Search NASA⌕ Search

SEARCH · Search NASA

Results for “algorithmic differentiation”

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 397 records · Page 22

Uncertainty Quantification of Geophysical Inversion Using Stochastic Partial Differential Equations (LDRD #218329)

This report summarizes work completed under the Laboratory Directed Research and Development (LDRD) project "Uncertainty Quantification of Geophysical Inversion Using Stochastic Differential Equations." Geophysical inversions often require computationally expensive algorithms to find even one solution, let alone propagating uncertainties through to the solution domain. The primary purpose of this project was to find more computationally efficient means to approximate solution uncertainty in geophysical inversions. We found multiple computationally efficient methods of propagating Earth model uncertainty into uncertainties in solutions of full waveform seismic moment tensor inversions. However, the optimum method of approximating the uncertainty in these seismic source solutions was to use the Karhunen-Love theorem with data misfit residuals. This method was orders of magnitude more computationally efficient than traditional Monte Carlo methods and yielded estimates of uncertainty that closely approximated those of Monte Carlo. We will summarize the various methods we evaluated for estimating uncertainty in seismic source inversions as well as work toward this goal in the realm of 3-D seismic tomographic inversion uncertainty.

58 GEOSCIENCES↗

Measurement of Piston Deposit Thickness Using Laser Profilometer

In this work, a novel method of measuring the thickness of deposits on engine pistons, including within ring grooves, is described. A laser profilometer is used in conjunction with a rotary stage to measure a continuous profile of the piston with deposits, and the measurement is repeated after cleaning the deposits from the piston. Algorithms for aligning the scans and determining the differential thickness are described, and results are compared to deposit thickness measurements using a contact-type magnetic induction coating thickness gauge. This method allows continuous measurement of deposit thickness across the entire surface including lands and ring grooves as well as quantification of the volume of the deposits.

42 ENGINEERING↗

Progress on the National Solar Radiation Data Base (NSRDB): A new DNI computation

This study introduces a new technique to compute direct normal irradiance (DNI) for improving the National Solar Radiation Data Base (NSRDB). A finite-surface integration algorithm is developed to compute solar radiation in differential solid angles and efficiently infer its contribution to a surface perpendicular to the solar direction. A lookup table of cloud bi-directional transmittance distribution function (BTDF) is developed by use of the discrete ordinates radiative transfer (DISORT) model for possible solar and observing directions and various cloud optical and microphysical properties. In each solar incident direction, DNI is given by the cloud BTDFs from approximately 200 differential solid angles. The simulated DNI is calibrated and evaluated using surface observations at the National Renewable Energy Laboratory's (NREL's) Solar Energy Research Laboratory (SRRL) and the Atmospheric Radiation Measurement (ARM) Southern Great Plains (SGP) facility.

54 ENVIRONMENTAL SCIENCES↗

Improved algorithm for processing grating-based phase contrast interferometry image sets

Grating-based X-ray and neutron interferometry tomography using phase-stepping methods generates large data sets. An improved algorithm is presented for solving for the parameters to calculate transmissions, differential phase contrast, and dark-field images. The method takes advantage of the vectorization inherent in high-level languages such as Mathematica and MATLAB and can solve a 16 × 1k × 1k data set in less than a second. In addition, the algorithm can function with partial data sets. This is demonstrated with processing of a 16-step grating data set with partial use of the original data chosen without any restriction. Also, we have calculated the reduced chi-square for the fit and notice the effect of grating support structural elements upon the differential phase contrast image and have explored expanded basis set representations to mitigate the impact.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

NEW TECHNIQUES FOR HIGH-CONTRAST IMAGING WITH ADI: THE ACORNS-ADI SEEDS DATA REDUCTION PIPELINE

We describe Algorithms for Calibration, Optimized Registration, and Nulling the Star in Angular Differential Imaging (ACORNS-ADI), a new, parallelized software package to reduce high-contrast imaging data, and its application to data from the SEEDS survey. We implement several new algorithms, including a method to register saturated images, a trimmed mean for combining an image sequence that reduces noise by up to {approx}20%, and a robust and computationally fast method to compute the sensitivity of a high-contrast observation everywhere on the field of view without introducing artificial sources. We also include a description of image processing steps to remove electronic artifacts specific to Hawaii2-RG detectors like the one used for SEEDS, and a detailed analysis of the Locally Optimized Combination of Images (LOCI) algorithm commonly used to reduce high-contrast imaging data. ACORNS-ADI is written in python. It is efficient and open-source, and includes several optional features which may improve performance on data from other instruments. ACORNS-ADI requires minimal modification to reduce data from instruments other than HiCIAO. It is freely available for download at www.github.com/t-brandt/acorns-adi under a Berkeley Software Distribution (BSD) license.

