Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “error localization”

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

Fermi–Löwdin orbital self-interaction correction of adsorption energies on transition metal ions

Density functional theory (DFT)-based descriptions of the adsorption of small molecules on transition metal ions are prone to self-interaction errors. Here, we show that such errors lead to a large over-estimation of adsorption energies of small molecules on Cu + , Zn + , Zn 2+ , and Mn + in local spin density approximation (LSDA) and Perdew, Burke, Ernzerhof (PBE) generalized gradient approximation calculations compared to reference values computed using the coupled-cluster with single, doubles, and perturbative triple excitations method. These errors are significantly reduced by removing self-interaction using the Perdew–Zunger self-interaction correction (PZ-SIC) in the Fermi–Löwdin Orbital (FLO) SIC framework. In the case of FLO-PBE, typical errors are reduced to less than 0.1 eV. Furthermore, analysis of the results using DFT energies evaluated on self-interaction-corrected densities [DFT(@FLO)] indicates that the density-driven contributions to the FLO-DFT adsorption energy corrections are roughly the same size in DFT = LSDA and PBE, but the total corrections due to removing self-interaction are larger in LSDA.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

PROTEUS: Machine Learning Driven Resilience for Extreme-scale Systems

The objective of this project is to design, develop, and evaluate scalable software to enhance resilience, data checkpointing, program restart, and analysis. The proposed tasks are to 1) develop scalable machine learning techniques to learn temporal change patterns in a scalable and in-situ manner, and to minimize data movement and maximize learning locally closest to data; 2) design a concise data representation and indexing mechanism to capture the distribution of changes in data that can guarantee point-wise user-defined tolerable errors while reducing the data storage requirements by an order of magnitude or more; 3) develop data reduction techniques as library modules; 4) exploit local SSD for minimizing data movement in storage hierarchy; 5) develop anomaly detection algorithms that can predict corruptions based on learning of emerging patterns; 6) develop software libraries to be incorporated within widely used data formats and APIs; and 7) evaluate the proposed software using DOE scientific applications. The outcomes of the proposed work are to satisfy many synergistic data reduction and resilience requirements for large-scale data intensive applications executed on extreme-scale computing systems. The developed mechanism for error-bound data approximation is directly applicable to existing scientific applications. Through machine learning from historical events and change distribution, this work will enable anomaly detection for DOE computer facility.

97 MATHEMATICS AND COMPUTING↗

Dynamic Stochastic Superresolution of sparsely observed turbulent systems

Real-time capture of the relevant features of the unresolved turbulent dynamics of complex natural systems from sparse noisy observations and imperfect models is a notoriously difficult problem. The resulting lack of observational resolution and statistical accuracy in estimating the important turbulent processes, which intermittently send significant energy to the large-scale fluctuations, hinders efficient parameterization and real-time prediction using discretized PDE models. This issue is particularly subtle and important when dealing with turbulent geophysical systems with an vast range of interacting spatio-temporal scales and rough energy spectra near the mesh scale of numerical models. Here, we introduce and study a suite of general Dynamic Stochastic Superresolution (DSS) algorithms and show that, by appropriately filtering sparse regular observations with the help of cheap stochastic exactly solvable models, one can derive stochastically ‘superresolved’ velocity fields and gain insight into the important characteristics of the unresolved dynamics, including the detection of the so-called black swans. The DSS algorithms operate in Fourier domain and exploit the fact that the coarse observation network aliases high-wavenumber information into the resolved waveband. It is shown that these cheap algorithms are robust and have significant skill on a test bed of turbulent solutions from realistic nonlinear turbulent spatially extended systems in the presence of a significant model error. In particular, the DSS algorithms are capable of successfully capturing time-localized extreme events in the unresolved modes, and they provide good and robust skill for recovery of the unresolved processes in terms of pattern correlation. Moreover, we show that DSS improves the skill for recovering the primary modes associated with the sparse observation mesh which is equally important in applications. The skill of the various DSS algorithms depends on the energy spectrum of the turbulent signal and the observation time relative to the decorrelation time of the turbulence at a given spatial scale in a fashion elucidated here. The DSS technique exploiting a simple Gaussian closure of the nonlinear stochastic forecast model emerges as the most suitable trade-off between the superresolution skill and computational complexity associated with estimating the cross-correlations between the aliasing modes of the sparsely observed turbulent signal. Such techniques offer a promising and efficient approach to constraining unresolved turbulent fluxes through stochastic superparameterization and a subsequent improvement in coarse-grained filtering and prediction of the next generation atmosphere–ocean system (AOS) models.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Quantitative and three-dimensional assessment of holdup material

