Search NASA⌕ Search

SEARCH · Search NASA

Results for “stochastic approximation”

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 811 records · Page 45

Optimization models for integrated biorefinery operations

Variations of physical and chemical characteristics of biomass lead to an uneven flow of biomass in a biorefinery, which reduces equipment utilization and increases operational costs. Uncertainty of biomass supply and high processing costs increase the risk of investing in the US’s cellulosic biofuel industry. We propose a stochastic programming model to streamline processes within a biorefinery. A chance constraint models system’s reliability requirement that the reactor is operating at a high utilization rate given uncertain biomass moisture content, particle size distribution, and equipment failure. The model identifies operating conditions of equipment and inventory level to maintain a continuous flow of biomass to the reactor. Furthermore, the sample average approximation method approximates the chance constraint and a bisection search-based heuristic solves this approximation. A case study is developed using real-life data collected at Idaho National Laboratory’s biomass processing facility. An extensive computational analysis indicates that sequencing of biomass bales based on moisture level, increasing storage capacity, and managing particle size distribution, increases utilization of the reactor and reduces operational costs.

09 BIOMASS FUELS↗

Multiwavelength Variability Study of the Classical BL Lac Object PKS 0735+178 on Timescales Ranging from Decades to Minutes

We present the results of our power spectral analysis for the BL Lac object PKS 0735+178, utilizing the Fermi -LAT survey at high-energy γ -rays, several ground-based optical telescopes, and single-dish radio telescopes operating at GHz frequencies. The novelty of our approach is that, by combining long-term and densely sampled intra-night light curves in the optical regime, we were able to construct for the first time the optical power spectrum of the blazar for a time domain extending from 23 years down to minutes. Our analysis reveals that: (1) the optical variability is consistent with a pure red noise, for which the power spectral density can be well approximated by a single power law throughout the entire time domain probed; (2) the slope of power spectral density at high-energy γ -rays (∼1) is significantly flatter than that found at radio and optical frequencies (∼2) within the corresponding time variability range; (3) for the derived power spectra, we did not detect any low-frequency flattening, nor do we see any evidence for cutoffs at the highest frequencies down to the noise floor levels due to measurement uncertainties. We interpret our findings in terms of a model where the blazar variability is generated by the underlying single stochastic process (at radio and optical frequencies), or a linear superposition of such processes (in the γ -ray regime). Along with the detailed PSD analysis, we also present the results of our extended (1998–2015) intra-night optical monitoring program and newly acquired optical photo-polarimetric data for the source.

79 ASTRONOMY AND ASTROPHYSICS↗

Theoretical and experimental study of the formation conditions of stepped leaders in negative flashes

Natural lightning flashes are stochastic and uncontrollable, and thus, it is difficult to observe the formation process of a downward negative stepped leader (NSL) directly and in detail. This situation has led to some dispute over the actual NSL formation mechanism, and thus has hindered improvements in the lightning shielding analysis model. In this paper, on the basis of controllable long air gap discharge experiments, the formation conditions required for NSLs in negative flashes have been studied. First, a series of simulation experiments on varying scales were designed and carried out. The NSL formation processes were observed, and several of the characteristic process parameters, including the scale, the propagation velocity, and the dark period, were obtained. By comparing the acquired formation processes and the characteristic parameters with those in natural lightning flashes, the similarity between the NSLs in the simulation experiments and those in natural flashes was proved. Then, based on the local thermodynamic equation and the space charge estimation method, the required NSL formation conditions were deduced, and the space background electric field (E{sub b}) was proposed as the primary parameter for NSL formation. Finally, the critical value of E{sub b} required for the formation of NSLs in natural flashes was determined to be approximately 75 kV/m by extrapolation of the results of the simulation experiments.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Dichotomous noise models of gene switches

Molecular noise in gene regulatory networks has two intrinsic components, one part being due to fluctuations caused by the birth and death of protein or mRNA molecules which are often present in small numbers and the other part arising from gene state switching, a single molecule event. Stochastic dynamics of gene regulatory circuits appears to be largely responsible for bifurcations into a set of multi-attractor states that encode different cell phenotypes. The interplay of dichotomous single molecule gene noise with the nonlinear architecture of genetic networks generates rich and complex phenomena. In this paper, we elaborate on an approximate framework that leads to simple hybrid multi-scale schemes well suited for the quantitative exploration of the steady state properties of large-scale cellular genetic circuits. Through a path sum based analysis of trajectory statistics, we elucidate the connection of these hybrid schemes to the underlying master equation and provide a rigorous justification for using dichotomous noise based models to study genetic networks. Numerical simulations of circuit models reveal that the contribution of the genetic noise of single molecule origin to the total noise is significant for a wide range of kinetic regimes.

