Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “MARKOV PROCESS”

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 901 records · Page 50

Parallel simulated annealing with embedded machine learning and multifidelity models for reactor core design

This paper presents extensions to a penalty-free, parallel simulated annealing (SA) algorithm for multi-constrained combinatorial optimization with the aim of embedding multi-fidelity physics models into the annealing procedure. The method uses a low-fidelity, quickly executing model for rapid design space exploration and a high-fidelity model for detailed constraint resolution and on-the-fly bias correction. Machine learning models updated within the annealing procedure were used to bridge the gap between the multi-fidelity models, which led to accurate rapid exploration and efficient detailed constraint resolution. A software implementation of the new multi-fidelity optimization methods, called ML-PSA, was demonstrated on a continuous multi-fidelity optimization problem and a constrained combinatorial PWR lattice design problem. These problems demonstrate some of the features, parallel performance characteristics, and extensible nature of the multi-fidelity SA methods. This paper shows that the developed software and procedure are a general optimization tool that can be applied to a wide variety of scientific and engineering design optimization applications. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Measuring small absorptions by exploiting photothermal self-phase modulation

We present a method for the measurement of small optical absorption coefficients. The method exploits the deformation of cavity Airy peaks that occur if the cavity contains an absorbing material with a nonzero thermorefractive coefficient dn/dT or a nonzero expansion coefficient ath. Light absorption leads to a local temperature change and to an intensity-dependent phase shift, i.e., to a photothermal self-phase modulation. The absorption coefficient is derived from a comparison of time-resolved measurements with a numerical time-domain simulation applying a Markov-chain Monte Carlo algorithm. We apply our method to the absorption coefficient of lithium niobate doped with 7mol.% magnesium oxide and derive a value of {alpha}LN=(5.9{+-}0.9)x10{sup -4}/cm. Our method should also apply to materials with much lower absorption coefficients. Based on our modeling, we estimate that, with cavity finesse values of the order of 10{sup 4}, absorption coefficients of as low as 10{sup -8}/cm can be measured.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Methodology for Path Planning with Dynamic Data-Driven Flight Capability Estimation

This paper presents methodology to enable path planning for an unmanned aerial vehicle that uses dynamic data-driven flight capability estimation. The main contribution of the work is a general mathematical approach that leverages offline high-fidelity physics-based modeling together with onboard sensor measurements to achieve dynamic path planning. The mathematical framework, expressed as a constrained partially observable Markov decision process, accounts for vehicle capability constraints and is robust to modeling error and disturbances in both the vehicle process and measurement models. Vehicle capability constraints are incorporated using probabilistic support vector machine surrogates of high-fidelity physics-based models that adequately capture the richness of the vehicle dynamics. Sensor measurements are treated in a general manner and can include combinations of multiple modalities such as that through global positioning system signals, inertial mass unit outputs, and structural strain data of the airframe. Results are presented for a simulated three-dimensional environment and point-mass airplane model. The vehicle can dynamically adjust its trajectory according to the observations it receives about its current state of health, thereby retaining a high probability of survival and mission success.

42 ENGINEERING↗

Pore Formation in the Laminar Structure of Wear-Resistant Modified Powder Coatings

The open porosity of wear-resistant modified coatings is investigated, and their macrostructure is analyzed. A statistical method is proposed for describing the formation of porosity in the laminar structure of a powder coating. The matrix of transition probabilities is determined for a uniform Markov chain. The porosity of a coating on a smooth substrate is calculated.

36 MATERIALS SCIENCE↗

Approximating Nash Equilibrium in Day-ahead Electricity Market Bidding with Multi-agent Deep Reinforcement Learning

In this paper, a day-ahead electricity market bidding problem with multiple strategic generation company (GEN-CO) bidders is studied. The problem is formulated as a Markov game model, where GENCO bidders interact with each other todevelop their optimal day-ahead bidding strategies. Considering unobservable information in the problem, a model-free and data-driven approach, known as multi-agent deep deterministic policy gradient (MADDPG), is applied for approximating the Nash equilibrium (NE) in the above Markov game. The MADDPG algorithm has the advantage of generalization due to the automatic feature extraction ability of the deep neural networks. The algorithm is tested on an IEEE 30-bus system with three competitive GENCO bidders in both an uncongested caseand a congested case. Comparisons with a truthful bidding strategy and state-of-the-art deep reinforcement learning methods including deep Q network and deep deterministic policy gradient (DDPG) demonstrate that the applied MADDPG algorithm can find a superior bidding strategy for all the market participants with increased profit gains. In addition, the comparison with a conventional model-based method shows that the MADDPG algorithm has higher computational efficiency, which is feasible for real-world applications.

