Search NASA⌕ Search

SEARCH · Search NASA

Results for “kernel 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 901 records · Page 50

TADPLOT program, version 2.0: User's guide

The TADPLOT Program, Version 2.0 is described. The TADPLOT program is a software package coordinated by a single, easy-to-use interface, enabling the researcher to access several standard file formats, selectively collect specific subsets of data, and create full-featured publication and viewgraph quality plots. The user-interface was designed to be independent from any file format, yet provide capabilities to accommodate highly specialized data queries. Integrated with an applications software network, data can be assessed, collected, and viewed quickly and easily. Since the commands are data independent, subsequent modifications to the file format will be transparent, while additional file formats can be integrated with minimal impact on the user-interface. The graphical capabilities are independent of the method of data collection; thus, the data specification and subsequent plotting can be modified and upgraded as separate functional components. The graphics kernel selected adheres to the full functional specifications of the CORE standard. Both interface and postprocessing capabilities are fully integrated into TADPLOT.

Hammond, Dana P.↗

Deep-Focusing Time-Distance Helioseismology

Much progress has been made by measuring the travel times of solar acoustic waves from a central surface location to points at equal arc distance away. Depth information is obtained from the range of arc distances examined, with the larger distances revealing the deeper layers. This method we will call surface-focusing, as the common point, or focus, is at the surface. To obtain a clearer picture of the subsurface region, it would, no doubt, be better to focus on points below the surface. Our first attempt to do this used the ray theory to pick surface location pairs that would focus on a particular subsurface point. This is not the ideal procedure, as Born approximation kernels suggest that this focus should have zero sensitivity to sound speed inhomogeneities. However, the sensitivity is concentrated below the surface in a much better way than the old surface-focusing method, and so we expect the deep-focusing method to be more sensitive. A large sunspot group was studied by both methods. Inversions based on both methods will be compared.

Duvall, T. L., Jr.↗

WE-EF-303-04: An Advanced Image Processing Method to Improve the Spatial Resolution of Proton Radiographies

Purpose: We present an optimization method to improve the spatial resolution and the water equivalent thickness accuracy of proton radiographies. Methods: The method is designed for imaging systems measuring only the residual range of protons without relying on tracker detectors to determine the beam trajectory before and after the target. Specifically, the method was used for an imaging set-up consisting of a stack of 61 parallel-plate ionization chambers (PPIC) working as a range telescope. The method uses a decomposition approach of the residual range signal measured by the PPIC and constructs subimages with small size pixels geometrically rearranged and appropriately averaged to be merged into a final single radiography. The method was tested using Monte Carlo simulated and experimental proton radiographies of a PMMA step phantom and an anthropomorphic head phantom. Results: For the step phantom, the effective spatial resolution was found to be 4 and 3 times higher than the nominal resolution for the simulated and experimental radiographies, respectively. For the head phantom, a gamma index was calculated to quantify the conformity of the simulated proton radiographies with a digitally reconstructed X-ray radiography convolved with a Gaussian kernel equal to the proton beam spot-size. For DTA=2.5 mm and RD=2.5%, the passing ratio was 100%/85% for the optimized/non-optimized case, respectively. An extension of the method allows reducing the dose given to the patient during radiography acquisition. We show that despite a dose reduction of 25 times (leading to a dose of 0.016 mGy for the current imaging set-up), the image quality of the optimized radiographies remains fairly unaffected for both the simulated and experimental results. Conclusion: The optimization method leads to a significant increase of the spatial resolution allowing recovering image details that are unresolved in non-optimized radiographies. These results represent a major step towards clinical application of proton radiography.

60 APPLIED LIFE SCIENCES↗

A convolution-superposition dose calculation engine for GPUs

Purpose: Graphic processing units (GPUs) are increasingly used for scientific applications, where their parallel architecture and unprecedented computing power density can be exploited to accelerate calculations. In this paper, a new GPU implementation of a convolution/superposition (CS) algorithm is presented. Methods: This new GPU implementation has been designed from the ground-up to use the graphics card's strengths and to avoid its weaknesses. The CS GPU algorithm takes into account beam hardening, off-axis softening, kernel tilting, and relies heavily on raytracing through patient imaging data. Implementation details are reported as well as a multi-GPU solution. Results: An overall single-GPU acceleration factor of 908x was achieved when compared to a nonoptimized version of the CS algorithm implemented in PlanUNC in single threaded central processing unit (CPU) mode, resulting in approximatively 2.8 s per beam for a 3D dose computation on a 0.4 cm grid. A comparison to an established commercial system leads to an acceleration factor of approximately 29x or 0.58 versus 16.6 s per beam in single threaded mode. An acceleration factor of 46x has been obtained for the total energy released per mass (TERMA) calculation and a 943x acceleration factor for the CS calculation compared to PlanUNC. Dose distributions also have been obtained for a simple water-lung phantom to verify that the implementation gives accurate results. Conclusions: These results suggest that GPUs are an attractive solution for radiation therapy applications and that careful design, taking the GPU architecture into account, is critical in obtaining significant acceleration factors. These results potentially can have a significant impact on complex dose delivery techniques requiring intensive dose calculations such as intensity-modulated radiation therapy (IMRT) and arc therapy. They also are relevant for adaptive radiation therapy where dose results must be obtained rapidly.

