Search NASA⌕ Search

SEARCH · Search NASA

Results for “variational method”

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 559 records · Page 31

Adapting Reinforcement Learning Treatment Policies Using Limited Data to Personalize Critical Care

We report reinforcement learning (RL) demonstrates promise for developing effective treatment policies in critical care settings. However, existing RL methods often require large and comprehensive patient data sets and do not readily lend themselves to settings in which certain patient subpopulations are severely underrepresented. In this study, we develop a new method, noisy Bayesian policy updates (NBPU), for selecting high-performing reinforcement learning–based treatment policies for underrepresented patient subpopulations using limited observations. Our method uses variational inference to learn a probability distribution over treatment policies based on a reference patient subpopulation for which sufficient data are available. It then exploits limited data from an underrepresented patient subpopulation to update this probability distribution and adapts its recommendations to this subpopulation. We demonstrate our method’s utility on a data set of ICU patients receiving intravenous blood anticoagulant medication. Our results show that NBPU outperforms state-of-the-art methods in terms of both selecting effective treatment policies for patients with nontypical clinical characteristics and predicting the corresponding policies’ performance for these patients.

60 APPLIED LIFE SCIENCES↗

Numerical schemes for dynamically orthogonal equations of stochastic fluid and ocean flows

The quantification of uncertainties is critical when systems are nonlinear and have uncertain terms in their governing equations or are constrained by limited knowledge of initial and boundary conditions. Such situations are common in multiscale, intermittent and non-homogeneous fluid and ocean flows. The dynamically orthogonal (DO) field equations provide an adaptive methodology to predict the probability density functions of such flows. The present work derives efficient computational schemes for the DO methodology applied to unsteady stochastic Navier-Stokes and Boussinesq equations, and illustrates and studies the numerical aspects of these schemes. Semi-implicit projection methods are developed for the mean and for the DO modes, and time-marching schemes of first to fourth order are used for the stochastic coefficients. Conservative second-order finite-volumes are employed in physical space with new advection schemes based on total variation diminishing methods. Other results include: (i) the definition of pseudo-stochastic pressures to obtain a number of pressure equations that is linear in the subspace size instead of quadratic; (ii) symmetric advection schemes for the stochastic velocities; (iii) the use of generalized inversion to deal with singular subspace covariances or deterministic modes; and (iv) schemes to maintain orthonormal modes at the numerical level. To verify our implementation and study the properties of our schemes and their variations, a set of stochastic flow benchmarks are defined including asymmetric Dirac and symmetric lock-exchange flows, lid-driven cavity flows, and flows past objects in a confined channel. Different Reynolds number and Grashof number regimes are employed to illustrate robustness. Optimal convergence under both time and space refinements is shown as well as the convergence of the probability density functions with the number of stochastic realizations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Density functional study on electronic and optical properties of C (or N)-doped cubic cerium dioxide

Density functional calculations were performed on electronic and optical properties of C (or N)-doped cubic cerium dioxide (CeO{sub 2}). When O is replaced by C (or N) in CeO{sub 2}, obvious band-gap (E{sub g}) reduction is observed. Meanwhile, it is interesting to find that the substitutional doping of C (or N) in CeO{sub 2} obviously increases the O 2p-Ce 4f transition intensity and also the refractive index. The increase in the O 2p-Ce 4f transition intensity on going from undoped, N-doped and C-doped CeO{sub 2} was related to the covalent character of the Ce-O bond. Compared with the undoped CeO{sub 2}, the C (or N)-doped CeO{sub 2}, with steep absorption peaks at lower energy, can be used for visible-light absorption applications. - Graphical abstract: By first-principles calculations, it is predicted that substitutional doping of C (or N) in c-CeO{sub 2} lowers E{sub g} and also O 2p-Ce 4f gap, and increases O 2p-Ce 4f transition intensity.