Nuclear material deposited in equipment, transfer lines, and ventilation systems of a processing facility is usually referred to as holdup. Operators need to know the location and amount of holdup in their facility for radiation safety and material accountability. In this work, we propose to use an array of detectors co-axial to the inspected pipe to measure the holdup material. This method is implementable into an automated system capable of crawling on surfaces and pipes of various curvatures, which would enable faster, easier, and more accurate holdup safeguards measurements. We first demonstrated that the current holdup assay procedure could lead to a non-negligible bias in the estimate of special nuclear material mass, due to the simplified assumption of deposited geometry introduced by the Generalized Geometry Holdup (GGH) model. The new approach consists of imaging the inner holdup material by characterizing the detector array’s response and unfolding it from the measured light output. Our experimental proof of principle consists of three NaI(Tl) detectors surrounding an aluminum pipe containing two cesium-137 ( 137 Cs) sources. We derived the source distribution inside the pipe by first calculating the detector response matrix using a method adaptive to the surface geometry of the object containing the measured holdup material. Creating a matrix of the detector array’s measured counts, we then proceed to solve an inverse problem, resulting in an accurately located source position and activity distribution within the response matrix’s spatial resolution. Furthermore, we then developed a simulated model of the envisioned experimental setup, which accurately described both the activity and position of the source in 2D. Finally, we extended our model onto a discretized three-dimensional model of the system, encompassing 36 detectors. The model was simulated to measure the source distribution and activity of multiple sources in pipes of varied geometries, which could then be conveniently translated to a real-time holdup measurement protocol. For the 3D simulation of four different source geometries, the model accurately localized the source position in 3D, while the activity retained a maximum relative error of ±5.32%.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Density Matrix Implementation of the Fermi–Löwdin Orbital Self-Interaction Correction Method

The Fermi–Löwdin orbital self-interaction correction (FLOSIC) method effectively provides a transformation from canonical orbitals to localized Fermi–Löwdin orbitals which are used to remove the self-interaction error in the Perdew–Zunger (PZ) framework. This transformation is solely determined by a set of points in space, called Fermi–Löwdin descriptors (FODs), and the occupied canonical orbitals or the density matrix. In this work, we provide a detailed workflow for the implementation of the FLOSIC method for removal of self-interaction error in DFT calculations in an orbital-by-orbital basis that takes advantage of the unitary invariant nature of the FLOSIC method. In this way, it is possible to cast the self-consistent energy minimization at fixed FODs in the same manner than standard Kohn–Sham with one additional term in the Kohn–Sham Hamiltonian that introduces the PZ self-interaction correction. Each energy minimization iteration is divided in two substeps, one for the density matrix and one for the FODs. Expressions for the effective Kohn–Sham matrix and FOD gradients are provided such that its implementation is suitable for most electronic structure codes. Here, we analyze the convergence characteristics of the algorithm and present applications for the evaluation of NMR shielding constants and real-time time-dependent DFT simulations based on the Liouville–von Neumann equation to calculate excitation energies.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Computing x-ray absorption spectra from linear-response particles atop optimized holes

State specific orbital optimized density functional theory (OO-DFT) methods, such as restricted open-shell Kohn–Sham (ROKS), can attain semiquantitative accuracy for predicting x-ray absorption spectra of closed-shell molecules. OO-DFT methods, however, require that each state be individually optimized. In this Communication, we present an approach to generate an approximate core-excited state density for use with the ROKS energy ansatz, which is capable of giving reasonable accuracy without requiring state-specific optimization. Herein, this is achieved by fully optimizing the core-hole through the core-ionized state, followed by the use of electron-addition configuration interaction singles to obtain the particle level. This hybrid approach can be viewed as a DFT generalization of the static-exchange (STEX) method and can attain ~0.6 eV rms error for the K-edges of C–F through the use of local functionals, such as PBE and OLYP. This ROKS(STEX) approach can also be used to identify important transitions for full OO ROKS treatment and can thus help reduce the computational cost of obtaining OO-DFT quality spectra. ROKS(STEX), therefore, appears to be a useful technique for the efficient prediction of x-ray absorption spectra.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Bond length alternation of π -conjugated polymers predicted by the Fermi–Löwdin orbital self-interaction correction method

π-conjugated polymers have been used in a wide range of practical applications, partly due to their unique properties that originate in the delocalization of electrons through the polymer backbone. The level of delocalization can be characterized by the induced bond length alternation (BLA), with shorter BLA connected with strong delocalization and vice versa. The accurate description of this structural parameter can be considered a benchmark for testing the capability of different electronic structure methods for self-interaction error (SIE) removal and electron correlation inclusion. Density functional theory (DFT), in its local or semi-local flavors, suffers from SIE and, thus, underestimates the BLA compared to self-interaction-free methods. In this work, we utilize the Fermi–Löwdin orbital self-interaction correction (FLOSIC) method for one-electron self-interaction removal to characterize the BLA of five oligomers with increasing length extrapolated to the polymeric limit. We compare the self-interaction-free BLA to several DFT approximations, Møller–Plesset second-order perturbation theory (MP2), and the BLA obtained with the domain based local pair natural orbital CCSD(T) [DLPNO-CCSD(T)] approximation. Our findings show that FLOSIC corrects for the small BLA given by (semi-)local DFT approximations, but it tends to overcorrect with respect to CAM-B3LYP, MP2, and DLPNO-CCSD(T).

