Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Markov Model”

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 1,657 records · Page 92

Analysis of structural perturbations in systems via cost decomposition methods

It has long been common practice to analyze linear dynamic systems by decomposing the total response in terms of individual contributions which are easier to analyze. Examples of this philosophy include the expansion of transfer functions using: (1) the superposition principle, (2) residue theory and partial fraction expansions, (3) Markov parameters, Hankel matrices, and (4) regular and singular perturbations. This paper summarizes a new and different kind of expansion designed to decompose the norm of the response vector rather than the response vector itself. This is referred to as "cost-decomposition' of the system. The notable advantages of this type of decomposition are: (a) easy application to multi-input, multi-output systems, (b) natural compatibility with Linear Quadratic Gaussian Theory, (c) applicability to the analysis of more general types of structural perturbations involving inputs, outputs, states, parameters. Property (c) makes the method suitable for problems in model reduction, measurement/actuator selections, and sensitivity analysis.

Skelton, R. E.↗

Spectroscopic measurements of the electron and ion temperatures and effective ion charge in current sheets formed in two- and three-dimensional magnetic configurations

The spatial distributions of the electron temperature and density, the effective and average ion charges, and the thermal and directed ion velocities in current sheets formed in two-dimensional magnetic fields and three-dimensional magnetic configurations with an X line were studied using spectroscopic and interference holographic methods. The main attention was paid to studying the time evolution of the intensities of spectral lines of the working-gas (argon) and impurity ions under different conditions. Using these data, the electron temperature was calculated with the help of an original mathematical code based on a collisional-radiative plasma model incorporating the processes of ionization and excitation, as well as MHD plasma flows generated in the stage of the current-sheet formation. It is shown that the electron temperature depends on the longitudinal magnetic field, whereas the ion temperature is independent of it. The effective ion charge of the current-sheet plasma was determined for the first time.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

A stochastic model and Monte Carlo algorithm for fluctuation-induced H{sub 2} formation on the surface of interstellar dust grains

A stochastic algorithm for simulation of fluctuation-induced kinetics of H{sub 2} formation on grain surfaces is suggested as a generalization of the technique developed in our recent studies [1] where this method was developed to describe the annihilation of spatially separate electrons and holes in a disordered semiconductor. The stochastic model is based on the spatially inhomogeneous, nonlinear integro-differential Smoluchowski equations with random source term. In this paper we derive the general system of Smoluchowski type equations for the formation of H{sub 2} from two hydrogen atoms on the surface of interstellar dust grains with physisorption and chemisorption sites. We focus in this study on the spatial distribution, and numerically investigate the segregation in the case of a source with a continuous generation in time and randomly distributed in space. The stochastic particle method presented is based on a probabilistic interpretation of the underlying process as a stochastic Markov process of interacting particle system in discrete but randomly progressed time instances. The segregation is analyzed through the correlation analysis of the vector random field of concentrations which appears to be isotropic in space and stationary in time.

79 ASTRONOMY AND ASTROPHYSICS↗

Nonadiabatic transition paths from quantum jump trajectories

We present a means of studying rare reactive pathways in open quantum systems using transition path theory and ensembles of quantum jump trajectories. This approach allows for the elucidation of reactive paths for dissipative, nonadiabatic dynamics when the system is embedded in a Markovian environment. Here, we detail the dominant pathways and rates of thermally activated processes and the relaxation pathways and photoyields following vertical excitation in a minimal model of a conical intersection. We find that the geometry of the conical intersection affects the electronic character of the transition state as defined through a generalization of a committor function for a thermal barrier crossing event. Similarly, the geometry changes the mechanism of relaxation following a vertical excitation. Relaxation in models resulting from small diabatic coupling proceeds through pathways dominated by pure dephasing, while those with large diabatic coupling proceed through pathways limited by dissipation. The perspective introduced here for the nonadiabatic dynamics of open quantum systems generalizes classical notions of reactive paths to fundamentally quantum mechanical processes.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

THE PANCHROMATIC HUBBLE ANDROMEDA TREASURY. IV. A PROBABILISTIC APPROACH TO INFERRING THE HIGH-MASS STELLAR INITIAL MASS FUNCTION AND OTHER POWER-LAW FUNCTIONS