61 RADIATION PROTECTION AND DOSIMETRY↗

Carbothermic Synthesis of 820 m UN Kernels: Literature Review, Thermodynamics, Analysis, and Related Experiments

The U.S. Department of Energy is considering a new nuclear fuel that would be less susceptible to ruptures during a loss-of-coolant accident. The fuel would consist of tristructural isotropic coated particles with large, dense uranium nitride (UN) kernels. This effort explores many factors involved in using gel-derived uranium oxide-carbon microspheres to make large UN kernels. Analysis of recent studies with sufficient experimental details is provided. Extensive thermodynamic calculations are used to predict carbon monoxide and other pressures for several different reactions that may be involved in conversion of uranium oxides and carbides to UN. Experimentally, the method for making the gel-derived microspheres is described. These were used in a microbalance with an attached mass spectrometer to determine details of carbothermic conversion in argon, nitrogen, or vacuum. A quantitative model is derived from experiments for vacuum conversion to an uranium oxide-carbide kernel.

carothermic↗

Python for Development of OpenMP and CUDA Kernels for Multidimensional Data

Design of data structures for high performance computing (HPC) is one of the principal challenges facing researchers looking to utilize heterogeneous computing machinery. Heterogeneous systems derive cost, power, and speed efficiency by being composed of the appropriate hardware for the task. Yet, each type of processor requires a specific organization of the application state in order to achieve peak performance. Discovering this and refactoring the code can be a challenging and time-consuming task for the researcher, as the data structures and the computational model must be co-designed. We present a methodology that uses Python as the environment for which to explore tradeoffs in both the data structure design as well as the code executing on the computation accelerator. Our method enables multi-dimensional arrays to be used effectively in any target environment. We have chosen to focus on OpenMP and CUDA environments, thus exploring the development of optimized kernels for the two most common classes of computing hardware available today: multi-core CPU and GPU. Python s large palette of file and network access routines, its associative indexing syntax and support for common HPC environments makes it relevant for diverse hardware ranging from laptops through computing clusters to the highest performance supercomputers. Our work enables researchers to accelerate the development of their codes on the computing hardware of their choice.

43 PARTICLE ACCELERATORS↗

Monte Carlo mesh tallies based on a Kernel Density Estimator approach using integrated particle tracks

A new Monte Carlo mesh tally based on a Kernel Density Estimator (KDE) approach using integrated particle tracks is presented. We first derive the KDE integral-track estimator and present a brief overview of its implementation as an alternative to the MCNP fmesh tally. To facilitate a valid quantitative comparison between these two tallies for verification purposes, there are two key issues that must be addressed. The first of these issues involves selecting a good data transfer method to convert the nodal-based KDE results into their cell-averaged equivalents (or vice versa with the cell-averaged MCNP results). The second involves choosing an appropriate resolution of the mesh, since if it is too coarse this can introduce significant errors into the reference MCNP solution. After discussing both of these issues in some detail, we present the results of a convergence analysis that shows the KDE integral-track and MCNP fmesh tallies are indeed capable of producing equivalent results for some simple 3D transport problems. In all cases considered, there was clear convergence from the KDE results to the reference MCNP results as the number of particle histories was increased. (authors)

97 MATHEMATICS AND COMPUTING↗

Improved dosimetry for targeted radionuclide therapy using nonrigid registration on sequential SPECT images