Chemistry↗

Lifting primordial non-Gaussianity above the noise

Primordial non-Gaussianity (PNG) in Large Scale Structures is obfuscated by the many additional sources of non-linearity. Within the Effective Field Theory approach to Standard Perturbation Theory, we show that matter non-linearities in the bispectrum can be modeled sufficiently well to strengthen current bounds with near future surveys, such as Euclid. We find that the EFT corrections are crucial to this improvement in sensitivity. Yet, our understanding of non-linearities is still insufficient to reach important theoretical benchmarks for equilateral PNG, while, for local PNG, our forecast is more optimistic. We consistently account for the theoretical error intrinsic to the perturbative approach and discuss the details of its implementation in Fisher forecasts.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Procrustean entanglement concentration in quantum-classical networking

The success of a future quantum internet will rest in part on the ability of quantum and classical signals to coexist in the same optical fiber infrastructure, a challenging endeavor given the orders of magnitude differences in flux of single-photon-level quantum fields and bright classical traffic. Here, we theoretically describe and experimentally implement Procrustean entanglement concentration for polarization-entangled states contaminated with classical light, showing significant mitigation of crosstalk noise in dense wavelength-division multiplexing. Our approach leverages a pair of polarization-dependent loss emulators to attenuate highly polarized crosstalk that results from imperfect isolation of conventional signals copropagating on shared fiber links. We demonstrate our technique both on the tabletop and over a deployed quantum local area network, finding a substantial improvement of two-qubit entangled state fidelity from approximately 75% to over 92%. This local filtering technique could be used as a preliminary step to reduce asymmetric errors, potentially improving the overall efficiency when combined with more complex error-mitigation techniques in future quantum repeater networks.

97 MATHEMATICS AND COMPUTING↗

SU-E-J-34: Setup Accuracy in Spine SBRT Using CBCT 6D Image Guidance in Comparison with 6D ExacTrac

Purpose Volumetric information of the spine captured on CBCT can potentially improve the accuracy in spine SBRT setup that has been commonly performed through 2D radiographs. This work evaluates the setup accuracy in spine SBRT using 6D CBCT image guidance that recently became available on Varian systems. Methods ExacTrac radiographs have been commonly used for Spine SBRT setup. The setup process involves first positioning patients with lasers followed by localization imaging, registration, and repositioning. Verification images are then taken providing the residual errors (ExacTracRE) before beam on. CBCT verification is also acquired in our institute. The availability of both ExacTrac and CBCT verifications allows a comparison study. 41 verification CBCT of 16 patients were retrospectively registered with the planning CT enabling 6D corrections, giving CBCT residual errors (CBCTRE) which were compared with ExacTracRE. Results The RMS discrepancies between CBCTRE and ExacTracRE are 1.70mm, 1.66mm, 1.56mm in vertical, longitudinal and lateral directions and 0.27°, 0.49°, 0.35° in yaw, roll and pitch respectively. The corresponding mean discrepancies (and standard deviation) are 0.62mm (1.60mm), 0.00mm (1.68mm), −0.80mm (1.36mm) and 0.05° (0.58°), 0.11° (0.48°), −0.16° (0.32°). Of the 41 CBCT, 17 had high-Z surgical implants. No significant difference in ExacTrac-to-CBCT discrepancy was observed between patients with and without the implants. Conclusion Multiple factors can contribute to the discrepancies between CBCT and ExacTrac: 1) the imaging iso-centers of the two systems, while calibrated to coincide, can be different; 2) the ROI used for registration can be different especially if ribs were included in ExacTrac images; 3) small patient motion can occur between the two verification image acquisitions; 4) the algorithms can be different between CBCT (volumetric) and ExacTrac (radiographic) registrations.

60 APPLIED LIFE SCIENCES↗

SU-G-206-11: The Effect of Table Height On CTDIvol and SSDE in CT Scanning: A Phantom Study

