Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Probabilities Mathematics”

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 1,135 records · Page 63

Statistical Coverage Concerns in a Revised k-Effective of the World Problem [Slides]

A revised version of k-Effective of the World is defined specifically to stress current Monte Carlo power iteration techniques. Results of numerous independent Monte Carlo calculations show that even for batch sizes that are typically considered reasonable (5-10K), incorrect results for k are obtained with non-trivial probability. The causes for what makes a problem prone to having these difficulties are identified and advice for criticality safety practitioners is provided.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

A Probabilistic Approach To Selecting Build Configurations in Package Managers

In the past decade software has grown significantly in complexity and scale. Likewise the number of dependencies for most software has increased with typical software packages depending on tens to hundreds of other packages. Such large numbers of dependencies place a significant burden on package managers to correctly maintain dependency lists and constraints between them. Due to this package managers have incorporated sophisticated tooling such as SAT solvers into their dependency management mechanisms. Despite these tools package managers still rely on many human annotated constraints for dependency and version selection. These are error-prone and require a significant amount of labor to constantly update and test. In this paper we propose a methodology to make use of historical build results in selecting the version for package dependencies. Our method utilizes the flexibility of the Spack package manager’s heavily parameterized package configurations to incorporate a machine learning model trained to predict the probability of build outcomes. This work is able to build and install packages with a 13% higher success rate than the default version selection mechanism in Spack.

97 MATHEMATICS AND COMPUTING↗

Graphical methods for the sensitivity analysis in discriminant analysis

Similar to regression, many measures to detect influential data points in discriminant analysis have been developed. Many follow similar principles as the diagnostic measures used in linear regression in the context of discriminant analysis. Here we focus on the impact on the predicted classification posterior probability when a data point is omitted. The new method is intuitive and easily interpretative compared to existing methods. We also propose a graphical display to show the individual movement of the posterior probability of other data points when a specific data point is omitted. This enables the summaries to capture the overall pattern of the change.

97 MATHEMATICS AND COMPUTING↗

Monitoring the state of the human airways by analysis of respiratory sound

A mechanism whereby sound is generated by the motion of vortices in the human lung is described. This mechanism is believed to be responsible for most of the sound which is generated both on inspiration and expiration in normal lungs. Mathematical expressions for the frequencies of sound generated, which depend only upon the axial flow velocity and diameters of the bronchi, are derived. This theory allows the location within the bronchial tree from which particular sounds emanate to be determined. Redistribution of pulmonary blood volume following transition from earth gravity to the weightless state probably alters the caliber of certain airways and doubtless alters sound transmission properties of the lung. We believe that these changes can be monitored effectively and non-invasively by spectral analysis of pulmonary sound.

Hardin, J. C.↗

Monitoring the state of the human airways by analysis of respiratory sound

A mechanism whereby sound is generated by the motion of vortices in the human lung is described. This mechanism is believed to be responsible for most of the sound which is generated both on inspiration and expiration in normal lungs. Mathematical expressions for the frequencies of sound generated, which depend only upon the axial flow velocity and diameters of the bronchi, are derived. This theory allows the location within the bronchial tree from which particular sounds emanate to be determined. Redistribution of pulmonary blood volume following transition from Earth gravity to the weightless state probably alters the caliber of certain airways and doubtless alters sound transmission properties of the lung. We believe that these changes can be monitored effectively and non-invasively by spectral analysis of pulmonary sound.

Lung/anatomy & histology/physiology↗

Bayesian Inference with Latent Hamiltonian Neural Networks (L-HNNs)

When sampling for Bayesian inference, one popular approach is to use Hamiltonian Monte Carlo (HMC) and the No-U-Turn Sampler (NUTS). However, HMC and NUTS can require numerous numerical gradients of the target density and can prove slow in practice. We propose Hamiltonian neural networks (HNNs) with HMC and NUTS for solving Bayesian inference problems [1, 2]. Once trained, HNNs do not require gradients of the target density while sampling. Moreover, they satisfy important properties such as perfect time reversibility and Hamiltonian conservation, making them well suited for use within HMC and NUTS because stationarity can be shown. We also propose an HNN extension called latent HNNs (L-HNNs), which predict latent variable outputs. Compared to HNNs, L-HNNs offer improved expressivity and a reduction in integration errors. Finally, we propose employing L-HNNs in NUTS with an online error monitoring scheme to prevent degeneracy of the sampling in regions of low probability density. We demonstrate L-HNNs in NUTS with online error monitoring by using several example cases involving complex, heavy-tailed, and high local curvature probability densities. Overall, L-HNNs in NUTS with online error monitoring satisfactorily inferred these probability densities. Compared to traditional NUTS, L-HNNs in NUTS with online error monitoring improved the effective sample size (ESS) per gradient by an order of magnitude.