We present a probabilistic approach for inferring the parameters of the present-day power-law stellar mass function (MF) of a resolved young star cluster. This technique (1) fully exploits the information content of a given data set; (2) can account for observational uncertainties in a straightforward way; (3) assigns meaningful uncertainties to the inferred parameters; (4) avoids the pitfalls associated with binning data; and (5) can be applied to virtually any resolved young cluster, laying the groundwork for a systematic study of the high-mass stellar MF (M {approx}> 1 M {sub Sun }). Using simulated clusters and Markov Chain Monte Carlo sampling of the probability distribution functions, we show that estimates of the MF slope, {alpha}, are unbiased and that the uncertainty, {Delta}{alpha}, depends primarily on the number of observed stars and on the range of stellar masses they span, assuming that the uncertainties on individual masses and the completeness are both well characterized. Using idealized mock data, we compute the theoretical precision, i.e., lower limits, on {alpha}, and provide an analytic approximation for {Delta}{alpha} as a function of the observed number of stars and mass range. Comparison with literature studies shows that {approx}3/4 of quoted uncertainties are smaller than the theoretical lower limit. By correcting these uncertainties to the theoretical lower limits, we find that the literature studies yield ({alpha}) = 2.46, with a 1{sigma} dispersion of 0.35 dex. We verify that it is impossible for a power-law MF to obtain meaningful constraints on the upper mass limit of the initial mass function, beyond the lower bound of the most massive star actually observed. We show that avoiding substantial biases in the MF slope requires (1) including the MF as a prior when deriving individual stellar mass estimates, (2) modeling the uncertainties in the individual stellar masses, and (3) fully characterizing and then explicitly modeling the completeness for stars of a given mass. The precision on MF slope recovery in this paper are lower limits, as we do not explicitly consider all possible sources of uncertainty, including dynamical effects (e.g., mass segregation), unresolved binaries, and non-coeval populations. We briefly discuss how each of these effects can be incorporated into extensions of the present framework. Finally, we emphasize that the technique and lessons learned are applicable to more general problems involving power-law fitting.

79 ASTRONOMY AND ASTROPHYSICS↗

Intrepid MCMC: Metropolis-Hastings with exploration

In engineering examples, one often encounters the need to sample from unnormalized distributions with complex shapes that may also be implicitly defined through a physical or numerical simulation model, making it computationally expensive to evaluate the associated density function. For such cases, MCMC has proven to be an invaluable tool. Random-walk Metropolis Methods (also known as Metropolis-Hastings (MH)), in particular, are highly popular for their simplicity, flexibility, and ease of implementation. However, most MH algorithms suffer from significant limitations when attempting to sample from distributions with multiple modes (particularly disconnected ones). Here, in this paper, we present Intrepid MCMC - a novel MH scheme that utilizes a simple coordinate transformation to significantly improve the mode-finding ability and convergence rate to the target distribution of random-walk Markov chains while retaining most of the simplicity of the vanilla MH paradigm. Through multiple examples, we showcase the improvement in the performance of Intrepid MCMC over vanilla MH for a wide variety of target distribution shapes. We also provide an analysis of the mixing behavior of the Intrepid Markov chain, as well as the efficiency of our algorithm for increasing dimensions. A thorough discussion is presented on the practical implementation of the Intrepid MCMC algorithm. Finally, its utility is highlighted through a Bayesian parameter inference problem for a two-degree-of-freedom oscillator under free vibration.