36 MATERIALS SCIENCE↗

Charge constrained density functional molecular dynamics for simulation of condensed phase electron transfer reactions

We present a plane-wave basis set implementation of charge constrained density functional molecular dynamics (CDFT-MD) for simulation of electron transfer reactions in condensed phase systems. Following the earlier work of Wu and Van Voorhis [Phys. Rev. A 72, 024502 (2005)], the density functional is minimized under the constraint that the charge difference between donor and acceptor is equal to a given value. The classical ion dynamics is propagated on the Born-Oppenheimer surface of the charge constrained state. We investigate the dependence of the constrained energy and of the energy gap on the definition of the charge and present expressions for the constraint forces. The method is applied to the Ru{sup 2+}-Ru{sup 3+} electron self-exchange reaction in aqueous solution. Sampling the vertical energy gap along CDFT-MD trajectories and correcting for finite size effects, a reorganization free energy of 1.6 eV is obtained. This is 0.1-0.2 eV lower than a previous estimate based on a continuum model for solvation. The smaller value for the reorganization free energy can be explained by the fact that the Ru-O distances of the divalent and trivalent Ru hexahydrates are predicted to be more similar in the electron transfer complex than for the separated aqua ions.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Influence of density on N-H bond stretch vibration in plasma enhanced chemical vapor deposited SiN{sub x}:H

The infrared absorption of hydrogenated silicon nitride (SiN{sub x}:H) films provides information about the average configuration of covalent bonds in the film. In plasma enhanced chemical vapor deposited SiN{sub x}:H, the absorption spectrum of the N-H bonds shows both a systematic shift and a tailing toward lower frequency with the increase in the film density. First-principles computations of N-H vibrations show an increased degree of anharmonicity when the H site is weakly interacting with a second neighboring N atom. This interaction becomes more pronounced in films with higher density, which consequently shifts the vibrational spectrum down.

36 MATERIALS SCIENCE↗

Ab initio potentials of F+Li{sub 2} accessible at ultracold temperatures

Ab initio calculations for the strongly exoergic Li{sub 2}+F harpoon reaction are presented using density-functional theory, complete active space self-consistent field, and multireference configuration interaction methods to argue that this reaction would be an ideal candidate for investigation with ultracold molecules. The lowest six states are calculated with the aug-correlation-consistent polarized valence triple-zeta basis set and at least two can be accessed by a ground rovibronic Li{sub 2} molecule with zero collision energy at all reaction geometries. The large reactive cross section (characteristic of harpoon reactions) and chemiluminescent products are additional attractive features of these reactions.

74 ATOMIC AND MOLECULAR PHYSICS↗

Spin-polarized self-consistent-field equations for paired orbitals

Unrestricted Hartree-Fock-like equations are proposed to find multiple spin-symmetry-broken states of the molecular systems. Developed equations are pseudo-eigenvalue-type equations for the Fock-type operators constructed in such a way to include an effective field which makes different-spin orbitals biorthogonal. The eigenvectors of these operators are noncanonical Hartree-Fock orbitals becoming Loewdin-Amos-Hall paired (corresponding) orbitals after self-consistency is achieved. The eigenvalues of the modified Fock operators appear to be the energies of the paired orbitals. Because the paired orbitals do not follow the spatial symmetry of the molecular nuclear core, the equations allow one to obtain the broken symmetry states with relative ease as demonstrated for the model H{sub 6} hexagon molecule. For this molecule, the Cizek-Paldus instability matrix analysis predicts the existence of three spin-symmetry-broken states. All these solutions are systematically achieved by the paired equations, unlike the standard unrestricted equations which basically converge to a single solution. The proposed approach is also valid for the density functional theory in which the spin-polarized Kohn-Sham equations might be transformed to paired equations.