97 MATHEMATICS AND COMPUTING↗

Bayesian Monte Carlo Evaluation of Imperfect (n, 233 U) Data and Model

Conventional nuclear data evaluation methods using generalized linear least squares make the following assumptions: prior and posterior probability distribution functions (PDFs) of all model parameters and data are normal (Gaussian); the linear approximation is sufficiently accurate to minimize the cost function (even for nonlinear models); the model (e.g., of neutron cross section) and experimental data (including covariance data) are without defect and prior PDFs of parameters and measured data are known perfectly. Neglect of covariance between model parameters and measured data in conventional evaluations contributes to imperfections. These assumptions are inherent to the generalized linear least squares minimization method commonly used for resolved resonance region neutron cross section evaluations but are often not justified due to the presence of non-normal PDFs, nonlinear models (e.g., R-matrix formalism), and inherent imperfections in data and models (e.g., imperfect covariance data). Here, these assumptions are removed in a mathematical framework of Bayes’ theorem, which is implemented using the Metropolis-Hastings Monte Carlo method. Most importantly, new parameters are introduced to parameterize discrepancies between the theoretical model and measured data to quantify judgement about discrepancies or imperfections in a reproducible manner. An evaluation of 233U in the eV region using the ENDF-B/VIII.0 library and transmission data (Guber et al.) is presented, and posterior parameters are compared to those obtained by conventional evaluation methods. This example illustrates the effects of removing the most harmful assumption: that of model-data perfection.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

On-the-fly generation of differential resonance scattering probability distribution functions for Monte Carlo codes

Current Monte Carlo codes use one of three models to model neutron scattering in the epithermal energy range: (1) the asymptotic scattering model, (2) the free gas scattering model, or (3) the S({alpha},{beta}) model, depending on the neutron energy and the specific Monte Carlo code. The free gas scattering model assumes the scattering cross section is constant over the neutron energy range, which is usually a good approximation for light nuclei, but not for heavy nuclei where the scattering cross section may have several resonances in the epithermal region. Several researchers in the field have shown that using the free gas scattering model in the vicinity of the resonances in the lower epithermal range can under-predict resonance absorption due to the up-scattering phenomenon. Existing methods all involve performing the collision analysis in the center-of-mass frame, followed by a conversion back to the laboratory frame. In this paper, we will present a new sampling methodology that (1) accounts for the energy-dependent scattering cross sections in the collision analysis and (2) acts in the laboratory frame, avoiding the conversion to the center-of-mass frame. The energy dependence of the scattering cross section was modeled with even-ordered polynomials to approximate the scattering cross section in Blackshaw's equations for the moments of the differential scattering PDFs. These moments were used to sample the outgoing neutron speed and angle in the laboratory frame on-the-fly during the random walk of the neutron. Results for criticality studies on fuel pin and fuel assembly calculations using these methods showed very close comparison to results using the reference Doppler-broadened rejection correction (DBRC) scheme. (authors)

97 MATHEMATICS AND COMPUTING↗

Hellmann–Feynman connection for the relative Fisher information

The (i) reciprocity relations for the relative Fisher information (RFI, hereafter) and (ii) a generalized RFI–Euler theorem are self-consistently derived from the Hellmann–Feynman theorem. These new reciprocity relations generalize the RFI–Euler theorem and constitute the basis for building up a mathematical Legendre transform structure (LTS, hereafter), akin to that of thermodynamics, that underlies the RFI scenario. This demonstrates the possibility of translating the entire mathematical structure of thermodynamics into a RFI-based theoretical framework. Virial theorems play a prominent role in this endeavor, as a Schrödinger-like equation can be associated to the RFI. Lagrange multipliers are determined invoking the RFI–LTS link and the quantum mechanical virial theorem. An appropriate ansatz allows for the inference of probability density functions (pdf’s, hereafter) and energy-eigenvalues of the above mentioned Schrödinger-like equation. The energy-eigenvalues obtained here via inference are benchmarked against established theoretical and numerical results. A principled theoretical basis to reconstruct the RFI-framework from the FIM framework is established. Numerical examples for exemplary cases are provided. - Highlights: • Legendre transform structure for the RFI is obtained with the Hellmann–Feynman theorem. • Inference of the energy-eigenvalues of the SWE-like equation for the RFI is accomplished. • Basis for reconstruction of the RFI framework from the FIM-case is established. • Substantial qualitative and quantitative distinctions with prior studies are discussed.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Dynamic Capacity and Surface Fatigue Life for Spur and Helical Gears

A mathematical model for surface fatigue life of gear, pinion, or entire meshing gear train is given. The theory is based on a previous statistical approach for rolling-element bearings. Equations are presented which give the dynamic capacity of the gear set. The dynamic capacity is the transmitted tangential load which gives a 90 percent probability of survival of the gear set for one million pinion revolutions. The analytical results are compared with test data for a set of AISI 9310 spur gears operating at a maximum Hertz stress of 1.71 billion N/sq m and 10,000 rpm. The theoretical life predictions are shown to be good when material constants obtained from rolling-element bearing tests were used in the gear life model.

