Search NASA⌕ Search

SEARCH · Search NASA

Results for “multilevel 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 109 records · Page 6

Upscaling of Mixed Finite Element Discretization Problems by the Spectral AMGe Method

Here, we propose two multilevel spectral techniques for constructing coarse discretization spaces for saddle-point problems corresponding to PDEs involving a divergence constraint, with a focus on mixed finite element discretizations of scalar self-adjoint second order elliptic equations on general unstructured grids. We use element agglomeration algebraic multigrid (AMGe), which employs coarse elements that can have nonstandard shape since they are agglomerates of fine-grid elements. The coarse basis associated with each agglomerated coarse element is constructed by solving local eigenvalue problems and local mixed finite element problems. This construction leads to stable upscaled coarse spaces and guarantees the inf-sup compatibility of the upscaled discretization. Also, the approximation properties of these upscaled spaces improve by adding more local eigenfunctions to the coarse spaces. The higher accuracy comes at the cost of additional computational effort, as the sparsity of the resulting upscaled coarse discretization (referred to as operator complexity) deteriorates when we introduce additional functions in the coarse space. We also provide an efficient solver for the coarse (upscaled) saddle-point system by employing hybridization, which leads to a symmetric positive definite (s.p.d.) reduced system for the Lagrange multipliers, and to solve the latter s.p.d. system, we use our previously developed spectral AMGe solver. Numerical experiments, in both two and three dimensions, are provided to illustrate the efficiency of the proposed upscaling technique.

97 MATHEMATICS AND COMPUTING↗

Optimizing shift selection in multilevel Monte Carlo for disconnected diagrams in lattice QCD

The calculation of disconnected diagram contributions to physical signals is a computationally expensive task in Lattice QCD. To extract the physical signal, the trace of the inverse Lattice Dirac operator, a large sparse matrix, must be stochastically estimated. Because the variance of the stochastic estimator is typically large, variance reduction techniques must be employed. Multilevel Monte Carlo (MLMC) methods reduce the variance of the trace estimator by utilizing a telescoping sequence of estimators. Frequency Splitting is one such method that uses a sequence of inverses of shifted operators to estimate the trace of the inverse lattice Dirac operator, however there is no a priori way to select the shifts that minimize the cost of the multilevel trace estimation. Here we present a sampling and interpolation scheme that is able to predict the variances associated with Frequency Splitting under displacements of the underlying space time lattice. The interpolation scheme is able to predict the variances to high accuracy and therefore chooses shifts that correspond to an approximate minimum of the cost for the trace estimation. We show that Frequency Splitting with the chosen shifts displays significant speedups over multigrid deflation, and that these shifts can be used for multiple configurations within the same ensemble with no penalty to performance.

97 MATHEMATICS AND COMPUTING↗

Interpolation as a means of shift selection for multilevel Monte Carlo with lattice displacements

The calculation of disconnected diagram contributions to physical signals is a computationally expensive task in Lattice QCD. To extract the physical signal, the trace of the inverse Lattice Dirac operator, a large sparse matrix, must be stochastically estimated. Because the variance of the stochastic estimator is typically large, variance reduction techniques must be employed. Multilevel Monte Carlo (MLMC) methods reduce the variance of the trace estimator by utilizing a telescoping sequence of estimators. Frequency Splitting is one such method that uses a sequence of inverses of shifted operators to estimate the trace of the inverse of the lattice Dirac operator, however there is no a priori way to select the shifts that minimize the cost of the multilevel trace estimation. We present a sampling and interpolation scheme that is able to predict the variances associated with Frequency Splitting under displacements of the underlying space time lattice. The interpolation scheme is able to predict the variances to high accuracy and therefore choose shifts that correspond to an approximate minimum of the cost for the trace estimation. We show that Frequency Splitting with the chosen shifts displays significant speedups over multigrid deflation