74 ATOMIC AND MOLECULAR PHYSICS↗

Density functional theory beyond the linear regime: Validating an adiabatic local density approximation

We present a local density approximation (LDA) for one-dimensional (1D) systems interacting via the soft-Coulomb interaction based on quantum Monte Carlo calculations. Results for the ground-state energies and ionization potentials of finite 1D systems show excellent agreement with exact calculations obtained by exploiting the mapping of an N-electron system in d dimensions onto a single electron in Nxd dimensions, properly symmetrized by the Young diagrams. We conclude that 1D LDA is of the same quality as its three-dimensional (3D) counterpart, and we infer conclusions about 3D LDA. The linear and nonlinear time-dependent responses of 1D model systems using LDA, exact exchange, and the exact solution are investigated and show very good agreement in both cases, except for the well-known problem of missing double excitations. Consequently, the 3D LDA is expected to be of good quality beyond the linear response. In addition, the 1D LDA should prove useful in modeling the interaction of atoms with strong laser fields, where this specific 1D model is often used.

74 ATOMIC AND MOLECULAR PHYSICS↗

Density functional, density matrix functional, and the virial theorem

On the basis of the virial theorem for finite motion of interacting electrons in a static external field, it is shown that ground-state energy is determined evidently by the one-particle density matrix. The exact expression for ground-state energy is found and can be applied to various problems, in particular, in studying the electronic structure of atoms, molecules, and other bound complexes.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

High-order-harmonic generation in homonuclear and heteronuclear diatomic molecules: Exploration of multiple orbital contributions

We present a time-dependent density functional theory (TDDFT) approach with proper asymptotic long-range potential for nonperturbative treatment of high-order harmonic generation (HHG) of diatomic molecules with their molecular axis parallel to the laser field polarization. A time-dependent two-center generalized pseudospectral method in prolate spheroidal coordinate system is used for accurate and efficient treatment of the TDDFT equations in space and time. The theory is applied to a detailed all-electron nonperturbative investigation of HHG processes of homonuclear (N{sub 2} and F{sub 2}) and heteronuclear (CO, BF, and HF) molecules in intense ultrashort laser pulses with the emphasis on the role of multiple molecular orbitals (MOs). The results reveal intriguing and substantially different nonlinear optical response behaviors for homonuclear and heteronuclear molecules. In particular, we found that the HHG spectrum for homonuclear molecules features a destructive interference of MO contributions while heteronuclear molecules show mostly constructive interference of orbital contributions.

74 ATOMIC AND MOLECULAR PHYSICS↗

Effects of metal impurities on the optical properties of polyethylene in the warm dense-matter regime

For the warm dense-matter regime, we have examined the effects on the equation of state, conductivity, and optical properties of the introduction of a metal impurity (Al) into a poorly conducting mixture (CH{sub 2}) by means of quantum molecular-dynamics simulations employing temperature-dependent density-functional theory for temperatures between 1 and 4 eV and densities from ambient solid to a few times compressed. The properties such as dc conductivity and Rosseland mean opacity exhibit significant departures from the pure CH{sub 2} results only for metal concentrations above about 50%. The system is representative of a wide range of environments that include planetary interiors, inertial confinement fusion capsules, and laser-produced plasmas.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Band offsets in cubic GaN/AlN superlattices

The presently unknown band offset in nonpolar cubic GaN/AlN superlattices is investigated by inter-sub-band and interband spectroscopies as well as ab initio calculations. On one hand, the conduction-band offset (CBO) has been determined from the comparison of the measured transition energies with model calculations within the effective mass approximation. On the other hand, the valence-band offset (VBO) and the CBO are accurately simulated by calculating many-body corrections within the GW approximation on top of hybrid-functional density functional theory calculations. Thus, a CBO of (1.4{+-}0.1) eV and a VBO of (0.5{+-}0.1) eV is obtained as a result of both approaches.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

DFT calculations of molecular excited states using an orbital-dependent nonadiabatic exchange kernel

A density functional method for computing molecular excitation spectra is presented that uses a frequency-dependent kernel and takes into account the nonlocality of exchange interaction. Owing to its high numerical stability and the use of a nonadiabatic (frequency-dependent) exchange kernel, the proposed approach provides a qualitatively correct description of the asymptotic behavior of charge-transfer excitation energies.