Coy, J. J.↗

A kinetic-theory approach to turbulent chemically reacting flows

The paper examines the mathematical and physical foundations for the kinetic theory of reactive turbulent flows, discussing the differences and relation between the kinetic and averaged equations, and comparing some solutions of the kinetic equations obtained by the Green's function method with those obtained by the approximate bimodal method. The kinetic method described consists essentially in constructing the probability density functions of the chemical species on the basis of solutions of the Langevin stochastic equation for the influence of eddies on the behavior of fluid elements. When the kinetic equations are solved for the structure of the diffusion flame established in a shear layer by the bimodal method, discontinuities in gradients of the mean concentrations at the two flame edges appear. This is a consequence of the bimodal approximation of all distribution functions by two dissimilar half-Maxwellian functions, which is a very crude approximation. These discontinuities do not appear when the solutions are constructed by the Green's function method described here.

Chung, P. M.↗

Optical character recognition of handwritten Arabic using hidden Markov models

The problem of optical character recognition (OCR) of handwritten Arabic has not received a satisfactory solution yet. In this paper, an Arabic OCR algorithm is developed based on Hidden Markov Models (HMMs) combined with the Viterbi algorithm, which results in an improved and more robust recognition of characters at the sub-word level. Integrating the HMMs represents another step of the overall OCR trends being currently researched in the literature. The proposed approach exploits the structure of characters in the Arabic language in addition to their extracted features to achieve improved recognition rates. Useful statistical information of the Arabic language is initially extracted and then used to estimate the probabilistic parameters of the mathematical HMM. A new custom implementation of the HMM is developed in this study, where the transition matrix is built based on the collected large corpus, and the emission matrix is built based on the results obtained via the extracted character features. The recognition process is triggered using the Viterbi algorithm which employs the most probable sequence of sub-words. The model was implemented to recognize the sub-word unit of Arabic text raising the recognition rate from being linked to the worst recognition rate for any character to the overall structure of the Arabic language. Numerical results show that there is a potentially large recognition improvement by using the proposed algorithms.

Character recognition↗

Relaxation times and modes of disturbed aggregate distribution in micellar solutions with fusion and fission of micelles

We have performed direct numerical calculations of the kinetics of relaxation in the system of surfactant spherical micelles under joint action of the molecular mechanism with capture and emission of individual surfactant molecules by molecular aggregates and the mechanism of fusion and fission of the aggregates. As a basis, we have taken the difference equations of aggregation and fragmentation in the form of the generalized kinetic Smoluchowski equations for aggregate concentrations. The calculations have been made with using the droplet model of molecular surfactant aggregates and two modified Smoluchowski models for the coefficients of aggregate-monomer and aggregate-aggregate fusions which take into account the effects of the aggregate size and presence of hydrophobic spots on the aggregate surface. A full set of relaxation times and corresponding relaxation modes for nonequilibrium aggregate distribution in the aggregation number has been found. The dependencies of these relaxation times and modes on the total concentration of surfactant in the solution and the special parameter controlling the probability of fusion in collisions of micelles with other micelles have been studied.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

A Pseudoreversible Normalizing Flow for Stochastic Dynamical Systems with Various Initial Distributions

Here, we present a pseudoreversible normalizing flow method for efficiently generating samples of the state of a stochastic differential equation (SDE) with various initial distributions. The primary objective is to construct an accurate and efficient sampler that can be used as a surrogate model for computationally expensive numerical integration of SDEs, such as those employed in particle simulation. After training, the normalizing flow model can directly generate samples of the SDE’s final state without simulating trajectories. The existing normalizing flow model for SDEs depends on the initial distribution, meaning the model needs to be retrained when the initial distribution changes. The main novelty of our normalizing flow model is that it can learn the conditional distribution of the state, i.e., the distribution of the final state conditional on any initial state, such that the model only needs to be trained once and the trained model can be used to handle various initial distributions. This feature can provide a significant computational saving in studies of how the final state varies with the initial distribution. Additionally, we propose to use a pseudoreversible network architecture to define the normalizing flow model, which has sufficient expressive power and training efficiency for a variety of SDEs in science and engineering, e.g., in particle physics. We provide a rigorous convergence analysis of the pseudoreversible normalizing flow model to the target probability density function in the Kullback–Leibler divergence metric. Numerical experiments are provided to demonstrate the effectiveness of the proposed normalizing flow model.

