning, the daily variations in patients who were not replanned assessed, and the predictive power of the deviation after 1, 5, and 15 fractions determined. Results: Dose deviations were significantly reduced following replanning, compared to if the original plan would have been delivered for the entire course. Early deviations were significantly correlated with total deviations (p<0.01). Using the criteria that a 10% increase in the final delivered dose indicates a replan may be needed earlier in the treatment course, the following guidelines can be made with a 90% specificity after the first 5 fractions: deviations of 7% in the mean dose to the inferior constrictors and 5% in the mean dose to the parotid glands and submandibular glands. No significant dose deviations were observed in any patients for the CTV -70Gy (max deviation 4%). Conclusions: A 5–7% increase in mean dose to normal tissues within the first 5 fractions strongly correlate to an overall deviatios in the delivered dose for HN patients. This work is funded in part by NIH 2P01CA059827-16.

60 APPLIED LIFE SCIENCES↗

NIBBS-Search for Fast and Accurate Prediction of Phenotype-Biased Metabolic Systems

Understanding of genotype-phenotype associations is important not only for furthering our knowledge on internal cellular processes, but also essential for providing the foundation necessary for genetic engineering of microorganisms for industrial use (e.g., production of bioenergy or biofuels). However, genotype-phenotype associations alone do not provide enough information to alter an organism’s genome to either suppress or exhibit a phenotype. It is important to look at the phenotype-related genes in the context of the genome-scale network to understand how the genes interact with other genes in the organism. Identification of metabolic subsystems involved in the expression of the phenotype is one way of placing the phenotype-related genes in the context of the entire network. A metabolic system refers to a metabolic network subgraph; nodes are compounds and edges labels are the enzymes that catalyze the reaction. The metabolic subsystem could be part of a single metabolic pathway or span parts of multiple pathways. Arguably, comparative genome-scale metabolic network analysis is a promising strategy to identify these phenotype-related metabolic subsystems. Network Instance-Based Biased Subgraph Search (NIBBS) is a graph-theoretic method for genome-scale metabolic network comparative analysis that can identify metabolic systems that are statistically biased toward phenotypeexpressing organismal networks. We set up experiments with target phenotypes like hydrogen production, TCA expression, and acid-tolerance. We show via extensive literature search that some of the resulting metabolic subsystems are indeed phenotype-related and formulate hypotheses for other systems in terms of their role in phenotype expression. NIBBS is also orders of magnitude faster than MULE, one of the most efficient maximal frequent subgraph mining algorithms that could be adjusted for this problem. Also, the set of phenotype-biased metabolic systems output by NIBBS comes very close to the set of phenotype-biased subgraphs output by an exact maximally-biased subgraph enumeration algorithm ( MBS-Enum ). The code (NIBBS and the module to visualize the identified subsystems) is available at http:// freescience.org/cs/NIBBS.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Superheater Corrosion In Biomass Boilers: Today's Science and Technology