Whyte, Travis↗

Multilevel segmentation of intracranial aneurysms in CT angiography images

Purpose: Segmentation of aneurysms plays an important role in interventional planning. Yet, the segmentation of both the lumen and the thrombus of an intracranial aneurysm in computed tomography angiography (CTA) remains a challenge. This paper proposes a multilevel segmentation methodology for efficiently segmenting intracranial aneurysms in CTA images. Methods: The proposed methodology first uses the lattice Boltzmann method (LBM) to extract the lumen part directly from the original image. Then, the LBM is applied again on an intermediate image whose lumen part is filled by the mean gray-level value outside the lumen, to yield an image region containing part of the aneurysm boundary. After that, an expanding disk is introduced to estimate the complete contour of the aneurysm. Finally, the contour detected is used as the initial contour of the level set with ellipse to refine the aneurysm. Results: The results obtained on 11 patients from different hospitals showed that the proposed segmentation was comparable with manual segmentation, and that quantitatively, the average segmentation matching factor (SMF) reached 86.99%, demonstrating good segmentation accuracy. Chan–Vese method, Sen’s model, and Luca’s model were used to compare the proposed method and their average SMF values were 39.98%, 40.76%, and 77.11%, respectively. Conclusions: The authors have presented a multilevel segmentation method based on the LBM and level set with ellipse for accurate segmentation of intracranial aneurysms. Compared to three existing methods, for all eleven patients, the proposed method can successfully segment the lumen with the highest SMF values for nine patients and second highest SMF values for the two. It also segments the entire aneurysm with the highest SMF values for ten patients and second highest SMF value for the one. This makes it potential for clinical assessment of the volume and aspect ratio of the intracranial aneurysms.

60 APPLIED LIFE SCIENCES↗

Multilevel Monte Carlo Estimators For Derivative-Free Optimization Under Uncertainty

Optimization is a key tool for scientific and engineering applications; however, in the presence of models affected by uncertainty, the optimization formulation needs to be extended to consider statistics of the quantity of interest. Optimization under uncertainty (OUU) deals with this endeavor and requires uncertainty quantification analyses at several design locations; i.e., its overall computational cost is proportional to the cost of performing a forward uncertainty analysis at each design location. An OUU workflow has two main components: an inner loop strategy for the computation of statistics of the quantity of interest, and an outer loop optimization strategy tasked with finding the optimal design, given a merit function based on the inner loop statistics. Here, in this work, we propose to alleviate the cost of the inner loop uncertainty analysis by leveraging the so-called multilevel Monte Carlo (MLMC) method, which is able to allocate resources over multiple models with varying accuracy and cost. The resource allocation problem in MLMC is formulated by minimizing the computational cost given a target variance for the estimator. We consider MLMC estimators for statistics usually employed in OUU workflows and solve the corresponding allocation problem. For the outer loop, we consider a derivative-free optimization strategy implemented in the SNOWPAC library; our novel strategy is implemented and released in the Dakota software toolkit. We discuss several numerical test cases to showcase the features and performance of our approach with respect to its Monte Carlo single fidelity counterpart.

97 MATHEMATICS AND COMPUTING↗

Hazard analysis for identifying common cause failures of digital safety systems using a redundancy-guided systems-theoretic approach

Replacing the existing aging analog instrumentation and control (I&C) systems with modern safety control and protection, digital technology offers one of the foremost means of performance improvements and cost reductions for the existing nuclear power plants (NPPs). However, the qualification of digital I&C systems remains a challenge, especially considering the issue of software common-cause failures (CCFs), which are difficult to address. With the application and upgrades of advanced digital I&C systems, software CCFs have become a potential threat to plant safety because most redundant designs use similar digital platforms or software in the operating and application systems. With complex designs of multilayer redundancy to meet the single-failure criterion, digital I&C safety systems (e.g., engineered safety-features actuation system [ESFAS]) are of a particular concern in the U.S. Nuclear Regulatory Commission (NRC) licensing procedures. Here, this paper applies a modularized approach to conduct redundancy-guided systems-theoretic hazard analysis for an advanced digital ESFAS with multilevel redundancy designs. Systematic methods and risk-informed tools are incorporated to address both hardware and software CCFs, which provide guidance to eliminate the causal factors of potential single points of failure in the design of digital safety systems in advanced plant designs.