97 MATHEMATICS AND COMPUTING↗

Gaussian Process Classification for Galaxy Blend Identification in LSST

Abstract A significant fraction of observed galaxies in the Rubin Observatory Legacy Survey of Space and Time (LSST) will overlap at least one other galaxy along the same line of sight, in a so-called “blend.” The current standard method of assessing blend likelihood in LSST images relies on counting up the number of intensity peaks in the smoothed image of a blend candidate, but the reliability of this procedure has not yet been comprehensively studied. Here we construct a realistic distribution of blended and unblended galaxies through high-fidelity simulations of LSST-like images, and from this we examine the blend classification accuracy of the standard peak-finding method. Furthermore, we develop a novel Gaussian process blend classifier model, and show that this classifier is competitive with both the peak finding method as well as with a convolutional neural network model. Finally, whereas the peak-finding method does not naturally assign probabilities to its classification estimates, the Gaussian process model does, and we show that the Gaussian process classification probabilities are generally reliable.

79 ASTRONOMY AND ASTROPHYSICS↗

HUMAN FAILURE EVENT DEPENDENCE: WHAT ARE THE LIMITS

Abstract Objectives: This paper discusses the differences between classical human reliability analysis (HRA) dependence and the full spectrum of probabilistic dependence. Positive influence suggests an error increases the likelihood of subsequent errors or success increases the likelihood of subsequent success. Currently the typical method for dependence in HRA implements the Technique for Human Error Rate Prediction (THERP) positive dependence equations. This assumes that the dependence between two human failure events varies at discrete levels between zero and complete dependence (as defined by THERP). Dependence in THERP does not consistently span dependence values between 0 and 1. In contrast, probabilistic dependence employs Bayes Law, and addresses a continuous range of dependence. Methods: Using the laws of probability, complete dependence and maximum positive dependence do not always agree. Maximum dependence is when two events overlap to their fullest amount. Maximum negative dependence is the smallest amount that two events can overlap. When the minimum probability of two events overlapping is less than independence, negative dependence occurs. For example, negative dependence is when an operator fails to actuate Pump A, thereby increasing his or her chance of actuating Pump B. The initial error actually increases the chance of subsequent success. Results: Comparing THERP and probability theory yields different results in certain scenarios; with the latter addressing negative dependence. Given that most human failure events are rare, the minimum overlap is typically 0. And when the second event is smaller than the first event the max dependence is less than 1, as defined by Bayes Law. As such alternative dependence equations are provided along with a look-up table defining the maximum and maximum negative dependence given the probability of two events. Conclusions: THERP dependence has been used ubiquitously for decades, and has provided approximations of the dependencies between two events. Since its inception, computational abilities have increased exponentially, and alternative approaches that follow the laws of probability dependence need to be implemented. These new approaches need to consider negative dependence and identify when THERP output is not appropriate.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Modeling Damage Complexity-Dependent Non-Homologous End-Joining Repair Pathway

Non-homologous end joining (NHEJ) is the dominant DNA double strand break (DSB) repair pathway and involves several repair proteins such as Ku, DNA-PKcs, and XRCC4. It has been experimentally shown that the choice of NHEJ proteins is determined by the complexity of DSB. In this paper, we built a mathematical model, based on published data, to study how NHEJ depends on the damage complexity. Under an appropriate set of parameters obtained by minimization technique, we can simulate the kinetics of foci track formation in fluorescently tagged mammalian cells, Ku80-EGFP and DNA-PKcs-YFP for simple and complex DSB repair, respectively, in good agreement with the published experimental data, supporting the notion that simple DSB undergo fast repair in a Ku-dependent, DNA-PKcs-independent manner, while complex DSB repair requires additional DNA-PKcs for end processing, resulting in its slow repair, additionally resulting in slower release rate of Ku and the joining rate of complex DNA ends. Based on the numerous experimental descriptions, we investigated several models to describe the kinetics for complex DSB repair. An important prediction of our model is that the rejoining of complex DSBs is through a process of synapsis formation, similar to a second order reaction between ends, rather than first order break filling/joining. The synapsis formation (SF) model allows for diffusion of ends before the synapsis formation, which is precluded in the first order model by the rapid coupling of ends. Therefore, the SF model also predicts the higher number of chromosomal aberrations observed with high linear energy transfer (LET) radiation due to the higher proportion of complex DSBs compared to low LET radiation, and an increased probability of misrejoin following diffusion before the synapsis is formed, while the first order model does not provide a mechanism for the increased effectiveness in chromosomal aberrations observed.

59 BASIC BIOLOGICAL SCIENCES↗