Search NASA⌕ Search

SEARCH · Search NASA

Results for “approximation methods”

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 721 records · Page 40

Nonadiabatic Quantum Molecular Dynamics in Dense Manifolds of Electronic States

Most nonadiabatic molecular dynamics methods require the determination of a basis of adiabatic or diabatic electronic states at every time step, but in dense manifolds of electronic states, such approaches become intractable. A notable exception is Ehrenfest molecular dynamics, which can be implemented without explicit determination of such a basis but suffers from unphysical behavior when propagation on a mean-field potential energy surface (PES) does not accurately reflect the true dynamics on multiple electronic states. Here we introduce the multiple cloning for dense manifolds of states (MCDMS) method, a systematically improvable approximation to the multiple cloning method. MCDMS avoids both the mean-field PES problem and the need to compute the full electronic spectrum. This is achieved by reformulating multiple cloning to use a subspace of approximate eigenstates constructed from the time-dependent Ehrenfest electronic wave function. Finally, by application to model systems, we show that this approach allows a substantial reduction in the size of the required electronic basis without significant loss in accuracy.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Resolution of the Gross-Pitaevskii equation with the imaginary-time method on a Lagrange mesh

The Lagrange-mesh method is an approximate variational calculation which has the simplicity of a mesh calculation. Combined with the imaginary-time method, it is applied to the iterative resolution of the Gross-Pitaevskii equation. Two variants of a fourth-order factorization of the exponential of the Hamiltonian and two types of mesh (Lagrange-Hermite and Lagrange-sinc) are employed and compared. The accuracy is checked with the help of these comparisons and of the virial theorem. The Lagrange-Hermite mesh provides very accurate results with short computing times for values of the dimensionless parameter of the nonlinear term up to 10{sup 4}. For higher values up to 10{sup 7}, the Lagrange-sinc mesh is more efficient. Examples are given for anisotropic and nonseparable trapping potentials.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Fractional dynamics of an erbium-doped fiber laser model

In this paper we investigate the model of the time-fractional dynamics of an erbium-doped fiber laser model (TFDEFL) with Liouville–Caputo (LC), Caputo–Fabrizio–Caputo (CFC) and Atangana–Baleanu–Caputo (ABC) time-fractional derivatives. We employ the homotopy analysis transform method (HATM) to calculate approximate solutions for the TFDEFL model. This method gives the solution in the form of a rapidly convergent series that can ensure the convergence in solving the resultant series. We study the convergence analysis of HATM by computing the interval of convergence through the h-curves, the residual error function and the average residual error, respectively. We also show the effectiveness and accuracy of this method by comparing the approximate solutions based upon the LC, CFC and ABC time-fractional derivatives.

36 MATERIALS SCIENCE↗

Perspective on Kramers symmetry breaking and restoration in relativistic electronic structure methods for open-shell systems

Without rigorous symmetry constraints, solutions to approximate electronic structure methods may artificially break symmetry. In the case of the relativistic electronic structure, if time-reversal symmetry is not enforced in calculations of molecules not subject to a magnetic field, it is possible to artificially break Kramers degeneracy in open shell systems. This leads to a description of excited states that may be qualitatively incorrect. Despite this, different electronic structure methods to incorporate correlation and excited states can partially restore Kramers degeneracy from a broken symmetry solution. For single-reference techniques, the inclusion of double and possibly triple excitations in the ground state provides much of the needed correction. Formally, however, this imbalanced treatment of the Kramers-paired spaces is a multi-reference problem, and so methods such as complete-active-space methods perform much better at recovering much of the correct symmetry by state averaging. Using multi-reference configuration interaction, any additional corrections can be obtained as the solution approaches the full configuration interaction limit. A recently proposed “Kramers contamination” value is also used to assess the magnitude of symmetry breaking.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Acceleration of diffusive molecular dynamics simulations through mean field approximation and subcycling time integration

Highlights: • Two numerical methods are presented to accelerate DMD simulations. • The methods greatly reduce the computation time without much loss of accuracy. • The accelerated DMD model is capable of simulating long-term mass transport at atomistic length scales. • Demonstrated for the propagation of an atomically sharp phase boundary over 30 seconds. - Abstract: Diffusive Molecular Dynamics (DMD) is a class of recently developed computational models for the simulation of long-term diffusive mass transport at atomistic length scales. Compared to previous atomistic models, e.g., transition state theory based accelerated molecular dynamics, DMD allows the use of larger time-step sizes, but has a higher computational complexity at each time-step due to the need to solve a nonlinear optimization problem at every time-step. This paper presents two numerical methods to accelerate DMD based simulations. First, we show that when a many-body potential function, e.g., embedded atom method (EAM), is employed, the cost of DMD is dominated by the computation of the mean of the potential function and its derivatives, which are high-dimensional random variables. To reduce the cost, we explore both first- and second-order mean field approximations. Specifically, we show that the first-order approximation, which uses a point estimate to calculate the mean, can reduce the cost by two to three orders of magnitude, but may introduce relatively large error in the solution. We show that adding an approximate second-order correction term can significantly reduce the error without much increase in computational cost. Second, we show that DMD can be significantly accelerated through subcycling time integration, as the cost of integrating the empirical diffusion equation is much lower than that of the optimization solver. To assess the DMD model and the numerical approximation methods, we present two groups of numerical experiments that simulate the diffusion of hydrogen in palladium nanoparticles. In particular, we show that the computational framework is capable of capturing the propagation of an atomically sharp phase boundary over a time window of more than 30 seconds. The effects of the proposed numerical methods on solution accuracy and computation time are also assessed quantitatively.