42 ENGINEERING↗

A scalable nonlinear fluid–structure interaction solver based on a Schwarz preconditioner with isogeometric unstructured coarse spaces in 3D

Nonlinear fluid-structure interaction (FSI) problems on unstructured meshes in 3D appear many applications in science and engineering, such as vibration analysis of aircrafts and patient-specific diagnosis of cardiovascular diseases. In this work, we develop a highly scalable, parallel algorithmic and software framework for FSI problems consisting of a nonlinear fluid system and a nonlinear solid system, that are coupled monolithically. The FSI system is discretized by a stabilized finite element method in space and a fully implicit backward difference scheme in time. To solve the large, sparse system of nonlinear algebraic equations at each time step, we propose an inexact Newton-Krylov method together with a multilevel, smoothed Schwarz preconditioner with isogeometric coarse meshes generated by a geometry preserving coarsening algorithm. Here ''geometry'' includes the boundary of the computational domain and the wet interface between the fluid and the solid. We show numerically that the proposed algorithm and implementation are highly scalable in terms of the number of linear and nonlinear iterations and the total compute time on a supercomputer with more than 10,000 processor cores for several problems with hundreds of millions of unknowns.

97 MATHEMATICS AND COMPUTING↗

Linear-Response Time-Dependent Embedded Mean-Field Theory

We present a time-dependent (TD) linear-response description of excited electronic states within the framework of embedded mean-field theory (EMFT). TD-EMFT allows for subsystems to be described at different mean-field levels of theory, enabling straightforward treatment of excited states and transition properties. We provide benchmark demonstrations of TD-EMFT for both local and nonlocal excitations in organic molecules, as well as applications to chlorophyll a, solvatochromic shifts of a dye in solution, and sulfur K-edge X-ray absorption spectroscopy (XAS). It is found that mixed-basis implementations of TD-EMFT lead to substantial errors in terms of transition properties; however, as previously found for ground-state EMFT, these errors are largely eliminated with the use of Fock-matrix corrections. These results indicate that TD-EMFT is a promising method for the efficient, multilevel description of excited-state electronic structure and dynamics in complex systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Multigrid-Reduction-in-Time for the Rotating Shallow Water Equations

We consider multilevel time-parallel methods for the numerical solution of the rotating shallow water equations. In particular, the multigrid-reduction-in-time (MGRIT) algorithm is used for the parallel time integration. An asymptotic model is used at the coarse levels while the full model is employed at the finer levels. The asymptotic model is well-suited for highly oscillatory partial differential equations like the rotating shallow water equations because it can accurately and stably take the required large time-steps on coarse levels. Our work exploits the flexibility of the MGRIT algorithm in terms of the number of levels and relaxation schemes to show some computational benefits, especially with respect to FCF-relaxation and data reuse.

97 MATHEMATICS AND COMPUTING↗

Region-adaptive, Error-controlled Scientific Data Compression using Multilevel Decomposition

The increase of computer processing speed is significantly outpacing improvements in network and storage bandwidth, leading to the big data challenge in modern science, where scientific applications can quickly generate much more data than that can be transferred and stored. As a result, big scientific data must be reduced by a few orders of magnitude while the accuracy of the reduced data needs to be guaranteed for further scientific explorations. Moreover, scientists are often interested in some specific spatial/temporal regions in their data, where higher accuracy is required. The locations of the regions requiring high accuracy can sometimes be prescribed based on application knowledge, while other times they must be estimated based on general spatial/temporal variation. In this paper, we develop a novel multilevel approach which allows users to impose region-wise compression error bounds. Our method utilizes the byproduct of a multilevel compressor to detect regions where details are rich and we provide the theoretical underpinning for region-wise error control. With spatially varying precision preservation, our approach can achieve significantly higher compression ratios than single-error bounded compression approaches and control errors in the regions of interest.We conduct the evaluations on two climate use cases – one targeting small-scale, node features and the other focusing on long, areal features. For both use cases, the locations of the features were unknown ahead of the compression. By selecting approximately 16% of the data based on multi-scale spatial variations and compressing those regions with smaller error tolerances than the rest, our approach improves the accuracy of post-analysis by approximately 2 × compared to single-error-bounded compression at the same compression ratio. Using the same error bound for the region of interest, our approach can achieve an increase of more than 50% in overall compression ratio.