This report broadens a previous review of published literature on corrosion of recovery boiler superheater tube materials to consider the performance of candidate materials at temperatures near the deposit melting temperature in advanced boilers firing coal, wood-based fuels, and waste materials as well as in gas turbine environments. Discussions of corrosion mechanisms focus on the reactions in fly ash deposits and combustion gases that can give corrosive materials access to the surface of a superheater tube. Setting the steam temperature of a biomass boiler is a compromise between wasting fuel energy, risking pluggage that will shut the unit down, and creating conditions that will cause rapid corrosion on the superheater tubes and replacement expenses. The most important corrosive species in biomass superheater corrosion are chlorine compounds and the most corrosion resistant alloys are typically FeCrNi alloys containing 20-28% Cr. Although most of these materials contain many other additional additions, there is no coherent theory of the alloying required to resist the combination of high temperature salt deposits and flue gases that are found in biomass boiler superheaters that may cause degradation of superheater tubes. After depletion of chromium by chromate formation or chromic acid volatilization exceeds a critical amount, the protective scale gives way to a thick layer of Fe 2 O 3 over an unprotective (FeCrNi) 3 O 4 spinel. This oxide is not protective and can be penetrated by chlorine species that cause further acceleration of the corrosion rate by a mechanism called active oxidation. Active oxidation, cited as the cause of most biomass superheater corrosion under chloride ash deposits, does not occur in the absence of these alkali salts when the chloride is present as HCl gas. Although a deposit is more corrosive at temperatures where it is molten than at temperatures where it is frozen, increasing superheater tube temperatures through the measured first melting point of fly ash deposits does not necessarily produce a step increase in corrosion rate. Corrosion rate typically accelerates at temperatures below the first melting temperature and mixed deposits may have a broad melting temperature range. Although the environment at a superheater tube surface is initially that of the ash deposits, this chemistry typically changes as the deposits mature. The corrosion rate is controlled by the environment and temperature at the tube surface, which can only be measured indirectly. Some results are counter-intuitive. Two boiler manufacturers and a consortium have developed models to predict fouling and corrosion in biomass boilers in order to specify tube materials for particular operating conditions. It would be very useful to compare the predictions of these models regarding corrosion rates and recommended alloys in the boiler environments where field tests will be performed in the current program. Manufacturers of biomass boilers have concluded that it is more cost-effective to restrict steam temperatures, to co-fire biofuels with high sulfur fuels and/or to use fuel additives rather than try to increase fuel efficiency by operating with superheater tube temperatures above melting temperature of fly ash deposits. Similar strategies have been developed for coal fired and waste-fired boilers. Additives are primarily used to replace alkali metal chloride deposits with higher melting temperature and less corrosive alkali metal sulfate or alkali aluminum silicate deposits. Design modifications that have been shown to control superheater corrosion include adding a radiant pass (empty chamber) between the furnace and the superheater, installing cool tubes immediately upstream of the superheater to trap high chloride deposits, designing superheater banks for quick replacement, using an external superheater that burns a less corrosive biomass fuel, moving circulating fluidized bed (CFB) superheaters from the convective pass into the hot recirculated fluidizing medium and adding an insulating layer to superheater tubes to raise their surface temperature above the dew point temperature of alkali chlorides. These design changes offer advantages but introduce other challenges. For example, operating with superheater temperatures above the dew point of alkali chlorides could require the use of creep-resistant tube alloys and doesn't eliminate chloride corrosion. Improved test methods that can be applied within this project include automated dimensional metrology to make a statistical analysis of depth of penetration and corrosion product thickness, and simultaneous thermal analysis measurements to quantify the melting of complex ashes and avoid the unreliability of the standard ash fusion test. Other important developments in testing include the installation of individually-temperature-controlled superheater loops for corrosion testing in operating boilers and temperature gradient testing.

09 BIOMASS FUELS↗

Beyond Generalized Eigenvalues

Two analysis techniques, the generalized eigenvalue method (GEM) or Prony's method (PM), are commonly used to analyze statistical estimates of correlation functions produced in lattice quantum field theory calculations. GEM takes full advantage of the matrix structure of correlation functions but only considers individual pairs of time separations when much more data exists. PM can be applied to many time separations and many individual matrix elements simultaneously but does not fully exploit the matrix structure of the correlation function. We combine both these methods into a single framework based on matrix polynomials which we call block Prony method (BPM).

Fleming, George T.↗

A function space approach to state and model error estimation for elliptic systems

An approach is advanced for the concurrent estimation of the state and of the model errors of a system described by elliptic equations. The estimates are obtained by a deterministic least-squares approach that seeks to minimize a quadratic functional of the model errors, or equivalently, to find the vector of smallest norm subject to linear constraints in a suitably defined function space. The minimum norm solution can be obtained by solving either a Fredholm integral equation of the second kind for the case with continuously distributed data or a related matrix equation for the problem with discretely located measurements. Solution of either one of these equations is obtained in a batch-processing mode in which all of the data is processed simultaneously or, in certain restricted geometries, in a spatially scanning mode in which the data is processed recursively. After the methods for computation of the optimal estimates are developed, an analysis of the second-order statistics of the estimates and of the corresponding estimation error is conducted. Based on this analysis, explicit expressions for the mean-square estimation error associated with both the state and model error estimates are then developed.