24 POWER TRANSMISSION AND DISTRIBUTION↗

THE ORBIT AND MASS OF THE THIRD PLANET IN THE KEPLER-56 SYSTEM

While the vast majority of multiple-planet systems have orbital angular momentum axes that align with the spin axis of their host star, Kepler-56 is an exception: its two transiting planets are coplanar yet misaligned by at least 40° with respect to the rotation axis of their host star. Additional follow-up observations of Kepler-56 suggest the presence of a massive, non-transiting companion that may help explain this misalignment. We model the transit data along with Keck/HIRES and HARPS-N radial velocity data to update the masses of the two transiting planets and infer the physical properties of the third, non-transiting planet. We employ a Markov Chain Monte Carlo sampler to calculate the best-fitting orbital parameters and their uncertainties for each planet. We find the outer planet has a period of 1002 ± 5 days and minimum mass of 5.61 ± 0.38 M{sub Jup}. We also place a 95% upper limit of 0.80 m s{sup −1} yr{sup −1} on long-term trends caused by additional, more distant companions.

79 ASTRONOMY AND ASTROPHYSICS↗

DEVELOPING ATMOSPHERIC RETRIEVAL METHODS FOR DIRECT IMAGING SPECTROSCOPY OF GAS GIANTS IN REFLECTED LIGHT. I. METHANE ABUNDANCES AND BASIC CLOUD PROPERTIES

Upcoming space-based coronagraphic instruments in the next decade will perform reflected light spectroscopy and photometry of cool directly imaged extrasolar giant planets. We are developing a new atmospheric retrieval methodology to help assess the science return and inform the instrument design for such future missions, and ultimately interpret the resulting observations. Our retrieval technique employs a geometric albedo model coupled with both a Markov chain Monte Carlo Ensemble Sampler ( emcee ) and a multimodal nested sampling algorithm ( MultiNest ) to map the posterior distribution. This combination makes the global evidence calculation more robust for any given model and highlights possible discrepancies in the likelihood maps. As a proof of concept, our current atmospheric model contains one or two cloud layers, methane as a major absorber, and a H{sub 2}–He background gas. This 6-to-9 parameter model is appropriate for Jupiter-like planets and can be easily expanded in the future. In addition to deriving the marginal likelihood distribution and confidence intervals for the model parameters, we perform model selection to determine the significance of methane and cloud detection as a function of expected signal-to-noise ratio in the presence of spectral noise correlations. After internal validation, the method is applied to realistic spectra of Jupiter, Saturn, and HD 99492c, a model observing target. We find that the presence or absence of clouds and methane can be determined with high confidence, while parameter uncertainties are model dependent and correlated. Such general methods will also be applicable to the interpretation of direct imaging spectra of cloudy terrestrial planets.

79 ASTRONOMY AND ASTROPHYSICS↗

RESOLVED MILLIMETER-WAVELENGTH OBSERVATIONS OF DEBRIS DISKS AROUND SOLAR-TYPE STARS

The presence of debris disks around young main-sequence stars hints at the existence and structure of planetary systems. Millimeter-wavelength observations probe large grains that trace the location of planetesimal belts. The Formation and Evolution of Planetary Systems Spitzer Legacy survey of nearby young solar analogues yielded a sample of five debris disk-hosting stars with millimeter flux suitable for interferometric follow-up. We present observations with the Submillimeter Array (SMA) and the Combined Array for Research in Millimeter-wave Astronomy at ∼2″ resolution that spatially resolve the debris disks around these nearby (d ∼ 50 pc) stars. Two of the five disks (HD 377, HD 8907) are spatially resolved for the first time and one (HD 104860) is resolved at millimeter wavelengths for the first time. We combine our new observations with archival SMA and Atacama Large Millimeter/Submillimeter Array data to enable a uniform analysis of the full five-object sample. We simultaneously model the broadband photometric data and resolved millimeter visibilities to constrain the dust temperatures and disk morphologies, and perform a Markov Chain Monte Carlo analysis to fit for basic structural parameters. We find that the radii and widths of the cold outer belts exhibit properties consistent with scaled-up versions of the Solar System's Kuiper Belt. All the disks exhibit characteristic grain sizes comparable to the blowout size, and all the resolved observations of emission from large dust grains are consistent with an axisymmetric dust distribution to within the uncertainties. These results are consistent with comparable studies carried out at infrared wavelengths.