Gong, Qian↗

Use of local noise power spectrum and wavelet analysis in quantitative image quality assurance for EPIDs

Purpose: To investigate the use of local noise power spectrum (NPS) to characterize image noise and wavelet analysis to isolate defective pixels and inter-subpanel flat-fielding artifacts for quantitative quality assurance (QA) of electronic portal imaging devices (EPIDs). Methods: A total of 93 image sets including custom-made bar-pattern images and open exposure images were collected from four iViewGT a-Si EPID systems over three years. Global quantitative metrics such as modulation transform function (MTF), NPS, and detective quantum efficiency (DQE) were computed for each image set. Local NPS was also calculated for individual subpanels by sampling region of interests within each subpanel of the EPID. The 1D NPS, obtained by radially averaging the 2D NPS, was fitted to a power-law function. The r-square value of the linear regression analysis was used as a singular metric to characterize the noise properties of individual subpanels of the EPID. The sensitivity of the local NPS was first compared with the global quantitative metrics using historical image sets. It was then compared with two commonly used commercial QA systems with images collected after applying two different EPID calibration methods (single-level gain and multilevel gain). To detect isolated defective pixels and inter-subpanel flat-fielding artifacts, Haar wavelet transform was applied on the images. Results: Global quantitative metrics including MTF, NPS, and DQE showed little change over the period of data collection. On the contrary, a strong correlation between the local NPS (r-square values) and the variation of the EPID noise condition was observed. The local NPS analysis indicated image quality improvement with the r-square values increased from 0.80 ± 0.03 (before calibration) to 0.85 ± 0.03 (after single-level gain calibration) and to 0.96 ± 0.03 (after multilevel gain calibration), while the commercial QA systems failed to distinguish the image quality improvement between the two calibration methods. With wavelet analysis, defective pixels and inter-subpanel flat-fielding artifacts were clearly identified as spikes after thresholding the inversely transformed images. Conclusions: The proposed local NPS (r-square values) showed superior sensitivity to the noise level variations of individual subpanels compared with global quantitative metrics such as MTF, NPS, and DQE. Wavelet analysis was effective in detecting isolated defective pixels and inter-subpanel flat-fielding artifacts. The proposed methods are promising for the early detection of imaging artifacts of EPIDs.

60 APPLIED LIFE SCIENCES↗

Scaling Structured Multigrid to 500K+ Cores Through Coarse-Grid Redistribution

The effcient solution of sparse, linear systems resulting from the discretization of partial differential equations is crucial to the performance of many physics-based simulations. The algorithmic optimality of multilevel approaches for common discretizations makes them a good candidate for an efficient parallel solver. Yet, modern architectures for high-performance computing systems continue to challenge the parallel scalability of multilevel solvers. While algebraic multigrid methods are robust for solving a variety of problems, the increasing importance of data locality and cost of data movement in modern architectures motivates the need to carefully exploit structure in the problem. Robust logically structured variational multigrid methods, such as Black Box Multigrid (BoxMG), maintain structure throughout the multigrid hierarchy. This avoids indirection and increased coarse- grid communication costs typical in parallel algebraic multigrid. Nevertheless, the parallel scalability of structured multigrid is challenged by coarse-grid problems where the overhead in communication dominates computation. In this paper, an algorithm is introduced for redistributing coarse-grid problems through incremental agglomeration. Guided by a predictive performance model, this algorithm provides robust redistribution decisions for structured multilevel solvers. A two-dimensional diffusion problem is used to demonstrate the significant gain in performance of this algorithm over the previous approach that used agglomeration to one processor. In addition, the parallel scalability of this approach