36 MATERIALS SCIENCE↗

Deuteron microscopic optical model potential

A deuteron microscopic optical model potential is obtained by the Green function method through nuclear-matter approximation and local-density approximation based on the effective Skyrme interaction. The microscopic optical model potential is used to calculate the deuteron reaction cross sections and the elastic scattering angular distributions for some target nuclei in the mass range 6<=A<=208 with incident deuteron energies up to 200 MeV. The calculated results are compared with the experimental data.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Communication-efficient Distributed Solutions to a System of Linear Equations with Laplacian Sparse Structure

Two communication-efficient distributed algorithms are proposed to solve a system of linear equations Ax = b with Laplacian sparse A. A system of linear equations with Laplacian sparse A can be found in many applications, e.g., the power flow problems and other network flow problems. The first algorithm is based on the gradient descent method in optimization and the agents only share two parts of the system state instead of that of the whole system state, which saves significant communication. The two parts shared by every agent through a communication link are the state information of its own and its neighbors connected by the communication link. The second method is obtained from an approximation of the Newton method, which converges faster. It requires twice as much communication as the first one but is still communicationefficient due to the low dimension of each part shared between agents. The convergence at a linear rate of both methods is proved. A comprehensive comparison of the convergence rate, communication burden, and computation costs between the methods is made. Finally, a simulation is conducted to show the effectiveness of both methods.

Wang, Peng↗

A Numerical Comparison of Rule Ensemble Methods and Support Vector Machines

Machine or statistical learning is a growing field that encompasses many scientific problems including estimating parameters from data, identifying risk factors in health studies, image recognition, and finding clusters within datasets, to name just a few examples. Statistical learning can be described as 'learning from data' , with the goal of making a prediction of some outcome of interest. This prediction is usually made on the basis of a computer model that is built using data where the outcomes and a set of features have been previously matched. The computer model is called a learner, hence the name machine learning. In this paper, we present two such algorithms, a support vector machine method and a rule ensemble method. We compared their predictive power on three supernova type 1a data sets provided by the Nearby Supernova Factory and found that while both methods give accuracies of approximately 95%, the rule ensemble method gives much lower false negative rates.

97 MATHEMATICS AND COMPUTING↗

Stochastic modeling of catalytic processes in nanoporous materials: Beyond mean-field approach

Transport and reaction in zeolites and other porous materials, such as mesoporous silica particles, has been a focus of interest in recent years. This is in part due to the possibility of anomalous transport effects (e.g. single-file diffusion) and its impact in the reaction yield in catalytic processes. Computational simulations are often used to study these complex nonequilibrium systems. Computer simulations using Molecular Dynamics (MD) techniques are prohibitive, so instead coarse grained one-dimensional models with the aid of Kinetic Monte Carlo (KMC) simulations are used. Both techniques can be computationally expensive, both time and resource wise. These coarse-grained systems can be exactly described by a set of coupled stochastic master equations, that describe the reaction-diffusion kinetics of the system. The equations can be written exactly, however, coupling between the equations and terms within the equations make it impossible to solve them exactly; approximations must be made. One of the most common methods to obtain approximate solutions is to use Mean Field (MF) theory. MF treatments yield reasonable results at high ratios of reaction rate k to hop rate h of the particles, but fail completely at low k=h due to the over-estimation of fluxes of particles within the pore. We develop a method to estimate fluxes and intrapore diffusivity in simple one- dimensional reaction-diffusion models at high and low k=h, where the pores are coupled to an equilibrated three-dimensional fluid. We thus successfully describe analytically these simple reaction-diffusion one-dimensional systems. Extensions to models considering behavior with long range steric interactions and wider pores require determination of multiple boundary conditions. We give a prescription to estimate the required parameters for these simulations. For one dimensional systems, if single-file diffusion is relaxed, additional parameters to describe particle exchange have to be introduced. We use Langevin Molecular Dynamics (MD) simulations to assess these parameters.