60 APPLIED LIFE SCIENCES↗

Identification of dominant gas transport frequencies during barometric pumping of fractured rock

Abstract We demonstrate that although barometric pressures are complicated signals comprised of numerous frequencies, it is a subset of these frequencies that drive the overwhelming majority of gas transport in fractured rock. Using an inverse numerical analysis, we demonstrate that a single barometric component with seasonally modulated amplitude approximates gas transport due to a measured barometric signal. If past barometric tendencies are expected to continue at a location, the identification of this frequency can facilitate accurate long term predictions of barometrically induced gas transport negating the need to consider stochastic realizations of future barometric variations. Additionally, we perform an analytical analysis that indicates that there is a set of barometric frequencies, consistent with the inverse numerical analysis, with high production efficiency. Based on the corroborating inverse numerical and analytical analyses, we conclude that there is a set of dominant gas transport frequencies in barometric records.

58 GEOSCIENCES↗

Probabilistic structure of the geodynamo

One of the most intriguing features of Earth's axial magnetic dipole field, well known from the geological record, is its occasional and unpredictable reversal of polarity. Understanding the phenomenon is rendered very difficult by the highly nonlinear nature of the underlying magnetohydrodynamic problem. Numerical simulations of the liquid outer core, where regeneration occurs, are only able to model conditions that are far from Earth-like. On the analytical front, the situation is not much better; basic calculations, such as relating the average rate of reversals to various core parameters, have apparently been intractable. Here, we present a framework for solving such problems. Starting with the magnetic induction equation, we show that by considering its sources to be stochastic processes with fairly general properties, we can derive a differential equation for the joint probability distribution of the dominant toroidal and poloidal modes. This can be simplified to a Fokker-Planck equation and, with the help of an adiabatic approximation, reduced even further to an equation for the dipole amplitude alone. From these equations various quantities related to the magnetic field, including the average reversal rate, field strength, and time to complete a reversal, can be computed as functions of a small number of numerical parameters. As a result, these parameters in turn can be computed from physical considerations or constrained by paleomagnetic, numerical, and experimental data.

58 GEOSCIENCES↗

Time development of a perturbed-spherical nucleus in a pure supercooled liquid. I - Power-law growth of morphological instabilities. II - Nonlinear development

The linear growth stage of the morphological instabilities developing at the liquid-solid interface during crystal growth from an axisymmetric spherical nucleus is analyzed. The corresponding growth rate parameters are calculated numerically, and it is shown that morphological instabilities for free growth evolve according to a power law in agreement with WKB results and contrary to an exponential law found in a quasi-stationary approximation. The time evolution of an arbitrary perturbed liquid-solid interface from a linear into the nonlinear stage is studied. The initial perturbations include the eigenfunctions for the linear problem, localized perturbations of the grain-boundary type, and a stochastic noise. It is shown that the perturbations grow and spread in a wavelike manner. The formation of a predendritic growth stage is characterized by establishment of constant values of tip radius and velocity.

Pines, V.↗

Optimal active control for Burgers equations

A method for active fluid flow control based on control theory is discussed. Dynamic programming and fixed point successive approximations are used to accommodate the nonlinear control problem. The long-term goal of this project is to establish an effective method applicable to complex flows such as turbulence and jets. However, in this report, the method is applied to stochastic Burgers equation as an intermediate step towards this goal. Numerical results are compared with those obtained by gradient search methods.

Ikeda, Yutaka↗

What does reflection from cloud sides tell us about vertical distribution of cloud droplets?

In order to accurately measure the interaction of clouds with aerosols, we have to resolve the vertical distribution of cloud droplet sizes and determine the temperature of glaciation for clean and polluted clouds. Knowledge of the droplet vertical profile is also essential for understanding precipitation. So far, all existing satellites either measure cloud microphysics only at cloud top (e.g., MODIS) or give a vertical profile of precipitation sized droplets (e.g., Cloudsat). What if one measures cloud microphysical properties in the vertical by retrieving them from the solar and infrared radiation reflected or emitted from cloud sides? This was the idea behind CLAIM-3D (A 3D - cloud aerosol interaction mission) recently proposed by NASA GSFC. This presentation will focus on the interpretation of the radiation reflected from cloud sides. In contrast to plane-parallel approximation, a conventional approach to all current operational retrievals, 3D radiative transfer will be used for interpreting the observed reflectances. As a proof of concept, we will show a few examples of radiation reflected from cloud fields generated by a simple stochastic cloud model with prescribed microphysics. Instead of fixed values of the retrieved effective radii, the probability density functions of droplet size distributions will serve as possible retrievals.