79 ASTRONOMY AND ASTROPHYSICS↗

Rapid Cellular Identification by Dynamic Electromechanical Response

Coupling between electrical and mechanical phenomena is ubiquitous in living systems. Here, we demonstrate rapid identification of cellular organisms using difference in electromechanical activity in a broad frequency range. Principal component analysis of the dynamic electromechanical response spectra bundled with neural network based recognition provides a robust identification algorithm based on their electromechanical signature, and allows unambiguous differentiation of model Micrococcus Lysodeikticus and Pseudomonas Fluorescens system. This methodology provides a universal pathway for biological identification obviating the need for well-defined analytical models of Scanning Probe Microscopy response.

60 APPLIED LIFE SCIENCES↗

Randomized algorithms for generalized singular value decomposition with application to sensitivity analysis

The generalized singular value decomposition (GSVD) is a valuable tool that has many applications in computational science. However, computing the GSVD for large-scale problems is challenging. Motivated by applications in hyper-differential sensitivity analysis (HDSA), in this work we propose new randomized algorithms for computing the GSVD which use randomized subspace iteration and weighted QR factorization. Detailed error analysis is given which provides insight into the accuracy of the algorithms and the choice of the algorithmic parameters. We demonstrate the performance of our algorithms on test matrices and a large-scale model problem where HDSA is used to study subsurface flow.

97 MATHEMATICS AND COMPUTING↗

Asymptotic analysis of boundary value and spectral problems in thin perforated regions with rapidly changing thickness and different limiting dimensions

Boundary value and spectral problems for an elliptic differential equation with rapidly oscillating coefficients in a thin perforated region with rapidly changing thickness are investigated. Descriptions of asymptotic algorithms for solutions of such problems in thin regions with different limiting dimensions are combined. For a mixed inhomogeneous boundary value problem a corrector is constructed and an asymptotic estimate in the corresponding Sobolev space is established. Asymptotic bounds for eigenvalues and eigenfunctions of the Neumann spectral problems are also found. Full asymptotic expansions for the eigenvalues and eigenfunctions are constructed under certain symmetry assumptions about the structure of the thin perforated region and the coefficients of the equations. Bibliography: 21 titles.

97 MATHEMATICS AND COMPUTING↗

RSVP - Flu Like Illness and Respiratory Syndromes COVID-19 Syndromic Reporting Tool Prototype

Individuals infected with SARS-CoV-2, the virus that causes COVID-19, may be infectious between 1-3 days prior to symptom onset. People may delay seeking medical care after symptom development due to multiple determinants of health seeking behavior like availability of testing, accessibility of providers, and ability to pay. Therefore, understanding symptoms in the general public is important to better predict and inform resource management plans and engage in reopening. As the influenza season looms, the ability to differentiate between clinical presentation of COVID-19 and seasonal influenza will also be important to health providers and public health response efforts. This project has developed an algorithm that when used with captured syndromic trends can help provide both differentiation to various influenza-like illnesses (ILI) as well as provide public health decision makers a better understanding regarding spatial and temporal trends. This effort has also developed a web-based tool to allow for the capturing of generalized syndromic trends and provide both spatial and temporal outputs on these trends. This page left blank

59 BASIC BIOLOGICAL SCIENCES↗

Terascale Optimal PDE Simulations

The Terascale Optimal PDE Solvers (TOPS) Integrated Software Infrastructure Center (ISIC) was created to develop and implement algorithms and support scientific investigations performed by DOE-sponsored researchers. These simulations often involve the solution of partial differential equations (PDEs) on terascale computers. The TOPS Center researched, developed and deployed an integrated toolkit of open-source, optimal complexity solvers for the nonlinear partial differential equations that arise in many DOE application areas, including fusion, accelerator design, global climate change and reactive chemistry. The algorithms created as part of this project were also designed to reduce current computational bottlenecks by orders of magnitude on terascale computers, enabling scientific simulation on a scale heretofore impossible.

97 MATHEMATICS AND COMPUTING↗

PHURBAS: AN ADAPTIVE, LAGRANGIAN, MESHLESS, MAGNETOHYDRODYNAMICS CODE. II. IMPLEMENTATION AND TESTS

We present an algorithm for simulating the equations of ideal magnetohydrodynamics and other systems of differential equations on an unstructured set of points represented by sample particles. The particles move with the fluid, so the time step is not limited by the Eulerian Courant-Friedrichs-Lewy condition. Full spatial adaptivity is required to ensure the particles fill the computational volume and gives the algorithm substantial flexibility and power. A target resolution is specified for each point in space, with particles being added and deleted as needed to meet this target. We have parallelized the code by adapting the framework provided by GADGET-2. A set of standard test problems, including 10{sup -6} amplitude linear magnetohydrodynamics waves, magnetized shock tubes, and Kelvin-Helmholtz instabilities is presented. Finally, we demonstrate good agreement with analytic predictions of linear growth rates for magnetorotational instability in a cylindrical geometry. This paper documents the Phurbas algorithm as implemented in Phurbas version 1.1.