36 MATERIALS SCIENCE↗

Analysis and Simulation of One-Dimensional Transport Models for Lithium Symmetric Cells

The main objective of this paper is to analyze transport models for lithium symmetric cells and arrive at an efficient model and code to simulate the same. Two one-dimensional models are considered. The first model uses the dilute solution Nernst–Planck equations in conjunction with the electroneutrality assumption (EN-NP model). For binary electrolytes, an analytical solution for electrolyte concentration is derived and compared with numerical solutions by an approximate finite volume method. The second approach relaxes the electroneutrality assumption by way of Poisson’s equation for the electrostatic potential (PNP model). The computational difficulties of the PNP model are tackled using the approximate finite volume method, with demonstrated convergence characteristics even with bulk dimensions several orders of magnitude larger than the characteristic double layer size. A robust code is developed for the PNP model. The computationally efficient transport models can facilitate simulations, physical understanding, and analysis. This is illustrated by a case study in which these models are coupled with modified kinetic models, which are then parameterized with experimental voltage response data using a systems-level approach. The estimated parameters provide further insight into the electrochemical phenomena underpinning voltage transitions in symmetric cells.

25 ENERGY STORAGE↗

Heuristic methods and performance bounds for photonic design

In the photonic design problem, a scientist or engineer chooses the physical parameters of a device to best match some desired device behavior. Many instances of the photonic design problem can be naturally stated as a mathematical optimization problem that is computationally difficult to solve globally. Because of this, several heuristic methods have been developed to approximately solve such problems. These methods often produce very good designs, and, in many practical applications, easily outperform ‘traditional’ designs that rely on human intuition. Yet, because these heuristic methods do not guarantee that the approximate solution found is globally optimal, the question remains of just how much better a designer might hope to do. This question is addressed by performance bounds or impossibility results, which determine a performance level that no design can achieve. We focus on algorithmic performance bounds, which involve substantial computation to determine. We illustrate a variety of both heuristic methods and performance bounds on two examples. In these examples (and many others not reported here) the performance bounds show that the heuristic designs are nearly optimal, and can be considered globally optimal in practice. This review serves to clearly set up the photonic design problem and unify existing approaches for calculating performance bounds, while also providing some natural generalizations and properties.

Angeris, Guillermo (ORCID:0000000249503990)↗

Approximation of periodic functions in the classes H{sub q}{sup {Omega}} by linear methods

The following result is proved: if approximations in the norm of L{sub {infinity}} (of H{sub 1}) of functions in the classes H{sub {infinity}}{sup {Omega}} (in H{sub 1}{sup {Omega}}, respectively) by some linear operators have the same order of magnitude as the best approximations, then the set of norms of these operators is unbounded. Also Bernstein's and the Jackson-Nikol'skii inequalities are proved for trigonometric polynomials with spectra in the sets Q(N) (in {Gamma}(N,{Omega})). Bibliography: 15 titles.

97 MATHEMATICS AND COMPUTING↗

Towards Off-policy Evaluation as a Prerequisite for Real-world Reinforcement Learning in Building Control

We present an initial study of off-policy evaluation (OPE), a problem prerequisite to real-world reinforcement learning (RL), in the context of building control. OPE is the problem of estimating a policy's performance without running it on the actual system, using historical data from the existing controller. It enables the control engineers to ensure a new, pretrained policy satisfies the performance requirements and safety constraints of a real-world system, prior to interacting with it. While many methods have been developed for OPE, no study has evaluated which ones are suitable for building operational data, which are generated by deterministic policies and have limited coverage of the state-action space. After reviewing existing works and their assumptions, we adopted the approximate model (AM) method. Furthermore, we used bootstrapping to quantify uncertainty and correct for bias. In a simulation study, we evaluated the proposed approach on 10 policies pretrained with imitation learning. On average, the AM method estimated the energy and comfort costs with 1.84% and 14.1% error, respectively.

Chen, B↗

Analyzing the Quantum Approximate Optimization Algorithm: Ansätze, Symmetries, and Lie Algebras

The quantum approximate optimization algorithm (QAOA) has been proposed as a method to obtain approximate solutions for combinatorial optimization tasks. In this work, we study the underlying algebraic properties of three QAOA ansätze for the maximum-cut problem on connected graphs, while focusing on the generated Lie algebras as well as their invariant subspaces. Specifically, we analyze the standard QAOA ansatz as well as the orbit and multiangle ansätze. We are able to fully characterize the Lie algebras of the multiangle ansatz across arbitrary connected graphs, finding that they only fall into one of just six families. Aside from the cycle and path graphs, the Lie dimensions for every graph are exponentially large in the system size, meaning that multiangle ansätze are extremely prone to exhibiting barren plateaus. Then, a similar quasi-graph-independent Lie-algebraic characterization beyond the multiangle ansatz is impeded as the circuit exhibits additional “hidden” symmetries besides those naturally arising from a certain parity-superselection operator and all automorphisms of the considered graph. Disregarding the “hidden” symmetries, we can upper bound the dimensions of the orbit and the standard Lie algebras, and the dimensions of the associated invariant subspaces are determined via explicit character formulas. To finish, we conjecture that (for most graphs) the standard Lie algebras have only components that are either exponential or that grow, at most, polynomially with the system size. This would imply that the QAOA is either prone to barren plateaus or classically simulable. More generally, our work provides a symmetry framework and tools to analyze any desired variational quantum algorithm.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Fast iterative refinement method for mixed systems of integral and fractional integro-differential equations