97 MATHEMATICS AND COMPUTING↗

A Multilevel, Hierarchical Sampling Technique for Spatially Correlated Random Fields

In this paper, we propose an alternative method to generate samples of a spatially correlated random field with applications to large-scale problems for forward propagation of uncertainty. A classical approach for generating these samples is the Karhunen--Loève (KL) decomposition. However, the KL expansion requires solving a dense eigenvalue problem and is therefore computationally infeasible for large-scale problems. Sampling methods based on stochastic partial differential equations provide a highly scalable way to sample Gaussian fields, but the resulting parametrization is mesh dependent. We propose a multilevel decomposition of the stochastic field to allow for scalable, hierarchical sampling based on solving a mixed finite element formulation of a stochastic reaction-diffusion equation with a random, white noise source function. Lastly, numerical experiments are presented to demonstrate the scalability of the sampling method as well as numerical results of multilevel Monte Carlo simulations for a subsurface porous media flow application using the proposed sampling method.

97 MATHEMATICS AND COMPUTING↗

Flexible evaluator for vehicle propulsion systems

A method for automatically generating a multi-component multilevel model for simulation of an automotive propulsion system and/or subsystem. The method comprises a capture component and a build component. The capture component operates to automatically parse all relevant information of an existing model and to generate one or more files indicative of the systems and subsystems of the model. The files are created such that they may be efficiently searched, manipulated, and applied for subsequent use in various user defined models created through the build component. The build component obtains the information in the one or more files and additional user defined files based on user selections to automatically create an executable model ready for analysis.

33 ADVANCED PROPULSION SYSTEMS↗

Understanding the bias introduced in quantum dot blinking using change point analysis

Quantum dots are semiconductor crystals that are less than 100 nanometers in size; that is, between 1 billionth and 100 billionths of a meter. They can be composed of one semiconductor material or a core of one semiconductor surrounded by a shell of another. The quantum dots studied in this work consisted of a cadmium selenide core surrounded by a cadmium sulfide shell. When a quantum dot is illuminated with light, some of it can be absorbed and excites the quantum dot to a higher energy state. To relax back down to its resting state, light can be emitted in the form of fluorescence. The nature of the fluorescence that is emitted from a single quantum dot differs from that which is emitted from a bulk sample (usually a solid film or a solution of quantum dots). While in bulk samples the fluorescence over time fluctuates randomly about some average, maintaining a continuous average intensity, single quantum dot fluorescence switches between two or more intensities. This phenomenon is referred to as quantum dot blinking, and its origin is of significant relevance to solar cell, light emitting diode and biological tracking applications. Quantum dot blinking can be studied by exciting an isolated quantum dot with light and recording the ensuing fluorescence using sensitive detectors. This is done over many minutes, resulting in a fluorescence intensity (in counts per second) versus time trajectory. These trajectories must be further analyzed in order to understand the nature of the blinking behavior. Conventionally, a bin and threshold analysis is used, whereby the arrival times of fluorescent photons are binned to reduce the signal noise, and a threshold is used to separate the trajectory into periods of high and low fluorescence. In this study, the effects of bin size and threshold on the blinking kinetics were investigated, and the assumption that there is only one bright and one dark state was tested. To do so, a statistically rigorous method called change point analysis was needed which does not require binning and which can determine the number of intensity levels present. Blinking in 17 quantum dots was studied using both analysis methods. The primary parameter that was compared between the two methods is called a truncation time. Increasing the bin size was found to generally lead to a larger truncation time while increasing the threshold always decreased the truncation time. Using the change point analysis method, 12 out of 17 quantum dots were found to exhibit more than two intensity levels. Lastly, the assumption that only two intensity levels exist was found to significantly affect the blinking statistics: when individual levels were combined into one bright level, the truncation time would increase. The change point analysis method was found to be more suitable for studying multilevel blinking in quantum dots than the conventional bin and threshold method, since the effects of binning and thresholding can be alleviated and the two-level assumption relaxed.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