97 - MATHEMATICS AND COMPUTING↗

SEVEN-YEAR WILKINSON MICROWAVE ANISOTROPY PROBE (WMAP ) OBSERVATIONS: GALACTIC FOREGROUND EMISSION

We present updated estimates of Galactic foreground emission using seven years of WMAP data. Using the power spectrum of differences between multi-frequency template-cleaned maps, we find no evidence for foreground contamination outside of the updated (KQ85y7) foreground mask. We place a 15 {mu}K upper bound on rms foreground contamination in the cleaned maps used for cosmological analysis. Further, the cleaning process requires only three power-law foregrounds outside of the mask. We find no evidence for polarized foregrounds beyond those from soft (steep-spectrum) synchrotron and thermal dust emission; in particular we find no indication in the polarization data of an extra 'haze' of hard synchrotron emission from energetic electrons near the Galactic center. We provide an updated map of the cosmic microwave background (CMB) using the internal linear combination method, updated foreground masks, and updates to point source catalogs using two different techniques. With additional years of data, we now detect 471 point sources using a five-band technique and 417 sources using a three-band CMB-free technique. In total there are 62 newly detected point sources, a 12% increase over the five-year release. Also new are tests of the Markov chain Monte Carlo foreground fitting procedure against systematics in the time-stream data, and tests against the observed beam asymmetry. Within a few degrees of the Galactic plane, the behavior in total intensity of low-frequency foregrounds is complicated and not completely understood. WMAP data show a rapidly steepening spectrum from 20 to 40 GHz, which may be due to emission from spinning dust grains, steepening synchrotron, or other effects. Comparisons are made to a 1 deg 408 MHz map (Haslam et al.) and the 11 deg ARCADE 2 data (Singal et al.). We find that spinning dust or steepening synchrotron models fit the combination of WMAP and 408 MHz data equally well. ARCADE data appear inconsistent with the steepening synchrotron model and consistent with the spinning dust model, though some discrepancies remain regarding the relative strength of spinning dust emission. More high-resolution data in the 10-40 GHz range would shed much light on these issues.

79 ASTRONOMY AND ASTROPHYSICS↗

Efficient GPU-Accelerated MultiSource Global Fit Pipeline for LISA Data Analysis

The large-scale analysis task of deciphering gravitational-wave signals in the LISA data stream will be difficult, requiring a large amount of computational resources and extensive development of computational methods. Its high dimensionality, multiple model types, and complicated noise profile require a global fit to all parameters and input models simultaneously. In this work, we detail our global fit algorithm, called “Erebor,” designed to accomplish this challenging task. It is capable of analyzing current state-of-the-art datasets and then growing into the future as more pieces of the pipeline are completed and added. We describe our pipeline strategy, the algorithmic setup, and the results from our analysis of the LDC2A Sangria dataset, which contains massive black hole binaries, compact galactic binaries, and a parametrized noise spectrum whose parameters are unknown to the user. The Erebor algorithm includes three unique and very useful contributions: GPU acceleration for enhanced computational efficiency; ensemble Markov Chain Monte Carlo (MCMC) sampling with multiple MCMC walkers per temperature for better mixing and parallelized sample creation; and special online updates to reversible-jump (or transdimensional) sampling distributions to ensure sampler mixing and accurate initial estimates for detectable sources in the data.We recover posterior distributions for all 15 (6) of the injected massive black hole binaries (MBHB) in the LDC2A training (hidden) dataset. We catalog ∼12000 galactic binaries (∼8000 as high confidence detections) for both the training and hidden datasets. All of the sources and their posterior distributions are provided in publicly available catalogs.

LISA↗

Compressive sampling of polynomial chaos expansions: Convergence analysis and sampling strategies

Sampling orthogonal polynomial bases via Monte Carlo is of interest for uncertainty quantification of models with random inputs, using Polynomial Chaos (PC) expansions. It is known that bounding a probabilistic parameter, referred to as coherence, yields a bound on the number of samples necessary to identify coefficients in a sparse PC expansion via solution to an ℓ{sub 1}-minimization problem. Utilizing results for orthogonal polynomials, we bound the coherence parameter for polynomials of Hermite and Legendre type under their respective natural sampling distribution. In both polynomial bases we identify an importance sampling distribution which yields a bound with weaker dependence on the order of the approximation. For more general orthonormal bases, we propose the coherence-optimal sampling: a Markov Chain Monte Carlo sampling, which directly uses the basis functions under consideration to achieve a statistical optimality among all sampling schemes with identical support. We demonstrate these different sampling strategies numerically in both high-order and high-dimensional, manufactured PC expansions. In addition, the quality of each sampling method is compared in the identification of solutions to two differential equations, one with a high-dimensional random input and the other with a high-order PC expansion. In both cases, the coherence-optimal sampling scheme leads to similar or considerably improved accuracy.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Deep Reinforcement Learning for Distribution System Operations: A Tutorial and Survey