Purpose: Voxel-level and patient-specific 3D dosimetry for targeted radionuclide therapy (TRT) typically involves serial nuclear medicine scans. Misalignment of the images can result in reduced dosimetric accuracy. Since the scans are typically performed over a period of several days, there will be patient movement between scans and possible nonrigid organ deformation. This work aims to implement and evaluate the use of nonrigid image registration on a series of quantitative SPECT (QSPECT) images for TRT dosimetry. Methods: A population of 4D extended cardiac torso phantoms, comprised of three In-111 Zevalin biokinetics models and three anatomical variations, was generated based on the patient data. The authors simulated QSPECT acquisitions at five time points. At each time point, individual organ and whole-body deformation between scans were modeled by translating/rotating organs and the body up to 5°/voxels, keeping ≤5% difference in organ volume. An analytical projector was used to generate realistic noisy projections for a medium energy general purpose collimator. Projections were reconstructed using OS-EM algorithm with geometric collimator detector response, attenuation, and scatter corrections. The QSPECT images were registered using organ-based nonrigid image registration method. The cumulative activity in each voxel was obtained by integrating the activity over time. Dose distribution images were obtained by convolving the cumulative activity images with a Y-90 dose kernel. Dose volume histograms (DVHs) for organs-of-interest were analyzed. Results: After nonrigid registration, the mean differences in organ doses compared to the case without misalignment were improved from (−15.50 ± 5.59)% to (−2.12 ± 1.05)% and (−7.28 ± 2.30)% to (−0.23 ± 0.71)% for the spleen and liver, respectively. For all organs, the cumulative DVHs showed improvement after nonrigid registration and the normalized absolute error of differential DVHs ranged from 6.79% to 22.70% for liver and 26.00% to 39.70% for spleen with different segmentation methods. Conclusions: These results demonstrated that nonrigid registration of sequential QSPECT images is feasible for TRT and improves the accuracy of 3D dosimetry.

60 APPLIED LIFE SCIENCES↗

Efficient 3-D velocity model building using joint inline and crossline plane-wave wave-equation migration velocity analyses

SUMMARY Wave-equation migration velocity analysis (WEMVA) is an image-domain inversion method for velocity model building. Automatic plane-wave WEMVA (PWEMVA) calculates the moveouts of plane-wave common-image gathers (CIGs) by searching a best-fitting parabola with semblance analysis and backprojects residual CIG moveouts into wavefield wave paths with a reflection tomographic kernel. However, 3-D PWEMVA is very computationally expensive because 3-D reflection tomographic inversion requires at least five 3-D reverse-time migrations per iteration and stores two types of source wavefields at model boundaries. We develop a joint inline and crossline PWEMVA method for efficient 3-D velocity model building. We alternatively implement the inline and crossline PWEMVAs with a constraint for each other, in which we iteratively construct the 3-D velocity model update through 1-D spline interpolation of 2-D gradients. The inline and crossline joint inversion is practical since PWEMVA only inverts for low-wavenumber velocity perturbations along wave paths, and the method can take less than 1 per cent of the computational cost of full 3-D PWEMVA. To construct unaliased plane waves for our joint inline and crossline PWEMVA, we develop a 3-D data interpolation method in the frequency–wavenumber (FK) domain to recover regularly and randomly missing traces. The method minimizes the misfit on sufficiently localized data subsets with iterative optimal step lengths and a gradient preconditioner that iteratively selects dominant dips along different azimuths. In numerical experiments, we use a 3-D synthetic seismic data set and a land 3-D field seismic data set acquired at the Farnsworth CO2-EOR (enhanced oil recovery) field to demonstrate the efficacy of our velocity model building and data interpolation methods.

Liu, Xuejian↗

Data-Efficient Strategies for Probabilistic Voltage Envelopes under Network Contingencies

This work presents an efficient data-driven method to construct probabilistic voltage envelopes (PVE) using power flow learning in grids with network contingencies. First, a network-aware Gaussian process (GP) termed Vertex-Degree Kernel (VDK-GP), developed in prior work, is used to estimate voltage–power functions for a few network configurations. The paper introduces a novel multi-task vertex degree kernel (MT-VDK) that amalgamates the learned VDK-GPs to determine power flows for unseen networks, with a significant reduction in the computational complexity and hyperparameter requirements compared to alternate approaches. Simulations on the IEEE 30-Bus network demonstrate the retention and transfer of power flow knowledge in both N-1 and N-2 contingency scenarios. The MT-VDK-GP approach achieves over 50 % reduction in mean prediction error for novel N-1 contingency network configurations in low training data regimes (50–250 samples) over VDK-GP. Additionally, MT-VDK-GP outperforms a hyper-parameter based transfer learning approach in over 75 % of N-2 contingency network structures, even without historical N-2 outage data. Furthermore, the proposed method demonstrates the ability to achieve PVEs using sixteen times fewer power flow solutions compared to Monte-Carlo sampling-based methods.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Simulating Hydrodynamics in Cosmology with CRK-HACC

