Search NASA⌕ Search

SEARCH · Search NASA

Results for “subspace 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 73 records · Page 4

A projection preconditioner for solving the implicit immersed boundary equations

This study presents a method for solving the linear semi-implicit immersed boundary equations which avoids the severe time step restriction presented by explicit-time methods. The Lagrangian variables are eliminated via a Schur complement to form a purely Eulerian saddle point system, which is preconditioned by a projection operator and then solved by a Krylov subspace method. From the viewpoint of projection methods, we derive an ideal preconditioner for the saddle point problem and compare the efficiency of a number of simpler preconditioners that approximate this perfect one. For low Reynolds number and high stiffness, one particular projection preconditioner yields an efficiency improvement of the explicit IB method by a factor around thirty. Substantial speed-ups over explicit-time method are achieved for Reynolds number below 100. In conclusion, this speedup increases as the Eulerian grid size and/or the Reynolds number are further reduced.

42 ENGINEERING↗

Nonlinear acceleration of SN transport calculations

The use of nonlinear iterative methods, Jacobian-Free Newton-Krylov (JFNK) in particular, for solving eigenvalue problems in transport applications has recently become an active subject of research. While JFNK has been shown to be effective for k-eigenvalue problems, there are a number of input parameters that impact computational efficiency, making it difficult to implement efficiently in a production code using a single set of default parameters. We show that different selections for the forcing parameter in particular can lead to large variations in the amount of computational work for a given problem. In contrast, we present a nonlinear subspace method that sits outside and effectively accelerates nonlinear iterations of a given form and requires only a single input parameter, the subspace size. It is shown to consistently and significantly reduce the amount of computational work when applied to fixed-point iteration, and this combination of methods is shown to be more efficient than JFNK for our application.

97 MATHEMATICS AND COMPUTING↗

$O(N)$ ab initio calculation scheme for large-scale moiré structures

Here we present a two-step method specifically tailored for band structure calculation of the small-angle moiré-pattern materials which contain tens of thousands of atoms in a unit cell. In the first step, the self-consistent field calculation for the ground state is performed with the O(N) Krylov subspace method implemented in openmx. Second, the crystal momentum-dependent Bloch Hamiltonian and overlap matrix are constructed from the results obtained in the first step and only a small number of eigenvalues near the Fermi energy are solved with shift-invert and Lanczos techniques. By systematically tuning two key parameters, the cutoff radius for electron hopping interaction and the dimension of the Krylov subspace, we obtained the band structures for both rigid and corrugated twisted bilayer graphene structures down to the first magic angle (θ = 1.08°) with high enough accuracy at affordable costs. The band structures are in good agreement with those from tight-binding models, continuum models, plane-wave pseudopotential based ab initio calculations, and experimental observations. This method is also shown to be efficient in twisted double-bilayer graphene and bilayer WSe 2 . We think this two-step method can play a crucial role in other twisted two-dimensional materials, especially those with much more complex band structure and where the effective model is hard to construct.

36 MATERIALS SCIENCE↗

Avoiding Communication in Primal and Dual Block Coordinate Descent Methods

Primal and dual block coordinate descent methods are iterative methods for solving regularized and unregularized optimization problems. Distributed-memory parallel implementations of these methods have become popular in analyzing large machine learning datasets. However, existing implementations communicate at every iteration, which, on modern data center and supercomputing architectures, often dominates the cost of floating-point computation. Recent results on communication-avoiding Krylov subspace methods suggest that large speedups are possible by re-organizing iterative algorithms to avoid communication. We show how applying similar algorithmic transformations can lead to primal and dual block coordinate descent methods that only communicate every $s$ iterations---where $s$ is a tuning parameter---instead of every iteration for the regularized least-squares problem. We show that the communication-avoiding variants reduce the number of synchronizations by a factor of $s$ on distributed-memory parallel machines without altering the convergence rate and attain strong scaling speedups of up to $6.1\times$ over the “standard algorithm" on a Cray XC30 supercomputer.