79 ASTRONOMY AND ASTROPHYSICS↗

OBSERVATIONAL SIGNATURES OF CORONAL LOOP HEATING AND COOLING DRIVEN BY FOOTPOINT SHUFFLING

The evolution of a coronal loop is studied by means of numerical simulations of the fully compressible three-dimensional magnetohydrodynamic equations using the HYPERION code. The footpoints of the loop magnetic field are advected by random motions. As a consequence, the magnetic field in the loop is energized and develops turbulent nonlinear dynamics characterized by the continuous formation and dissipation of field-aligned current sheets: energy is deposited at small scales where heating occurs. Dissipation is nonuniformly distributed so that only a fraction of the coronal mass and volume gets heated at any time. Temperature and density are highly structured at scales that, in the solar corona, remain observationally unresolved: the plasma of our simulated loop is multithermal, where highly dynamical hotter and cooler plasma strands are scattered throughout the loop at sub-observational scales. Numerical simulations of coronal loops of 50,000 km length and axial magnetic field intensities ranging from 0.01 to 0.04 T are presented. To connect these simulations to observations, we use the computed number densities and temperatures to synthesize the intensities expected in emission lines typically observed with the Extreme Ultraviolet Imaging Spectrometer on Hinode. These intensities are used to compute differential emission measure distributions using the Monte Carlo Markov Chain code, which are very similar to those derived from observations of solar active regions. We conclude that coronal heating is found to be strongly intermittent in space and time, with only small portions of the coronal loop being heated: in fact, at any given time, most of the corona is cooling down.

79 ASTRONOMY AND ASTROPHYSICS↗

Variations of the ISM compactness across the main sequence of star forming galaxies: Observations and simulations

The majority of star-forming galaxies follow a simple empirical correlation in the star formation rate (SFR) versus stellar mass (M{sub *}) plane, of the form SFR∝M{sub ∗}{sup α}, usually referred to as the star formation main sequence (MS). The physics that sets the properties of the MS is currently a subject of debate, and no consensus has been reached regarding the fundamental difference between members of the sequence and its outliers. Here we combine a set of hydro-dynamical simulations of interacting galactic disks with state-of-the-art radiative transfer codes to analyze how the evolution of mergers is reflected upon the properties of the MS. We present Chiburst, a Markov Chain Monte Carlo spectral energy distribution (SED) code that fits the multi-wavelength, broad-band photometry of galaxies and derives stellar masses, SFRs, and geometrical properties of the dust distribution. We apply this tool to the SEDs of simulated mergers and compare the derived results with the reference output from the simulations. Our results indicate that changes in the SEDs of mergers as they approach coalescence and depart from the MS are related to an evolution of dust geometry in scales larger than a few hundred parsecs. This is reflected in a correlation between the specific star formation rate, and the compactness parameter C, that parametrizes this geometry and hence the evolution of dust temperature (T{sub dust}) with time. As mergers approach coalescence, they depart from the MS and increase their compactness, which implies that moderate outliers of the MS are consistent with late-type mergers. By further applying our method to real observations of luminous infrared galaxies (LIRGs), we show that the merger scenario is unable to explain these extreme outliers of the MS. Only by significantly increasing the gas fraction in the simulations are we able to reproduce the SEDs of LIRGs.

79 ASTRONOMY AND ASTROPHYSICS↗

Multi-wavelength observations of GRB 111228A and implications for the fireball and its environment

Observations of very early multi-wavelength afterglows are critical to reveal the properties of the radiating fireball and its environment as well as the central engine of gamma-ray bursts (GRBs). We report our optical observations of GRB 111228A from 95 s to about 50 hr after the burst trigger and investigate its properties of the prompt gamma-rays and the ambient medium using our data and the data from the Swift and Fermi missions. Our joint optical and X-ray spectral fits to the afterglow data show that the ambient medium features a low dust-to-gas ratio. Incorporating the energy injection effect, our best fit to the afterglow light curves with the standard afterglow model via the Markov Chain Monte Carlo technique shows that ϵ{sub e}=(6.9±0.3)×10{sup −2}, ϵ{sub B}=(7.73±0.62)×10{sup −6},E{sub K}=(6.32±0.86)×10{sup 53} erg, n=0.100±0.014 cm{sup −3}. The low medium density likely implies that the afterglow jet may be in a halo or in a hot ISM. A chromatic shallow decay segment observed in the optical and X-ray bands is well explained with the long-lasting energy injection from the central engine, which would be a magnetar with a period of about 1.92 ms inferred from the data. The E{sub p} of its time-integrated prompt gamma-ray spectrum is ∼26 KeV. Using the initial Lorentz factor (Γ{sub 0}=476{sub −237}{sup +225}) derived from our afterglow model fit, it is found that GRB 111228A satisfies the L{sub iso}−E{sub p,z}−Γ{sub 0} relation and bridges the typical GRBs and low luminosity GRBs in this relation.