Abstract We introduce CRK-HACC, an extension of the Hardware/Hybrid Accelerated Cosmology Code (HACC), to resolve gas hydrodynamics in large-scale structure formation simulations of the universe. The new framework couples the HACC gravitational N -body solver with a modern smoothed-particle hydrodynamics (SPH) approach called conservative reproducing kernel SPH (CRKSPH). CRKSPH utilizes smoothing functions that exactly interpolate linear fields while manifestly preserving conservation laws (momentum, mass, and energy). The CRKSPH method has been incorporated to accurately model baryonic effects in cosmology simulations—an important addition targeting the generation of precise synthetic sky predictions for upcoming observational surveys. CRK-HACC inherits the codesign strategies of the HACC solver and is built to run on modern GPU-accelerated supercomputers. In this work, we summarize the primary solver components and present a number of standard validation tests to demonstrate code accuracy, including idealized hydrodynamic and cosmological setups, as well as self-similarity measurements.

79 ASTRONOMY AND ASTROPHYSICS↗

Semi-classical Kinetic Theory for Massive Spin-half Fermions with Leading-order Spin Effects

We consider the quantum kinetic-theory description for interacting massive spin-half fermions using the Wigner function formalism. We derive a general kinetic theory description assuming that the spin effects appear at the classical and quantum level. To track the effect of such different contributions we use the semi-classical expansion method to obtain the generalized dynamical equations including spin, analogous to classical Boltzmann equation. This approach can be used to obtain a collision kernel involving local as well as non-local collisions among the microscopic constituent of the system and eventually, a framework of spin hydrodynamics ensuring the conservation of the energy-momentum tensor and total angular momentum tensor.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Recovering pointwise values of discontinuous data within spectral accuracy

The pointwise values of a function, f(x), can be accurately recovered either from its spectral or pseudospectral approximations, so that the accuracy solely depends on the local smoothness of f in the neighborhood of the point x. Most notably, given the equidistant function grid values, its intermediate point values are recovered within spectral accuracy, despite the possible presence of discontinuities scattered in the domain. (Recall that the usual spectral convergence rate decelerates otherwise to first order, throughout). To this end, a highly oscillatory smoothing kernel is employed in contrast to the more standard positive unit-mass mollifiers. In particular, post-processing of a stable Fourier method applied to hyperbolic equations with discontinuous data, recovers the exact solution modulo a spectrally small error. Numerical examples are presented.

Gottlieb, D.↗

Recovering pointwise values of discontinuous data within spectral accuracy

The pointwise values of a function, f(x), can be accurately recovered either from its spectral or pseudospectral approximations, so that the accuracy solely depends on the local smoothness of f in the neighborhood of the point x. Most notably, given the equidistant function grid values, its intermediate point values are recovered within spectral accuracy, despite the possible presence of discontinuities scattered in the domain. (Recall that the usual spectral convergence rate decelerates otherwise to first order, throughout). To this end, a highly oscillatory smoothing kernel is employed in contrast to the more standard positive unit-mass mollifiers. In particular, post-processing of a stable Fourier method applied to hyperbolic equations with discontinuous data, recovers the exact solution modulo a spectrally small error. Numerical examples are presented.

Gottlieb, D.↗

Sensitivity analysis of a wing aeroelastic response

A variation of Sobieski's Global Sensitivity Equations (GSE) approach is implemented to obtain the sensitivity of the static aeroelastic response of a three-dimensional wing model. The formulation is quite general and accepts any aerodynamics and structural analysis capability. An interface code is written to convert one analysis's output to the other's input, and visa versa. Local sensitivity derivatives are calculated by either analytic methods or finite difference techniques. A program to combine the local sensitivities, such as the sensitivity of the stiffness matrix or the aerodynamic kernel matrix, into global sensitivity derivatives is developed. The aerodynamic analysis package FAST, using a lifting surface theory, and a structural package, ELAPS, implementing Giles' equivalent plate model are used.

Kapania, Rakesh K.↗

A Bayesian Learning Approach to Wireless Outdoor Heatmap Construction using Deep Gaussian Process