97 MATHEMATICS AND COMPUTING↗

An Efficient Sampling-Based Method for Sensitivity and Uncertainty Analysis Through RAVEN

Sensitivity analysis (SA) and uncertainty quantification (UQ) are now widely recognized as essential parts of analyses for complex nuclear systems. Plenty of well-established methodologies are already available for performing SA and UQ, such as Monte Carlo analysis, adjoint-based method, and variance decomposition. The focus of this summary is on developing an efficient sampling-based approach to perform SA and UQ through RAVEN. RAVEN (Risk Analysis and Virtual control Environment) is a software framework able to perform SA and UQ, currently under development at the Idaho National Laboratory. In literature, a number of approaches have been developed to accelerate the process of SA and UQ for large complex systems, such as principal component analysis (PCA) of input space, input-output correlation analysis, and intersection subspace methods. The fundamental idea is to identify the dominant active directions in the parameter space prior to conducting SA and UQ. The premise is that the number of active directions is much less than the dimensionality of input space. In this work, the PCA or Karhunen-Loeve expansion or SVD method is implemented in RAVEN to identify the active directions. This approach is also combined with the forward sampling approaches, such as Monte Carlo, Grid, Latin Hypercube, and Response Surface Design sampling approaches that are currently available in RAVEN, to perform SA and UQ more efficiently. The summary is organized as follows. In the following section, we present the PCA-based forward sampling approach. The effectiveness and efficiency are discussed in the numerical results section. The paper then ends with a concluding summary. In this summary, we have presented PCA-based forward sampling approach to perform SA and UQ. The goal is to reduce the computational overhead for repetitive executions of complex nuclear systems that are generally required when performing SA and UQ. The control of correlations between input parameters is implemented in RAVEN to better represent the input relationships. In addition, we are currently examining our implementations with more realistic models, and we are also focusing on implementation of some new high-dimensional input reduction techniques in RAVEN. (authors)

97 MATHEMATICS AND COMPUTING↗

Fast Simulating High Order Models Application to Micro Electro-Mechanical Systems (MEMS)

The approximation of high order systems by low order models is one of the important problems in system theory. The use of a reduced order model makes it easier to implement analysis, simulations and control system designs. Numerous methods are available in the literature for order reduction of linear continuous systems in time domain as well as in frequency domain. But, this is not the case for non linear systems. The well known Trajectory Piece-Wise Linear approach (TPWL) elaborated to nonlinear model order reduction guarantees a simplification and an accurate representation of the behaviour of strongly non linear systems handling local and global approximation. The present attempt is towards evolving an improvement for the TPWL order reduction technique, which ensures a good quality of approximation combining the advantages of the Krylov subspaces method and the local linearization. We illustrate the technique on a MEMS circuit (Micro Electro-Mechanical System)

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A time-parallel method for the solution of PDE-constrained optimization problems

We study a time-parallel approach to solving quadratic optimization problems with linear time-dependent partial differential equation (PDE) constraints. These problems arise in formulations of optimal control, optimal design and inverse problems that are governed by parabolic PDE models. They may also arise as subproblems in algorithms for the solution of optimization problems with nonlinear time-dependent PDE constraints, e.g., in sequential quadratic programming methods. We apply a piecewise linear finite element discretization in space to the PDE constraint, followed by the Crank-Nicolson discretization in time. The objective function is discretized using finite elements in space and the trapezoidal rule in time. At this point in the discretization, auxiliary state variables are introduced at each discrete time interval, with the goal to enable: (i) a decoupling in time; and (ii) a fixed-point iteration to recover the solution of the discrete optimality system. The fixed-point iterative schemes can be used either as preconditioners for Krylov subspace methods or as smoothers for multigrid (in time) schemes. We present promising numerical results for both use cases.

97 MATHEMATICS AND COMPUTING↗

Rapid sampling of stochastic displacements in Brownian dynamics simulations

Here, we present a new method for sampling stochastic displacements in Brownian Dynamics (BD) simulations of colloidal scale particles. The method relies on a new formulation for Ewald summation of the Rotne-Prager-Yamakawa (RPY) tensor, which guarantees that the real-space and wave-space contributions to the tensor are independently symmetric and positive-definite for all possible particle configurations. Brownian displacements are drawn from a superposition of two independent samples: a wave-space (far-field or long-ranged) contribution, computed using techniques from fluctuating hydrodynamics and non-uniform fast Fourier transforms; and a real-space (near-field or short-ranged) correction, computed using a Krylov subspace method. The combined computational complexity of drawing these two independent samples scales linearly with the number of particles. The proposed method circumvents the super-linear scaling exhibited by all known iterative sampling methods applied directly to the RPY tensor that results from the power law growth of the condition number of tensor with the number of particles. For geometrically dense microstructures (fractal dimension equal three), the performance is independent of volume fraction, while for tenuous microstructures (fractal dimension less than three), such as gels and polymer solutions, the performance improves with decreasing volume fraction. This is in stark contrast with other related linear-scaling methods such as the force coupling method and the fluctuating immersed boundary method, for which performance degrades with decreasing volume fraction. Calculations for hard sphere dispersions and colloidal gels are illustrated and used to explore the role of microstructure on performance of the algorithm. In practice, the logarithmic part of the predicted scaling is not observed and the algorithm scales linearly for up to 4×10 6 particles, obtaining speed ups of over an order of magnitude over existing iterative methods, and making the cost of computing Brownian displacements comparable to the cost of computing deterministic displacements in BD simulations. Finally, a high-performance implementation employing non-uniform fast Fourier transforms implemented on graphics processing units and integrated with the software package HOOMD-blue is used for benchmarking.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Low rank approach to computing first and higher order derivatives using automatic differentiation