Here, the rapid evolution of modern electric power distribution systems into complex networks of interconnected active devices, distributed generation (DG), and storage poses increasing difficulties for system operators. The large-scale integration of distributed energy resources (DERs) and the rapid exchange of measurement data via communication networks present major opportunities for advancing grid operations but also introduce greater uncertainty, higher data dimensionality, more complex network and device models, and challenging control and optimization problems. Deep reinforcement learning (DRL) algorithms are promising in addressing these challenges. However, they have not been effectively adapted for power systems applications, requiring extensive customization for implementation and evaluation. This has resulted in reproducibility challenges and a steep learning curve for researchers new to applying DRL algorithms to the power systems domain. To bridge these gaps, this tutorial aims to serve as a valuable resource for researchers interested in exploring learning-based algorithms to operate active power distribution networks. Specifically, this work presents a generalized process for translating sequential decision-making problems in power distribution systems into Markov decision process (MDP) formulations, illustrated through concrete grid service examples. Additionally, we introduce a simple environment design strategy to develop and evaluate example DRL algorithms for distribution system applications, complete with an included code repository to guide users through environment construction.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Correlation between surface chemistry and electrocatalytic properties of monodisperse Pt{sub x}Ni{sub 1-x} nanoparticles.

Monodisperse and homogeneous Pt{sub x}Ni{sub 1-x} alloy nanoparticles of various compositions are synthesized via an organic solution approach in order to reveal the correlation between surface chemistry and their electrocatalytic properties. Atomic-level microscopic analysis of the compositional profile and modeling of nanoparticle structure are combined to follow the dependence of Ni dissolution on the initial alloy composition and formation of the Pt-skeleton nanostructures. The developed approach and acquired knowledge about surface structure-property correlation can be further generalized and applied towards the design of advanced functional nanomaterials.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Numerical Methods for Large-Scale Data Factorization (Final Report (Milestone Date 9/30/11) for Subcontract No. B594099)

The following work has been performed by PI Hans De Sterck and graduate student Manda Winlaw for the required tasks 1-5 (as listed in the Statement of Work). Graduate student Manda Winlaw has visited LLNL January 31-March 11, 2011 and May 23-August 19, 2010, working with Van Henson and Mike O'Hara on non-negative matrix factorizations (NMF). She has investigated the dense subgraph clustering algorithm from 'Finding Dense Subgraphs for Sparse Undirected, Directed, and Bipartite Graphs' by Chen and Saad, testing this method on several term-document matrices and adapting it to cluster based on the rank of the subgraphs instead of the density. Manda Winlaw was awarded a first prize in the annual LLNL summer student poster competition for a poster on her NMF research. PI Hans De Sterck has developed a new adaptive algebraic multigrid algorithm for computing a few dominant or minimal singular triplets of sparse rectangular matrices. This work builds on adaptive algebraic multigrid methods that were further developed by the PI and collaborators (including Sanders and Henson) for Markov chains. The method also combines and extends existing multigrid algorithms for the symmetric eigenproblem. The PI has visited LLNL February 22-25, 2011, and has given a CASC seminar 'Algebraic Multigrid for the Singular Value Problem' on this work on February 23, 2011. During his visit, he has discussed this work and related topics with Van Henson, Geoffrey Sanders, Panayot Vassilevski, and others. He has tested the algorithm on PDE matrices and on a term-document matrix, with promising initial results. Manda Winlaw has also started to work, with O'Hara, on estimating probability distributions over undirected graph edges. The goal is to estimate probabilistic models from sets of undirected graph edges for the purpose of prediction, anomaly detection and support to supervised learning. Graduate student Manda Winlaw is writing a paper on the results obtained with O'Hara which will be submitted some time later in 2011 to a data mining conference. PI Hans De Sterck has developed a new optimization algorithm for canonical tensor approximation, formulating an extension of the nonlinear GMRES method to optimization problems. Numerical results for tensors with up to 8 modes show that this new method is efficient for sparse and dense tensors. He has written a paper on this which has been submitted to the SIAM Journal on Scientific Computing. PI Hans De Sterck has further developed his new optimization algorithm for canonical tensor approximation, formulating an extension in terms of steepest-descent preconditioning, which makes the approach generally applicable for nonlinear optimization. He has written a paper on this extension which has been submitted to Numerical Linear Algebra with Applications.