Rodriguez, G.↗

Verification of the accuracy of GEM-L2 in response to criticism by Lambeck and Coleman

The objective was to evaluate the accuracy of the Goddard Earth Model (GEM)-L2 and its improvement in estimating Lageos orbits in order to provide better baselines for plate tectonics, improved polar motion and Earth rotation. Analyses and comparisons with other models are made to verify accuracy in contrast to the Lambeck and Coleman result which denied the authors' accuracy estimates. The analysis presented completely verifies the accuracy of the GEM-L2 model and disproves the statistical methods of Lambeck and Coleman rejecting the accuracy of GEM-L2 as published in their report. The baselines derived from Lageos with GEM-L2 were estimated to have a 2 cm uncertainty due to the errors in the GEM-L2 gravity model which is supported by the results given. The calibration tests indicate that the errors in GEM-L2 should be reduced by about 30 percent which is in the opposite direction to the result obtained by Lambeck and Coleman.

Lerch, F. J.↗

Image processing in remote sensing data analysis - The state of the art

Image analysis techniques applicable to remote sensing data and covering image models, feature detection, segmentation and classification, texture analysis, and matching are studied. Model types for characterizing images examined include random-field, mosaic, and facet models. Edge and corner detection as well as global extraction of linear features are discussed. Pixel clustering and classification are covered in addition to the regional approach to segmentation. Autocorrelation, second-order gray level probability density, and the use of primitive element statistics are discussed in relation to texture analysis. Finally, reducing the cost of (sub)imaging matching methods (e.g., pixelwise comparison of gray levels and normalized cross-correlation between two images) as well as improving match sharpness is considered.

Rosenfeld, A.↗

A function space approach to state and model error estimation for elliptic systems

An approach is advanced for the concurrent estimation of the state and of the model errors of a system described by elliptic equations. The estimates are obtained by a deterministic least-squares approach that seeks to minimize a quadratic functional of the model errors, or equivalently, to find the vector of smallest norm subject to linear constraints in a suitably defined function space. The minimum norm solution can be obtained by solving either a Fredholm integral equation of the second kind for the case with continuously distributed data or a related matrix equation for the problem with discretely located measurements. Solution of either one of these equations is obtained in a batch-processing mode in which all of the data is processed simultaneously or, in certain restricted geometries, in a spatially scanning mode in which the data is processed recursively. After the methods for computation of the optimal esimates are developed, an analysis of the second-order statistics of the estimates and of the corresponding estimation error is conducted. Based on this analysis, explicit expressions for the mean-square estimation error associated with both the state and model error estimates are then developed. While this paper focuses on theoretical developments, applications arising in the area of large structure static shape determination are contained in a closely related paper (Rodriguez and Scheid, 1982).

Rodriguez, G.↗

An overview of the mathematical and statistical analysis component of RICIS

Mathematical and statistical analysis components of RICIS (Research Institute for Computing and Information Systems) can be used in the following problem areas: (1) quantification and measurement of software reliability; (2) assessment of changes in software reliability over time (reliability growth); (3) analysis of software-failure data; and (4) decision logic for whether to continue or stop testing software. Other areas of interest to NASA/JSC where mathematical and statistical analysis can be successfully employed include: math modeling of physical systems, simulation, statistical data reduction, evaluation methods, optimization, algorithm development, and mathematical methods in signal processing.