The authors’ new method of iterative refinement has been successfully applied to the linear fractional Fredholm integro-differential equation. In this work, we adapt our method to obtain approximate analytical solutions for linear Volterra equations of both the first and the second kind of fractional type. A detailed convergence analysis is presented for each kind. The authors also apply their method to the challenging first kind linear Fredholm integral equation. Three different cases are considered to test the efficacy of our method. These include mixed systems of various forms of linear integral and fractional integro-differential equations. We compare our method with six well-established methods: Picard’s successive approximations, an accelerated form of the latter, the Adomian decomposition, the variational iteration, the homotopy perturbation and the homotopy analysis. The results show the versatility of the new method in solving a wide variety of equations, its high convergence speed and accuracy and its capability of directly dealing with equations of the first kind. We also prove that Picard’s method for linear equations is a special case of our method.

97 MATHEMATICS AND COMPUTING↗

Two-dimensional gap solitons in elliptic-lattice potentials

We study two-dimensional (2D) matter-wave gap solitons trapped in an elliptically deformed concentric lattice potential, within the framework of the Gross-Pitaevskii equation (GPE) with self-attraction or self-repulsion. For a fixed eccentricity of the lattice, soliton families are found in both the repulsive and attractive models. In the former case, the analysis reveals two kinds of gap solitons trapped in the first oval trough (the ring-shaped potential minimum closest to the center): elliptic annular solitons (EASs), and double solitons (DSs), which are formed by two tightly localized density peaks located at diametrically opposite points of the trough, with zero phase difference between them. With the decrease of the norm, the density distribution in the EAS along the azimuthal direction changes from nearly uniform to double-peaked and, eventually, to the DS. In the attractive model, there exist only DSs in the oval trough, while EASs are not found. All such solitons without the angular momentum (l=0) are fully stable. For l{ne}0, vortical solitons--both EASs with a sufficiently large norm (in the repulsive model) and DSs (in models with both signs of the nonlinearity)--are quasistable, exhibiting rocking motion in the elliptic trough (we consider the cases of l=1 and l=2). At smaller values of the norm, the vortical annular solitons (in the repulsive model) are unstable. Stable fundamental solitons trapped in the central potential well are investigated, too, in both the attractive and repulsive models, by means of the variational approximation and numerical methods.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Rapid chain tracing of polypeptide backbones in electron-density maps

A method for rapid chain tracing of polypeptide backbones at moderate resolution is presented. A method for the rapid tracing of polypeptide backbones has been developed. The method creates an approximate chain tracing that is useful for visual evaluation of whether a structure has been solved and for use in scoring the quality of electron-density maps. The essence of the method is to (i) sample candidate C{sup α} positions at spacings of approximately 0.6 Å along ridgelines of high electron density, (ii) list all possible nonapeptides that satisfy simple geometric and density criteria using these candidate C{sup α} positions, (iii) score the nonapeptides and choose the highest scoring ones, and (iv) find the longest chains that can be made by connecting nonamers. An indexing and storage scheme that allows a single calculation of most distances and density values is used to speed up the process. The method was applied to 42 density-modified electron-density maps at resolutions from 1.5 to 3.8 Å. A total of 21 428 residues in these maps were traced in 24 CPU min with an overall r.m.s.d. of 1.61 Å for C{sup α} atoms compared with the known refined structures. The method appears to be suitable for rapid evaluation of electron-density map quality.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Electrical conductivity of a two-dimensional model for a structurally anisotropic composite

The electrical conductivity of a two-dimensional structurally anisotropic model for a composite is considered. The model represents an isotropic matrix with a system of nonconducting inclusions in the form of infinitely thin straight line segments (scratches). The scratches make an angle {theta} or -{theta} with a preferred axis (for definiteness, axis y) at the same probability, and their centers are chaotically distributed. An approximate effective medium method is used to obtain a general expression for the effective conductivity tensor of this model that is valid over a wide concentration range. In this approximation, both components of tensor are shown to vanish at the same percolation threshold, which is expressed explicitly. The conductivity of the model in a critical region is considered in terms of the similarity hypothesis.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