79 ASTRONOMY AND ASTROPHYSICS↗

Explicit physics-informed neural networks for nonlinear closure: The case of transport in tissues

In upscaling methods, closures for nonlinear problems present a well-known challenge. While a number of theoretical methods have been proposed for handling such closures, nonlinearities still remain a significant obstacle for many problems. In this work, we use a combination of formal upscaling and data-driven machine learning for explicitly closing a nonlinear transport and reaction process in multiscale tissues. The classical effectiveness factor model is used to formulate the macroscale reaction kinetics. We train a multilayer perceptron network using training data generated by direct numerical simulations over microscale examples. Once trained, the network is used in an algorithm for numerically solving the upscaled (coarse-grained) differential equation describing mass transport and reaction in two example tissues. The network is described as being explicit in the sense that the network is trained using macroscale concentrations and gradients of concentration as components of the feature space rather than incorporating them as part of a constraint in the optimization process. Network training and solutions to the macroscale transport equations were computed for two different tissues. The two tissue types (brain and liver) exhibit markedly different geometrical complexity and spatial scale (cell size and sample size). The upscaled solutions for the average concentration are compared with numerical solutions derived from the microscale concentration fields by a posteriori averaging. There are three outcomes of this work of particular note. 1) Our overall approach results in an upscaled nonlinear PDE. The PDE is closed using a neural network, and our approach results in the definition of the classical effectiveness factor for effecting closure. 2) We identify particular source terms for the closure problem that are important for representing the structure of the closure. These source terms involve macroscale concentrations and their gradients. We adopt these source terms to use as explicit features in the learning algorithm. We find the trained networks that include the macroscale source terms generate models that are able to predict the correction factor with increased fidelity over those that do not. 3) We find that the trained network exhibits good generalizability, and it is able to predict the effectiveness factor with high fidelity for realistically-structured tissues despite the significantly different scale and geometrical complexity of the two example tissue types. This latter result emphasizes our purposeful connection between conventional averaging methods with the use of machine learning for closure; this contrasts with some machine learning methods for upscaling where the exact form of the macroscale equation remains unknown.

97 MATHEMATICS AND COMPUTING↗

Modeling the diffusion of heat energy within composites of homogeneous materials using the uncertainty principle

The goal of this paper is to develop a highly accurate and efficient numerical method for the solution of a time-dependent partial differential equation with a piecewise constant coefficient, on a finite interval with periodic boundary conditions. The resulting algorithm can be used, for example, to model the diffusion of heat energy in one space dimension, in the case where the spatial domain represents a medium consisting of two homogeneous materials. The resulting model has, to our knowledge, not yet been solved in closed form through analytical methods, and is difficult to solve using existing numerical methods, thus suggesting an alternative approach. The approach presented in this paper is to represent the solution as a linear combination of wave functions that change frequencies at the interfaces between different materials. It is demonstrated through numerical experiments that using the Uncertainty Principle to construct a basis of such functions, in conjunction with a spectral method, a mathematical model for heat diffusion through different materials can be solved much more efficiently than with conventional time-stepping methods.

97 MATHEMATICS AND COMPUTING↗

SU-G-IeP1-01: A Novel MRI Post-Processing Algorithm for Visualization of the Prostate LDR Brachytherapy Seeds and Calcifications Based On B0 Field Inhomogeneity Correction and Hough Transform

Purpose: This study aims at developing an MRI-only workflow for post-implant dosimetry of the prostate LDR brachytherapy seeds. The specific goal here is to develop a post-processing algorithm to produce positive contrast for the seeds and prostatic calcifications and differentiate between them on MR images. Methods: An agar-based phantom incorporating four dummy seeds (I-125) and five calcifications of different sizes (from sheep cortical bone) was constructed. Seeds were placed arbitrarily in the coronal plane. The phantom was scanned with 3T Philips Achieva MR scanner using an 8-channel head coil array. Multi-echo turbo spin echo (ME-TSE) and multi-echo gradient recalled echo (ME-GRE) sequences were acquired. Due to minimal susceptibility artifacts around seeds, ME-GRE sequence (flip angle=15; TR/TE=20/2.3/2.3; resolution=0.7×0.7×2mm3) was further processed.The induced field inhomogeneity due to the presence of titaniumencapsulated seeds was corrected using a B0 field map. B0 map was calculated using the ME-GRE sequence by calculating the phase difference at two different echo times. Initially, the product of the first echo and B0 map was calculated. The features corresponding to the seeds were then extracted in three steps: 1) the edge pixels were isolated using “Prewitt” operator; 2) the Hough transform was employed to detect ellipses approximately matching the dimensions of the seeds and 3) at the position and orientation of the detected ellipses an ellipse was drawn on the B0-corrected image. Results: The proposed B0-correction process produced positive contrast for the seeds and calcifications. The Hough transform based on Prewitt edge operator successfully identified all the seeds according to their ellipsoidal shape and dimensions in the edge image. Conclusion: The proposed post-processing algorithm successfully visualized the seeds and calcifications with positive contrast and differentiates between them according to their shapes. Further assessments on more realistic phantoms and patient study are required to validate the outcome.