Hallum, Cecil R.↗

Automated navigation assessment for earth survey sensors using island targets

An automated method has been developed for performing navigation assessment on satellite-based Earth sensor data. The method utilizes islands as targets which can be readily located in the sensor data and identified with reference locations. The essential elements are an algorithm for classifying the sensor data according to source, a reference catalog of island locations, and a robust pattern-matching algorithm for island identification. The algorithms were developed and tested for the Sea-viewing Wide Field-of-view Sensor (SeaWiFS), an ocean color sensor. This method will allow navigation error statistics to be automatically generated for large numbers of points, supporting analysis over large spatial and temporal ranges.

Patt, Frederick S.↗

An Automated Method for Navigation Assessment for Earth Survey Sensors Using Island Targets

An automated method has been developed for performing navigation assessment on satellite-based Earth sensor data. The method utilizes islands as targets which can be readily located in the sensor data and identified with reference locations. The essential elements are an algorithm for classifying the sensor data according to source, a reference catalogue of island locations, and a robust pattern-matching algorithm for island identification. The algorithms were developed and tested for the Sea-viewing Wide Field-of-view Sensor (SeaWiFS), an ocean colour sensor. This method will allow navigation error statistics to be automatically generated for large numbers of points, supporting analysis over large spatial and temporal ranges.

Patt, F. S.↗

Space Launch System Complex Decision-Making Process

The Space Shuttle program has ended and elements of the Constellation Program have either been cancelled or transitioned to new NASA exploration endeavors. The National Aeronautics and Space Administration (NASA) has worked diligently to select an optimum configuration for the Space Launch System (SLS), a heavy lift vehicle that will provide the foundation for future beyond low earth orbit (LEO) large-scale missions for the next several decades. From Fall 2010 until Spring 2011, an SLS decision-making framework was formulated, tested, fully documented, and applied to multiple SLS vehicle concepts at NASA from previous exploration architecture studies. This was a multistep process that involved performing figure of merit (FOM)-based assessments, creating Pass/Fail gates based on draft threshold requirements, performing a margin-based assessment with supporting statistical analyses, and performing sensitivity analysis on each. This paper focuses on the various steps and methods of this process (rather than specific data) that allowed for competing concepts to be compared across a variety of launch vehicle metrics in support of the successful completion of the SLS Mission Concept Review (MCR) milestone.

Lyles, Garry↗

fuel cycle cost analysis of Th/U MOX fuel in a block-type HTR - 18619

With a rapid development of nuclear energy in these years, the supply of nuclear fuel is becoming more and more important. Thorium regains attention because of its abundant reserves and excellent physical properties. As a next generation reactor, the high temperature reactor (HTR) is a promising choice for thorium utilization because of its deeper burnup, harder spectrum and better inherent safety. Therefore, a 600 MWth 108-column modular high temperature reactor (MHR) core model with 20%UO{sub 2}-80%ThO{sub 2}; MOX fuel is adapted. A traditional two-step calculation scheme, lattice calculation then core calculation, is used to get the neutronic performance of equilibrium cycle including uranium enrichment and mass of fuel. Using these data, the fuel cycle cost of HTR can be calculated by the levelized lifetime cost method. In order to take into account the variation of unit price, a well-established, standard statistical analysis technique is used by Monte Carlo simulation, (thus, the probability distribution of fuel cycle cost can be obtained. Furthermore, sensitivity calculations are made to analyze the impact on the fuel cycle cost of variations in the unit prices for each fuel cycle component. The result shows that the most influential parameters are unit price of natural uranium, enrichment and discount rate. Finally a comparison of fuel cycle cost between 100%UO{sub 2} and 20%UO{sub 2}-80%ThO{sub 2} MOX fuel is made. The latter is 11% cheaper than the former. (authors)

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Linking big models to big data: efficient ecosystem model calibration through Bayesian model emulation