79 ASTRONOMY AND ASTROPHYSICS↗

MagAO imaging of long-period objects (MILO). I. A benchmark M dwarf companion exciting a massive planet around the sun-like star HD 7449

We present high-contrast Magellan adaptive optics images of HD 7449, a Sun-like star with one planet and a long-term radial velocity (RV) trend. We unambiguously detect the source of the long-term trend from 0.6–2.15 μm at a separation of ∼0.″54. We use the object’s colors and spectral energy distribution to show that it is most likely an M4–M5 dwarf (mass ∼0.1–0.2 M{sub ⊙}) at the same distance as the primary and is therefore likely bound. We also present new RVs measured with the Magellan/MIKE and Planet Finder Spectrograph spectrometers and compile these with archival data from CORALIE and HARPS. We use a new Markov chain Monte Carlo procedure to constrain both the mass (>0.17 M{sub ⊙} at 99% confidence) and semimajor axis (∼18 AU) of the M dwarf companion (HD 7449B). We also refine the parameters of the known massive planet (HD 7449Ab), finding that its minimum mass is 1.09{sub −0.19}{sup +0.52} M{sub J}, its semimajor axis is 2.33{sub −0.02}{sup +0.01} AU, and its eccentricity is 0.8{sub −0.06}{sup +0.08}. We use N-body simulations to constrain the eccentricity of HD 7449B to ≲0.5. The M dwarf may be inducing Kozai oscillations on the planet, explaining its high eccentricity. If this is the case and its orbit was initially circular, the mass of the planet would need to be ≲1.5 M{sub J}. This demonstrates that strong constraints on known planets can be made using direct observations of otherwise undetectable long-period companions.

79 ASTRONOMY AND ASTROPHYSICS↗

How dense is your gas? On the recoverability of LVG model parameters

We explore the recoverability of gas physical conditions with the Large Velocity Gradient (LVG) model, using the public code RADEX and the molecules HCN and CO. Examining a wide parameter range with a series of models of increasing complexity, we use both grid and Monte Carlo Markov Chain methods to recover the input conditions, and quantify the inherent and noise-induced uncertainties in the model results. We find that even with the benefit of generous assumptions, the LVG models struggle to recover any parameter better than to within half a dex, although we find no evidence of systemic offsets. Examining isotopologue lines, we demonstrate that it is always preferable to model the isotopologue abundance ratio as a free parameter due to large biases introduced in all other parameters when an incorrect ratio is assumed. Finally, we explore the effects of the background radiation temperature on CO and HCN line ratios, with an emphasis on the effect of the CMB at z>4, and show that while the effect on the line ratios is minor, the effect on the spectral line energy distribution peak is significant and that the CO(1–0) line luminosity to H{sub 2} mass conversion factor (α{sub CO}) needs to be altered to account for the loss of contrast against the hotter CMB as redshift increases.

79 ASTRONOMY AND ASTROPHYSICS↗

Isotropic at the break? 3D kinematics of Milky Way halo stars in the foreground of M31

We present the line-of-sight (LOS) velocities for 13 distant main sequence Milky Way halo stars with published proper motions (PMs). The PMs were measured using long baseline (5–7 years) multi-epoch Hubble Space Telescope/Advanced Camera for Surveys photometry, and the LOS velocities were extracted from deep (5–6 hr integrations) Keck II/DEIMOS spectra. We estimate the parameters of the velocity ellipsoid of the stellar halo using a Markov chain Monte Carlo ensembler sampler method. The velocity second moments in the directions of the Galactic (l, b, LOS) coordinate system are 〈v{sub l}{sup 2}〉{sup 1/2}=138{sub −26}{sup +43} km s{sup −1}, 〈v{sub b}{sup 2}〉{sup 1/2}=88{sub −17}{sup +28} km s{sup −1}, and 〈v{sub LOS}{sup 2}〉{sup 1/2}=91{sub −14}{sup +27} km s{sup −1}. We use these ellipsoid parameters to constrain the velocity anisotropy of the stellar halo. Ours is the first measurement of the anisotropy parameter β using 3D kinematics outside of the solar neighborhood. We find β=−0.3{sub −0.9}{sup +0.4}, consistent with isotropy and lower than solar neighborhood β measurements by 2σ (β{sub SN} ∼ 0.5–0.7). We identify two stars in our sample that are likely members of the known TriAnd substructure, and excluding these objects from our sample increases our estimate of the anisotropy to β=0.1{sub −1.0}{sup +0.4}, which is still lower than solar neighborhood measurements by 1σ. The potential decrease in β with Galactocentric radius is inconsistent with theoretical predictions, though consistent with recent observational studies, and may indicate the presence of large, shell-type structure (or structures) at r ∼ 25 kpc. The methods described in this paper will be applied to a much larger sample of stars with 3D kinematics observed through the ongoing HALO7D program.