97 MATHEMATICS AND COMPUTING↗

Cosmic microwave background with Brans-Dicke gravity. II. Constraints with the WMAP and SDSS data

Using the covariant formalism developed in a companion paper [F.-Q. Wu, L. E. Qiang, X. Wang, and X. Chen, preceding Article, Phys. Rev. D 82, 083002 (2010)] (paper I), we derive observational constraints on the Brans-Dicke model in a flat Friedmann-Lemaitre-Robertson-Walker universe with a cosmological constant and cold dark matter. The CMB observations we use include the Wilkinson Microwave Anisotropy Probe 5 yr data, the Arcminute Cosmology Bolometer Array Receiver 2007 data, the Cosmic Background Imager polarization data, and the Balloon Observations of Millimetric Extragalactic Radiation and Geophysics 2003 flight data. For the large scale structure we use the matter power spectrum data measured with the luminous red galaxy survey of the Sloan Digital Sky Survey Data Release 4. We parametrize the Brans-Dicke parameter {omega} with a new parameter {zeta}=ln(1/{omega}+1), and use the Markov-Chain Monte Carlo method to explore the parameter space. We find that using CMB data alone, one could place some constraints on positive {zeta} or {omega}, but negative {zeta} or {omega} could not be constrained effectively. However, with additional large scale structure data, one could break the degeneracy at {zeta}<0. The 2{sigma} (95.5%) bound on {zeta} is -0.008 37<{zeta}<0.010 18 (corresponding to {omega}<-120.0 or {omega}>97.8). We also obtained constraints on G/G, the rate of change of G at present, as -1.75x10{sup -12} yr{sup -1}<G/G<1.05x10{sup -12} yr{sup -1}, and {delta}G/G, the total variation of G since the epoch of recombination, as -0.083<{delta}G/G<0.095 at the 2{sigma} confidence level.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Interdiction of a Markovian evader

Network interdiction is a combinatorial optimization problem on an activity network arising in a number of important security-related applications. It is classically formulated as a bilevel maximin problem representing an 'interdictor' and an 'evader'. The evader tries to move from a source node to the target node along the shortest or safest path while the interdictor attempts to frustrate this motion by cutting edges or nodes. The interdiction objective is to find the optimal set of edges to cut given that there is a finite interdiction budget and the interdictor must move first. We reformulate the interdiction problem for stochastic evaders by introducing a model in which the evader follows a Markovian random walk guided by the least-cost path to the target. This model can represent incomplete knowledge about the evader and the graph as well as partial interdiction. We formulate the optimization problem for this model and show how, by exploiting topological ordering of the nodes, one can achieve an order-of-magnitude speedup in computing the objective function. We also introduce an evader-motion-based heuristic that can significantly improve solution quality by providing a global view of the network to approximation methods.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

GIGA-Lens: Fast Bayesian Inference for Strong Gravitational Lens Modeling

We present GIGA-Lens: a gradient-informed, GPU-accelerated Bayesian framework for modeling strong gravitational lensing systems, implemented in TensorFlow and JAX. The three components, optimization using multistart gradient descent, posterior covariance estimation with variational inference, and sampling via Hamiltonian Monte Carlo, all take advantage of gradient information through automatic differentiation and massive parallelization on graphics processing units (GPUs). We test our pipeline on a large set of simulated systems and demonstrate in detail its high level of performance. The average time to model a single system on four Nvidia A100 GPUs is 105 s. The robustness, speed, and scalability offered by this framework make it possible to model the large number of strong lenses found in current surveys and present a very promising prospect for the modeling of ${ \mathcal O }({10}^{5})$ lensing systems expected to be discovered in the era of the Vera C. Rubin Observatory, Euclid, and the Nancy Grace Roman Space Telescope.