Marshak, A.↗

Understanding the Driver of Energetic Electron Precipitation Using Coordinated Multisatellite Measurements

Magnetospheric plasma waves play a significant role in ring current and radiation belt dynamics, leading to pitch angle scattering loss and/or stochastic acceleration of the particles. During a nonstormtime dropout event on 24 September 2013, intense electromagnetic ion cyclotron (EMIC) waves were detected by Van Allen Probe A (RBSP-A). We quantitatively analyze a conjunction event when RBSP-A was located approximately along the same magnetic field line as MetOp- 01, which detected simultaneous precipitation of >30 keV protons and energetic electrons over an unexpectedly broad energy range (>~30 keV). Multi-point observations together with quasilinear theory provide direct evidence that the observed electron precipitation at higher energy (>~700 keV) is primarily driven by EMIC waves. However, the newly observed feature of the simultaneous electron precipitation extending down to ~30 keV is not supported by existing theories and raises an interesting question on whether EMIC waves can scatter such low-energy electrons.

79 ASTRONOMY AND ASTROPHYSICS↗

The functional integral with unconditional Wiener measure for anharmonic oscillator

In this article we propose the calculation of the unconditional Wiener measure functional integral with a term of the fourth order in the exponent by an alternative method as in the conventional perturbative approach. In contrast to the conventional perturbation theory, we expand into power series the term linear in the integration variable in the exponent. In such a case we can profit from the representation of the integral in question by the parabolic cylinder functions. We show that in such a case the series expansions are uniformly convergent and we find recurrence relations for the Wiener functional integral in the N-dimensional approximation. In continuum limit we find that the generalized Gelfand-Yaglom differential equation with solution yields the desired functional integral (similarly as the standard Gelfand-Yaglom differential equation yields the functional integral for linear harmonic oscillator)

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Real-Time Krylov Theory for Quantum Computing Algorithms

Quantum computers provide new avenues to access ground and excited state properties of systems otherwise difficult to simulate on classical hardware. New approaches using subspaces generated by real-time evolution have shown efficiency in extracting eigenstate information, but the full capabilities of such approaches are still not understood. In recent work, we developed the variational quantum phase estimation (VQPE) method, a compact and efficient real-time algorithm to extract eigenvalues on quantum hardware. Here we build on that work by theoretically and numerically exploring a generalized Krylov scheme where the Krylov subspace is constructed through a parametrized real-time evolution, which applies to the VQPE algorithm as well as others. We establish an error bound that justifies the fast convergence of our spectral approximation. We also derive how the overlap with high energy eigenstates becomes suppressed from real-time subspace diagonalization and we visualize the process that shows the signature phase cancellations at specific eigenenergies. We investigate various algorithm implementations and consider performance when stochasticity is added to the target Hamiltonian in the form of spectral statistics. To demonstrate the practicality of such real-time evolution, we discuss its application to fundamental problems in quantum computation such as electronic structure predictions for strongly correlated systems.

97 MATHEMATICS AND COMPUTING↗

Solving the $k$-Sparse Eigenvalue Problem with Reinforcement Learning

We examine the possibility of using a reinforcement learning (RL) algorithm to solve large-scale eigenvalue problems in which the desired the eigenvector can be approximated by a sparse vector with at most k nonzero elements, where k is relatively small compare to the dimension of the matrix to be partially diagonalized. Here, this type of problem arises in applications in which the desired eigenvector exhibits localization properties and in large-scale eigenvalue computations in which the amount of computational resource is limited. When the positions of these nonzero elements can be determined, we can obtain the k-sparse approximation to the original problem by computing eigenvalues of a k × k submatrix extracted from k rows and columns of the original matrix. We review a previously developed greedy algorithm for incrementally probing the positions of the nonzero elements in a k-sparse approximate eigenvector and show that the greedy algorithm can be improved by using an RL method to refine the selection of k rows and columns of the original matrix. We describe how to represent states, actions, rewards and policies in an RL algorithm designed to solve the k-sparse eigenvalue problem and demonstrate the effectiveness of the RL algorithm on two examples originating from quantum many-body physics.