This manuscript outlines a new approach for increasing the efficiency of applying automatic differentiation (AD) to large scale computational models. By using the principles of the Efficient Subspace Method (ESM), low rank approximations of the derivatives for first and higher orders can be calculated using minimized computational resources. The output obtained from nuclear reactor calculations typically has a much smaller numerical rank compared to the number of inputs and outputs. This rank deficiency can be exploited to reduce the number of derivatives that need to be calculated using AD. The effective rank can be determined according to ESM by computing derivatives with AD at random inputs. Reduced or pseudo variables are then defined and new derivatives are calculated with respect to the pseudo variables. Two different AD packages are used: OpenAD and Rapsodia. OpenAD is used to determine the effective rank and the subspace that contains the derivatives. Rapsodia is then used to calculate derivatives with respect to the pseudo variables for the desired order. The overall approach is applied to two simple problems and to MATWS, a safety code for sodium cooled reactors. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

MODAL FREQUENCY TRACKING: Performance Analysis on Noisy Experimental Ground Test Data

Mechanical devices operating in an environment contaminated by noise, uncertainties, and extraneous disturbances lead to low signal-to-noise-ratios creating an extremely challenging signal processing problem—especially for on-line, realtime, in-flight application. In this report, we discuss a model-based approach to incorporate physical device features into a dynamic structure that can be used to perform on-line tracking of modal frequencies. The approach we take, after pre-processing the raw measurement data, is to obtain a representation of the structural device under test that captures its underlying response to a suite of excitations and operating conditions. One approach is to apply modern system identification techniques based on stochastic subspace methods capable of both: (1) identifying the underlying black-box structure enabling the extraction of structural modes that can be used for analysis; and (2) modal tracking under continued vibrational excitation as in a real-time, in-flight like environment. The vibrational signature of the structural device is measured directly using a suite of tri-axial accelerometers characterizing a multiple input/multiple output (MIMO) dynamic system. In this analysis, we have 15 tri-axial accelerometer measurements or 45 total output channels along with a single excitation also measured with a tri-axial accelerometer providing a 3-input system for processing. We initially describe the experiment and then briefly develop the necessary background in stochastic subspace realization, discuss the algorithm and apply it to the isolated structural device, first, characterized by the set of noisy MIMO accelerometer measurements extracting the modal model from the data (subspace identification), extract the unique modal frequencies (modal coordinate transformation), track their evolution (real-time tracker) to evaluate the overall performance of the processor. Next we evaluate processor performance when the structural device is incorporated into a flight vehicle for further performance analysis. We summarize the results and predict performance.

42 ENGINEERING↗

The tensor t‐function: A definition for functions of third‐order tensors

Summary A definition for functions of multidimensional arrays is presented. The definition is valid for third‐order tensors in the tensor t‐product formalism, which regards third‐order tensors as block circulant matrices. The tensor function definition is shown to have similar properties as standard matrix function definitions in fundamental scenarios. To demonstrate the definition's potential in applications, the notion of network communicability is generalized to third‐order tensors and computed for a small‐scale example via block Krylov subspace methods for matrix functions. A complexity analysis for these methods in the context of tensors is also provided.

Lund, Kathryn↗

A low rank approach to automatic differentiation.

This manuscript introduces a new approach for increasing the efficiency of automatic differentiation (AD) computations for estimating the first order derivatives comprising the Jacobian matrix of a complex large-scale computational model. The objective is to approximate the entire Jacobian matrix with minimized computational and storage resources. This is achieved by finding low rank approximations to a Jacobian matrix via the Efficient Subspace Method (ESM). Low rank Jacobian matrices arise in many of today's important scientific and engineering problems, e.g. nuclear reactor calculations, weather climate modeling, geophysical applications, etc. A low rank approximation replaces the original Jacobian matrix J (whose size is dictated by the size of the input and output data streams) with matrices of much smaller dimensions (determined by the numerical rank of the Jacobian matrix). This process reveals the rank of the Jacobian matrix and can be obtained by ESM via a series of r randomized matrix-vector products of the form: Jq, and J{sup T} {omega} which can be evaluated by the AD forward and reverse modes, respectively.