We present a novel Bayesian learning approach to outdoor radio heatmap construction utilizing deep Gaussian process (GP). The proposed approach employs a two-layer hierarchy which consists of two cascaded Gaussian processes that are capable of modeling more complex input-output relations than standard single-layer Gaussian processes. Since deriving the exact model likelihood is challenging, a lower bound is optimized instead so that gradient descent-based methods can be performed to find out the optimal model parameters. Typically, inducing points are used in GPs to facilitate low-rank approximation of covariance (kernel) matrices for computation speedup. However, the inaccuracy induced by inducing points can accumulate when stacking multiple layers of GP which may hinder the performance of deep GP. Moreover, since inducing points need to be learned, having them at all layers of deep GP also incurs computational burden. To overcome the above challenges, in contrast to the canonical deep GP model, we use a modified architecture where a full standard GP resides in the first layer and inducing points are only introduced for the second layer. This modified architecture strikes a balance between model accuracy and training complexity. In the proposed model, the noise parameter of the first GP layer is also eliminated to improve the training efficiency as the noise parameter at the output of the second layer suffices to model the uncertainty in the output. The proposed approach is evaluated on real-world datasets, in the form of location-Received Signal Strength (RSS) pairs, collected from the Platform for Open Wireless Data-driven Experimental Research (POWDER) located at the campus of the University of Utah. Experiment results show that the proposed approach can achieve smaller prediction errors on various training and testing data configurations than DNN-based and GP-based methods.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Toward a better thermal scattering law of (C 5 O 2 H 8 ) n : Inelastic neutron scattering and oClimax + NJOY2016

With the advancements in technology (both experimental and computational) the determination of the“true” experimental phonon spectrum became more accessible. Here a methodology for producing thermal scattering libraries from the experimental data (namely the DFT+oClimax method) for lucite (C 5 O 2 H 8 ) n is discussed. Double differential scattering cross section (DDSCS) experiments were performed at the Spallation Neutron Source of Oak Ridge National Laboratory (SNS ORNL). Novel scattering kernel evaluations, based on the phonon spectrum for (C 5 O 2 H 8 ) n ,were created using oClimax and NJOY2016 codes. In order to compare and asses the performance of the newly created library, the experimental setup was simulated using MCNP6.1. Compared to the current ENDF/B-VIII.0, the resulting RPI (C 5 O 2 H 8 ) n library improved the calculation of both double differential scattering and total scattering cross sections. A set of criticality benchmarks containing (C 5 O 2 H 8 ) n from HEU-MET-THERM resulted in an overall improved calculation of K eƒƒ . The DFT + oClimax method is shown to be the most comprehensive method for analysis of moderator materials, due to the fact that it can be verified against all data measured at VISION, ARCS and SEQUOIA neutron spectrometers at SNS ORNL, and experimental total scattering cross section measurements. This technique also provides a new technique for calculating any phonon spectrum-related quantities such as scattering law kernel, specific heat capacity, thermal conductivity, etc. for any solid state material.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Magnetic Resonance Imaging-Based Radiation-Absorbed Dose Estimation of {sup 166}Ho Microspheres in Liver Radioembolization

Purpose: To investigate the potential of magnetic resonance imaging (MRI) for accurate assessment of the three-dimensional {sup 166}Ho activity distribution to estimate radiation-absorbed dose distributions in {sup 166}Ho-loaded poly (L-lactic acid) microsphere ({sup 166}Ho-PLLA-MS) liver radioembolization. Methods and Materials: MRI, computed tomography (CT), and single photon emission CT (SPECT) experiments were conducted on an anthropomorphic gel phantom with tumor-simulating gel samples and on an excised human tumor-bearing liver, both containing known amounts of {sup 166}Ho-PLLA-MS. Three-dimensional radiation-absorbed dose distributions were estimated at the voxel level by convolving the {sup 166}Ho activity distribution, derived from quantitative MRI data, with a {sup 166}Ho dose point-kernel generated by MCNP (Monte Carlo N-Particle transport code) and from Medical Internal Radiation Dose Pamphlet 17. MRI-based radiation-absorbed dose distributions were qualitatively compared with CT and autoradiography images and quantitatively compared with SPECT-based dose distributions. Both MRI- and SPECT-based activity estimations were validated against dose calibrator measurements. Results: Evaluation on an anthropomorphic phantom showed that MRI enables accurate assessment of local {sup 166}Ho-PLLA-MS mass and activity distributions, as supported by a regression coefficient of 1.05 and a correlation coefficient of 0.99, relating local MRI-based mass and activity calculations to reference values obtained with a dose calibrator. Estimated MRI-based radiation-absorbed dose distributions of {sup 166}Ho-PLLA-MS in an ex vivo human liver visually showed high correspondence to SPECT-based radiation-absorbed dose distributions. Quantitative analysis revealed that the differences in local and total amounts of {sup 166}Ho-PLLA-MS estimated by MRI, SPECT, and the dose calibrator were within 10%. Excellent agreement was observed between MRI- and SPECT-based dose-volume histograms. Conclusions: Quantitative MRI was demonstrated to provide accurate three-dimensional {sup 166}Ho-PLLA-MS activity distributions, enabling localized intrahepatic radiation-absorbed dose estimation by convolution with a {sup 166}Ho dose point-kernel for liver radioembolization treatment optimization and evaluation.

62 RADIOLOGY AND NUCLEAR MEDICINE↗