Data-model integration plays a critical role in assessing and improving our capacity to predict ecosystem dynamics. Similarly, the ability to attach quantitative statements of uncertainty around model forecasts is crucial for model assessment and interpretation and for setting field research priorities. Bayesian methods provide a rigorous data assimilation framework for these applications, especially for problems with multiple data constraints. However, the Markov chain Monte Carlo (MCMC) techniques underlying most Bayesian calibration can be prohibitive for computationally demanding models and large datasets. We employ an alternative method, Bayesian model emulation of sufficient statistics, that can approximate the full joint posterior density, is more amenable to parallelization, and provides an estimate of parameter sensitivity. Analysis involved informative priors constructed from a meta-analysis of the primary literature and specification of both model and data uncertainties, and it introduced novel approaches to autocorrelation corrections on multiple data streams and emulating the sufficient statistics surface. We report the integration of this method within an ecological workflow management software, Predictive Ecosystem Analyzer (PEcAn), and its application and validation with two process-based terrestrial ecosystem models: SIPNET and ED2. In a test against a synthetic dataset, the emulator was able to retrieve the true parameter values. A comparison of the emulator approach to standard brute-force MCMC involving multiple data constraints showed that the emulator method was able to constrain the faster and simpler SIPNET model's parameters with comparable performance to the brute-force approach but reduced computation time by more than 2 orders of magnitude. The emulator was then applied to calibration of the ED2 model, whose complexity precludes standard (brute-force) Bayesian data assimilation techniques. Both models are constrained after assimilation of the observational data with the emulator method, reducing the uncertainty around their predictions. Performance metrics showed increased agreement between model predictions and data. Our study furthers efforts toward reducing model uncertainties, showing that the emulator method makes it possible to efficiently calibrate complex models.

58 GEOSCIENCES↗

Statistical analysis of non-Maxwellian electron distribution functions measured with angularly resolved Thomson scattering

Angularly resolved Thomson scattering is a novel extension of Thomson scattering, enabling the measurement of the electron velocity distribution function over many orders of magnitude. Here, details of the theoretical basis of the technique and the instrument designed for this measurement are described. Angularly resolved Thomson-scattering data from several experiments are shown with descriptions of the corresponding distribution functions. A reduced model describing the distribution function is given and used to perform a Monte Carlo analysis of the uncertainty in the measurements. The electron density and temperature were determined to a precision of 12% and 21%, respectively, on average, while all other parameters defining the distribution function were generally determined to better than 20%. It was found that these uncertainties were primarily due to limited signal to noise and instrumental effects. Furthermore, measurements with this level of precision were sufficient to distinguish between Maxwellian and non-Maxwellian distribution functions.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Multi-Centrality Graph Spectral Decompositions and Their Application to Cyber Intrusion Detection

Many modern datasets can be represented as graphs and hence spectral decompositions such as graph principal component analysis (PCA) can be useful. Distinct from previous graph decomposition approaches based on subspace projection of a single topological feature, e.g., the centered graph adjacency matrix (graph Laplacian), we propose spectral decomposition approaches to graph PCA and graph dictionary learning that integrate multiple features, including graph walk statistics, centrality measures and graph distances to reference nodes. In this paper we propose a new PCA method for single graph analysis, called multi-centrality graph PCA (MC-GPCA), and a new dictionary learning method for ensembles of graphs, called multi-centrality graph dictionary learning (MC-GDL), both based on spectral decomposition of multi-centrality matrices. As an application to cyber intrusion detection, MC-GPCA can be an effective indicator of anomalous connectivity pattern and MC-GDL can provide discriminative basis for attack classification.

cybersecurity↗

Statistical Approaches to Aerosol Dynamics for Climate Simulation

In this work, we introduce two general non-parametric regression analysis methods for errors-in-variable (EIV) models: the compound regression, and the constrained regression. It is shown that these approaches are equivalent to each other and, to the general parametric structural modeling approach. The advantages of these methods lie in their intuitive geometric representations, their distribution free nature, and their ability to offer a practical solution when the ratio of the error variances is unknown. Each includes the classic non-parametric regression methods of ordinary least squares, geometric mean regression, and orthogonal regression as special cases. Both methods can be readily generalized to multiple linear regression with two or more random regressors.

Compound regression↗