97 MATHEMATICS AND COMPUTING↗

Multiscale two-stage solver for Biot’s poroelasticity equations in subsurface media

In this paper, we propose a two-stage preconditioner for accelerating the iterative solution by a Krylov subspace method of Biot’s poroelasticity equations based on a displacement-pressure formulation. The spatial discretization combines a finite element method for mechanics and a finite volume approach for flow. The fully implicit backward Euler scheme is used for time integration. The result is a 2 × 2 block linear system for each timestep. The preconditioning operator is obtained by applying a two-stage scheme. The first stage is a global preconditioner that employs multiscale basis functions to construct coarse-scale coupled systems using a Galerkin projection. This global stage is effective at damping low-frequency error modes associated with long-range coupling of the unknowns. The second stage is a local block-triangular smoothing preconditioner, which is aimed at high-frequency error modes associated with short-range coupling of the variables. Lastly, various numerical experiments are used to demonstrate the robustness of the proposed solver.

58 GEOSCIENCES↗

Approximate inverse-based block preconditioners in poroelasticity

We focus on the fully implicit solution of the linear systems arising from a three-field mixed finite element approximation of Biot’s poroleasticity equations. The objective is to develop algebraic block preconditioners for the efficient solution of such systems by Krylov subspace methods. In this work, we investigate the use of approximate inverse-based techniques to decouple the native system of equations and obtain explicit sparse approximations of the Schur complements related to the physics-based partitioning of the unknowns by field type. Here, the proposed methods are tested in various numerical experiments including real-world applications dealing with petroleum and geotechnical engineering.

54 ENVIRONMENTAL SCIENCES↗

A new relaxed splitting preconditioner for the generalized saddle point problems from the incompressible Navier–Stokes equations

Based on the modified relaxed splitting (MRS) preconditioner proposed by Fan and Zhu (Appl Math Lett 55:18–26, 2016), a new relaxed splitting preconditioner is proposed for the generalized saddle point problems arising from the incompressible Navier–Stokes equations. The eigenvalue distribution and an upper bound of the degree of the minimal polynomial of the preconditioned matrix are analyzed. Numerical results show that the proposed preconditioner is superior to the MRS preconditioner when they are used as preconditioners to accelerate the convergence rate of the Krylov subspace methods, such as GMRES.

97 MATHEMATICS AND COMPUTING↗

A low-order block preconditioner for saddle point linear systems

A preconditioner is proposed for the large and sparse linear saddle point problems, which is based on a low-order three-by-three block saddle point form. The eigenvalue distribution and an upper bound of the degree of the minimal polynomial for the preconditioned matrix are discussed. Numerical results show that the optimal convergence behavior can be achieved when the new preconditioner is used to accelerate the convergence rate of Krylov subspace methods such as GMRES.

97 MATHEMATICS AND COMPUTING↗

The parameterized preconditioner for the generalized saddle point problems from the incompressible Navier–Stokes equations

A parameterized preconditioner is proposed for the generalized saddle point problems arising from the incompressible Navier–Stokes equations. The eigenvalues and eigenvectors of the preconditioned matrix are analyzed. Numerical results show that the proposed preconditioner is efficient to accelerate the convergence rate of Krylov subspace methods, such as GMRES method.

97 MATHEMATICS AND COMPUTING↗

Performance of fully-coupled algebraic multigrid preconditioners for large-scale VMS resistive MHD

Here, we explore the current performance and scaling of a fully-implicit stabilized unstructured finite element (FE) variational multiscale (VMS) capability for large-scale simulations of 3D incompressible resistive magnetohydrodynamics (MHD). The large-scale linear systems that are generated by a Newton nonlinear solver approach are iteratively solved by preconditioned Krylov subspace methods. The efficiency of this approach is critically dependent on the scalability and performance of the algebraic multigrid preconditioner. Our study considers the performance of the numerical methods as recently implemented in the second-generation Trilinos implementation that is 64-bit compliant and is not limited by the 32-bit global identifiers of the original Epetra-based Trilinos. The study presents representative results for a Poisson problem on 1.6 million cores of an IBM Blue Gene/Q platform to demonstrate very large-scale parallel execution. Additionally, results for a more challenging steady-state MHD generator and a transient solution of a benchmark MHD turbulence calculation for the full resistive MHD system are also presented. These results are obtained on up to 131,000 cores of a Cray XC40 and one million cores of a BG/Q system.

97 MATHEMATICS AND COMPUTING↗