79 ASTRONOMY AND ASTROPHYSICS↗

Cosmological parameters from CMB maps without likelihood approximation

We propose an efficient Bayesian Markov chain Monte Carlo (MCMC) algorithm for estimating cosmological parameters from cosmic microwave background (CMB) data without the use of likelihood approximations. It builds on a previously developed Gibbs sampling framework that allows for exploration of the joint CMB sky signal and power spectrum posterior, P(s,C{sub ℓ}|d), and addresses a long-standing problem of efficient parameter estimation simultaneously in regimes of high and low signal-to-noise ratio. To achieve this, our new algorithm introduces a joint Markov chain move in which both the signal map and power spectrum are synchronously modified, by rescaling the map according to the proposed power spectrum before evaluating the Metropolis–Hastings accept probability. Such a move was already introduced by Jewell et al., who used it to explore low signal-to-noise posteriors. However, they also found that the same algorithm is inefficient in the high signal-to-noise regime, since a brute-force rescaling operation does not account for phase information. This problem is mitigated in the new algorithm by subtracting the Wiener filter mean field from the proposed map prior to rescaling, leaving high signal-to-noise information invariant in the joint step, and effectively only rescaling the low signal-to-noise component. To explore the full posterior, the new joint move is then interleaved with a standard conditional Gibbs move for the sky map. We apply our new algorithm to simplified simulations for which we can evaluate the exact posterior to study both its accuracy and its performance, and find good agreement with the exact posterior; marginal means agree to ≲0.006σ and standard deviations to better than ∼3%. The Markov chain correlation length is of the same order of magnitude as those obtained by other standard samplers in the field.

79 ASTRONOMY AND ASTROPHYSICS↗

Constraining white dwarf structure and neutrino physics in 47 Tucanae

We present a robust statistical analysis of the white dwarf cooling sequence in 47 Tucanae. We combine Hubble Space Telescope UV and optical data in the core of the cluster, Modules for Experiments in Stellar Evolution (MESA) white dwarf cooling models, white dwarf atmosphere models, artificial star tests, and a Markov Chain Monte Carlo sampling method to fit white dwarf cooling models to our data directly. We use a technique known as the unbinned maximum likelihood to fit these models to our data without binning. We use these data to constrain neutrino production and the thickness of the hydrogen layer in these white dwarfs. The data prefer thicker hydrogen layers (q{sub H}=3.2×10{sup −5}) and we can strongly rule out thin layers (q{sub H}=10{sup −6}). The neutrino rates currently in the models are consistent with the data. This analysis does not provide a constraint on the number of neutrino species.

79 ASTRONOMY AND ASTROPHYSICS↗

PROBING THE EXPANSION HISTORY OF THE UNIVERSE BY MODEL-INDEPENDENT RECONSTRUCTION FROM SUPERNOVAE AND GAMMA-RAY BURST MEASUREMENTS

To probe the late evolution history of the universe, we adopt two kinds of optimal basis systems. One of them is constructed by performing the principle component analysis, and the other is built by taking the multidimensional scaling approach. Cosmological observables such as the luminosity distance can be decomposed into these basis systems. These basis systems are optimized for different kinds of cosmological models that are based on different physical assumptions, even for a mixture model of them. Therefore, the so-called feature space that is projected from the basis systems is cosmological model independent, and it provides a parameterization for studying and reconstructing the Hubble expansion rate from the supernova luminosity distance and even gamma-ray burst (GRB) data with self-calibration. The circular problem when using GRBs as cosmological candles is naturally eliminated in this procedure. By using the Levenberg–Marquardt technique and the Markov Chain Monte Carlo method, we perform an observational constraint on this kind of parameterization. The data we used include the “joint light-curve analysis” data set that consists of 740 Type Ia supernovae and 109 long GRBs with the well-known Amati relation.

79 ASTRONOMY AND ASTROPHYSICS↗