79 ASTRONOMY AND ASTROPHYSICS↗

Response Surface Modeling Tool Suite, Version 1.x

The Response Surface Modeling (RSM) Tool Suite is a collection of three codes used to generate an empirical interpolation function for a collection of drag coefficient calculations computed with Test Particle Monte Carlo (TPMC) simulations. The first code, "Automated RSM", automates the generation of a drag coefficient RSM for a particular object to a single command. "Automated RSM" first creates a Latin Hypercube Sample (LHS) of 1,000 ensemble members to explore the global parameter space. For each ensemble member, a TPMC simulation is performed and the object drag coefficient is computed. In the next step of the "Automated RSM" code, a Gaussian process is used to fit the TPMC simulations. In the final step, Markov Chain Monte Carlo (MCMC) is used to evaluate the non-analytic probability distribution function from the Gaussian process. The second code, "RSM Area", creates a look-up table for the projected area of the object based on input limits on the minimum and maximum allowed pitch and yaw angles and pitch and yaw angle intervals. The projected area from the look-up table is used to compute the ballistic coefficient of the object based on its pitch and yaw angle. An accurate ballistic coefficient is crucial in accurately computing the drag on an object. The third code, "RSM Cd", uses the RSM generated by the "Automated RSM" code and the projected area look-up table generated by the "RSM Area" code to accurately compute the drag coefficient and ballistic coefficient of the object. The user can modify the object velocity, object surface temperature, the translational temperature of the gas, the species concentrations of the gas, and the pitch and yaw angles of the object. Together, these codes allow for the accurate derivation of an object's drag coefficient and ballistic coefficient under any conditions with only knowledge of the object's geometry and mass.

Walker, Andrew↗

Response Surface Modeling Tool Suite, Version 1.x

The Response Surface Modeling (RSM) Tool Suite is a collection of three codes used to generate an empirical interpolation function for a collection of drag coefficient calculations computed with Test Particle Monte Carlo (TPMC) simulations. The first code, "Automated RSM", automates the generation of a drag coefficient RSM for a particular object to a single command. "Automated RSM" first creates a Latin Hypercube Sample (LHS) of 1,000 ensemble members to explore the global parameter space. For each ensemble member, a TPMC simulation is performed and the object drag coefficient is computed. In the next step of the "Automated RSM" code, a Gaussian process is used to fit the TPMC simulations. In the final step, Markov Chain Monte Carlo (MCMC) is used to evaluate the non-analytic probability distribution function from the Gaussian process. The second code, "RSM Area", creates a look-up table for the projected area of the object based on input limits on the minimum and maximum allowed pitch and yaw angles and pitch and yaw angle intervals. The projected area from the look-up table is used to compute the ballistic coefficient of the object based on its pitch and yaw angle. An accurate ballistic coefficient is crucial in accurately computing the drag on an object. The third code, "RSM Cd", uses the RSM generated by the "Automated RSM" code and the projected area look-up table generated by the "RSM Area" code to accurately compute the drag coefficient and ballistic coefficient of the object. The user can modify the object velocity, object surface temperature, the translational temperature of the gas, the species concentrations of the gas, and the pitch and yaw angles of the object. Together, these codes allow for the accurate derivation of an object's drag coefficient and ballistic coefficient under any conditions with only knowledge of the object's geometry and mass.

Walker, Andrew↗

Optimized nested Markov chain Monte Carlo sampling: theory

Metropolis Monte Carlo sampling of a reference potential is used to build a Markov chain in the isothermal-isobaric ensemble. At the endpoints of the chain, the energy is reevaluated at a different level of approximation (the 'full' energy) and a composite move encompassing all of the intervening steps is accepted on the basis of a modified Metropolis criterion. By manipulating the thermodynamic variables characterizing the reference system we maximize the average acceptance probability of composite moves, lengthening significantly the random walk made between consecutive evaluations of the full energy at a fixed acceptance probability. This provides maximally decorrelated samples of the full potential, thereby lowering the total number required to build ensemble averages of a given variance. The efficiency of the method is illustrated using model potentials appropriate to molecular fluids at high pressure. Implications for ab initio or density functional theory (DFT) treatment are discussed.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