60 APPLIED LIFE SCIENCES↗

Distributed Automatic Differentiation for Ptychography

Synchrotron radiation light source facilities are leading the way to ultrahigh resolution X-ray imaging. High resolution imaging is essential to understanding the fundamental structure and interaction of materials at the smallest length scale possible. Diffraction based methods achieve nanoscale imaging by replacing traditional objective lenses by pixelated area detectors and computational image reconstruction. Among these methods, ptychography is quickly becoming the standard for sub-30 nanometer imaging of extended samples, but at the expense of increasingly high data rates and volumes. This paper presents a new distributed algorithm for solving the ptychographic image reconstruction problem based on automatic differentiation. Input datasets are subdivided between multiple graphics processing units (GPUs); each subset of the problem is then solved either entirely independent of other subsets (asynchronously) or through sharing gradient information with other GPUs (synchronously). The algorithm was evaluated on simulated and real data acquired at the Advanced Photon Source, scaling up to 192 GPUs. The synchronous variant of our method outperformed an existing multi-GPU implementation in terms of accuracy while running at a comparable execution time.

distributed algorithms↗

On Periodic Solutions of Autonomous Systems

We propose a method for the investigation of periodic solutions of autonomous dynamical systems described by ordinary differential equations with phase and integral restrictions. We formulate the general problem of periodic solutions as a boundary-value problem with restrictions. By introducing a fictitious control, we transform the boundary-value problem into a control problem for dynamical systems with phase and integral restrictions. The control problem is solved by reducing it to an integral Fredholm equation of the first kind. We establish necessary and sufficient conditions for the existence of periodic solutions and propose an algorithm for finding periodic solution according to the limit points of the minimizing sequences.

97 MATHEMATICS AND COMPUTING↗

Fast Multigrid Reduction-in-Time for Advection via Modified Semi-Lagrangian Coarse-Grid Operators

Many iterative parallel-in-time algorithms have been shown to be highly efficient for diffusion-dominated partial differential equations (PDEs) but are inefficient or even divergent when applied to advection-dominated PDEs. We consider the application of the multigrid reduction-in-time (MGRIT) algorithm to linear advection PDEs. Here, the key to efficient time integration with this method is using a coarse-grid operator that provides a sufficiently accurate approximation to the so-called ideal coarse-grid operator. For certain classes of semi-Lagrangian discretizations, we present a novel semi-Lagrangian-based coarse-grid operator that leads to fast and scalable multilevel time integration of linear advection PDEs. The coarse-grid operator is composed of a semi-Lagrangian discretization followed by a correction term, with the correction designed so that the leading-order truncation error of the composite operator is approximately equal to that of the ideal coarse-grid operator. Parallel results show substantial speed-ups over sequential time integration for variable-wave-speed advection problems in one and two spatial dimensions, and using high-order discretizations up to order five. The proposed approach establishes the first practical method that provides small and scalable MGRIT iteration counts for advection problems.

97 MATHEMATICS AND COMPUTING↗

Computational singular perturbation analysis of stochastic chemical systems with stiffness

Computational singular perturbation (CSP) is a useful method for analysis, reduction, and time integration of stiff ordinary differential equation systems. It has found dominant utility, in particular, in chemical reaction systems with a large range of time scales at continuum and deterministic level. On the other hand, CSP is not directly applicable to chemical reaction systems at micro or meso-scale, where stochasticity plays an non-negligible role and thus has to be taken into account. In this work we develop a novel stochastic computational singular perturbation (SCSP) analysis and time integration framework, and associated algorithm, that can be used to not only construct accurately and efficiently the numerical solutions to stiff stochastic chemical reaction systems, but also analyze the dynamics of the reduced stochastic reaction systems. Furthermore, the algorithm is illustrated by an application to a benchmark stochastic differential equation model, and numerical experiments are carried out to demonstrate the effectiveness of the construction.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