97 MATHEMATICS AND COMPUTING↗

Heat and mass transfer in flames

Heat- and mass-transfer processes in turbulent diffusion flames are discussed, considering turbulent mixing and the structure of single-phase flames, drop processes in spray flames, and nonluminous and luminous flame radiation. Interactions between turbulence and other phenomena are emphasized, concentrating on past work of the author and his associates. The conserved-scalar formalism, along with the laminar-flamelet approximation, is shown to provide reasonable estimates of the structure of gas flames, with modest levels of empiricism. Extending this approach to spray flames has highlighted the importance of drop/turbulence interactions; e.g., turbulent dispersion of drops, modification of turbulence by drops, etc. Stochastic methods being developed to treat these phenomena are yielding encouraging results.

Faeth, G. M.↗

Structural factoring approach for analyzing stochastic networks

The problem of finding the distribution of the shortest path length through a stochastic network is investigated. A general algorithm for determining the exact distribution of the shortest path length is developed based on the concept of conditional factoring, in which a directed, stochastic network is decomposed into an equivalent set of smaller, generally less complex subnetworks. Several network constructs are identified and exploited to reduce significantly the computational effort required to solve a network problem relative to complete enumeration. This algorithm can be applied to two important classes of stochastic path problems: determining the critical path distribution for acyclic networks and the exact two-terminal reliability for probabilistic networks. Computational experience with the algorithm was encouraging and allowed the exact solution of networks that have been previously analyzed only by approximation techniques.

Hayhurst, Kelly J.↗

Operating Procedure Changes to Improve Antiproton Production at the Fermilab Tevatron Collider

Since the start of Fermilab Collider Run II in 2001, the maximum weekly antiproton accumulation rate has increased from 400 x 10{sup 10} Pbars/week to approximately 3,700 x 10{sup 10} Pbars/week. There are many factors contributing to this increase, one of which involves changes to operational procedures that have streamlined and automated Antiproton Source production. Automation has been added to the beam line orbit control, stochastic cooling power level management, and RF settings. In addition, daily tuning efforts have been streamlined by implementing sequencer driven tuning software.

43 PARTICLE ACCELERATORS↗

An Atomic Mix Closure for Stochastic Media Transport Problems

The accurate modeling of stochastic media in radiation transport calculations remains on ongoing research challenge. The Levermore-Pomraning (LP) closure for the statistical transport equation is simple and inexpensive, but it is often inaccurate. More accurate closures have been proposed, but they can be prohibitively expensive. In this work we present a relatively inexpensive closure that we have observed to be more accurate than LP in almost all cases studied. Obtaining solutions to stochastic transport problems can be quite difficult. Transport calculations on an ensemble of explicit realizations generated by Monte Carlo sampling can be prohibitively expensive. Atomic mix or LP calculations are relatively inexpensive, but they can be inaccurate. The present work illustrates an approach that is less expensive than Monte Carlo sampling yet more accurate than atomic mix or LP approximations. We still need to analyze the reasons why the atomic mix closures derived for thicker problems lead to source iteration instabilities. It is not clear if the closure itself induces an ill-posed problem or if it is merely the iterative process that is problematic. We hope to examine alternative source terms to drive the subsidiary calculations. We also want to extend the work to multigroup and multidimensional problems. (authors)

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Brownian aggregation rate of colloid particles with several active sites

We theoretically analyze the aggregation kinetics of colloid particles with several active sites. Such particles (so-called “patchy particles”) are well known as chemically anisotropic reactants, but the corresponding rate constant of their aggregation has not yet been established in a convenient analytical form. Using kinematic approximation for the diffusion problem, we derived an analytical formula for the diffusion-controlled reaction rate constant between two colloid particles (or clusters) with several small active sites under the following assumptions: the relative translational motion is Brownian diffusion, and the isotropic stochastic reorientation of each particle is Markovian and arbitrarily correlated. This formula was shown to produce accurate results in comparison with more sophisticated approaches. Also, to account for the case of a low number of active sites per particle we used Monte Carlo stochastic algorithm based on Gillespie method. Simulations showed that such discrete model is required when this number is less than 10. Finally, we applied the developed approach to the simulation of immunoagglutination, assuming that the formed clusters have fractal structure.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