74 ATOMIC AND MOLECULAR PHYSICS↗

An Incremental Gradient Method for Optimization Problems With Variational Inequality Constraints

We consider minimizing a sum of agent-specific nondifferentiable merely convex functions over the solution set of a variational inequality (VI) problem in that each agent is associated with a local monotone mapping. This problem finds an application in computation of the best equilibrium in nonlinear complementarity problems arising in transportation networks. We develop an iteratively regularized incremental gradient method where at each iteration, agents communicate over a directed cycle graph to update their solution iterates using their local information about the objective and the mapping. The proposed method is single-timescale in the sense that it does not involve any excessive hard-to-project computation per iteration. We derive nonasymptotic agent-wise convergence rates for the suboptimality of the global objective function and infeasibility of the VI constraints measured by a suitably defined dual gap function. Finally, the proposed method appears to be the first fully iterative scheme equipped with iteration complexity that can address distributed optimization problems with VI constraints over cycle graphs.

convergence↗

The study of disorder and nanocrystallinity in C-S-H, supplementary cementitious materials and geopolymers using pair distribution function analysis

Significant progress was achieved with the application of Rietveld method to characterize the crystalline phases in portland cement paste. However, to obtain detailed information on the amorphous or poorly crystalline phases, it is necessary to analyze the total scattering data. The pair distribution function (PDF) method has been successfully used in the study of liquids and amorphous solids. The method takes the Sine Fourier transform of the measured structure factor over a wide momentum transfer range, providing a direct measure of the probability of finding an atom surrounding a central atom at a radial distance away. The obtained experimental characteristic distances can be also used to validate the predictions by the theoretical models, such as, molecular dynamics, ab initio simulations and density functional theory. The paper summarizes recent results of PDF analysis on silica fume, rice husk ash, fly ash, ASR gel, C-S-H and geopolymers.

36 MATERIALS SCIENCE↗

Exact-exchange kernel of time-dependent density functional theory: Frequency dependence and photoabsorption spectra of atoms

In this work we have calculated excitation energies and photoionization cross sections of Be and Ne in the exact-exchange (EXX) approximation of time-dependent density functional theory (TDDFT). The main focus has been on the frequency dependence of the EXX kernel and on how it affects the spectrum as compared to the corresponding adiabatic approximation. We show that for some discrete excitation energies the frequency dependence is essential to reproduce the results of time-dependent Hartree-Fock theory. Unfortunately, we have found that the EXX approximation breaks down completely at higher energies, producing a response function with the wrong analytic structure and making inner-shell excitations disappear from the calculated spectra. We have traced this failure to the existence of vanishing eigenvalues of the Kohn-Sham non-interacting response function. Based on the adiabatic TDDFT formalism we propose a new way of deriving the Fano parameters of autoionizing resonances.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

First principles theoretical study of the hole-assisted conversion of CO to CO{sub 2} on the anatase TiO{sub 2}(101) surface

First principles density functional theory calculations were carried out to investigate the adsorption and oxidation of CO on the positively charged (101) surface of anatase, as well as the desorption of CO{sub 2} from it. We find that the energy gain on adsorption covers the activation energy required for the oxidation, while the energy gain on the latter is sufficient for the desorption of CO{sub 2}, leaving an oxygen vacancy behind. Molecular dynamics simulations indicate that the process can be spontaneous at room temperature. The oxidation process described here happens only in the presence of the hole. The possibility of a photocatalytic cycle is discussed assuming electron scavenging by oxygen.

36 MATERIALS SCIENCE↗

Optimal control of the electronic current density: Application to one- and two-dimensional one-electron systems

Quantum optimal control theory is a powerful tool for engineering quantum systems subject to external fields such as the ones created by intense lasers. The formulation relies on a suitable definition for a target functional, that translates the intended physical objective to a mathematical form. We propose the use of target functionals defined in terms of the one-particle density and its current. A strong motivation for this is the possibility of using time-dependent density-functional theory for the description of the system dynamics. We exemplify this idea by defining an objective functional that on one hand attempts a large overlap with a target density and on the other hand minimizes the current. The latter requirement leads to optimized states with increased stability, which we prove with a few examples of one- and two-dimensional one-electron systems.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