Purpose: Localizer projection radiographs acquired prior to CT scans are used to estimate patient size, affecting the function of Automatic Tube Current Modulation (ATCM) and calculation of the Size Specific Dose Estimate (SSDE). Due to geometric effects, the projected patient size varies with scanner table height and with the orientation of the localizer (AP versus PA). Consequently, variations in scanner table height may affect both CTDIvol and the calculated size-corrected dose index (SSDE). This study sought to characterize these effects. Methods: An anthropomorphic phantom was imaged using an AP localizer, followed by a diagnostic scan using ATCM and our institution’s routine abdomen protocol. This was repeated at various scanner table heights, recording the scanner-reported CTDIvol for each diagnostic scan. The width of the phantom was measured from the localizer and diagnostic images using in-house software. The measured phantom width and scanner-reported CTDIvol were used to calculate SSDE. This was repeated using PA localizers followed by diagnostic scans. Results: 1) The localizer-based phantom width varied by up to 54% of the nominal phantom width between minimum and maximum table heights. 2) Changing the table height caused a variation in scanner-reported CTDIvol of a factor greater than 4.6 when using a PA localizer and almost 2 when using an AP localizer. 3) SSDE, calculated from measured phantom size and scanner-reported CTDIvol, varied by a factor of more than 2.8 when using a PA localizer and almost 1.5 when using an AP localizer. Conclusion: Our study demonstrates that off-center patient positioning affects the efficacy of ATCM, more severely when localizers are acquired in the PA rather than AP projection. Further, patient positioning errors can cause a large variation in the calculated SSDE. This hinders interpretation of SSDE for individual patients and aggregate SSDE data when evaluating CT protocols and clinical practices.

60 APPLIED LIFE SCIENCES↗

Local models for scatter estimation and descattering in polyenergetic X-ray tomography

We propose a new modeling approach for scatter estimation and descattering in polyenergetic X-ray computed tomography (CT) based on fitting models to local neighborhoods of a training set. X-ray CT is widely used in medical and industrial applications. X-ray scatter, if not accounted for during reconstruction, creates a loss of contrast in CT reconstructions and introduces severe artifacts including cupping, shading, and streaks. Even when these qualitative artifacts are not apparent, scatter can pose a major obstacle in obtaining quantitatively accurate reconstructions. Our approach to estimating scatter is, first, to generate a training set of 2D radiographs with and without scatter using particle transport simulation software. To estimate scatter for a new radiograph, we adaptively fit a scatter model to a small subset of the training data containing the radiographs most similar to it. We compared local and global (fit on full data sets) versions of several X-ray scatter models, including two from the recent literature, as well as a recent deep learning-based scatter model, in the context of descattering and quantitative density reconstruction of simulated, spherically symmetrical, single-material objects comprising shells of various densities. Our results show that, when applied locally, even simple models provide state-of-the-art descattering, reducing the error in density reconstruction due to scatter by more than half.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Studies of the Imperfection in Crab Crossing Scheme for Electron-Ion Collider

Crab crossing scheme is the essential scheme that accommodates large crossing angle without loss of luminosity in the design of Electron-Ion collider (EIC). The ideal optics and phase advances of the crab cavity pair are set to create a local crabbing bump in the interaction region (IR). However, there are always small errors in the actual lattice of IR. In this article, we will present the simulation and analytical studies on the imperfections in the crab crossing scheme in the EIC design. The tolerance of the imperfection and the possible remedies can be concluded from these studies.

Hao, Y.↗

Balancing aggregation and smoothing errors in inverse models

Inverse models use observations of a system (observation vector) to quantify the variables driving that system (state vector) by statistical optimization. When the observation vector is large, such as with satellite data, selecting a suitable dimension for the state vector is a challenge. A state vector that is too large cannot be effectively constrained by the observations, leading to smoothing error. However, reducing the dimension of the state vector leads to aggregation error as prior relationships between state vector elements are imposed rather than optimized. Here we present a method for quantifying aggregation and smoothing errors as a function of state vector dimension, so that a suitable dimension can be selected by minimizing the combined error. Reducing the state vector within the aggregation error constraints can have the added advantage of enabling analytical solution to the inverse problem with full error characterization. We compare three methods for reducing the dimension of the state vector from its native resolution: (1) merging adjacent elements (grid coarsening), (2) clustering with principal component analysis (PCA), and (3) applying a Gaussian mixture model (GMM) with Gaussian pdfs as state vector elements on which the native-resolution state vector elements are projected using radial basis functions (RBFs). The GMM method leads to somewhat lower aggregation error than the other methods, but more importantly it retains resolution of major local features in the state vector while smoothing weak and broad features.

Turner, A. J.↗

Design, fabrication and expected performance of the HEAO-B X ray telescope

The X-ray telescope of the second High Energy Astronomy Observatory (HEAO-B) is described. The design consists of four concentric paraboloid/hyperboloid pairs. Focal length is 3.44 meters, paraboloid diameters vary from 580-336 mm, grazing angles vary from 70 to 40 arcmin for paraxial rays, segment lengths are 511 mm. The reflecting surfaces are nickel-coated fused quartz and the resolution is limited by assembly errors (for small field angles) and design aberrations (for large field angles). The local slope, roundness, and surface roughness are identified as the primary optical tolerances connected with the individual elements.

Van Speybroeck, L. P.↗

Retrieval of thermodynamic variables with deep convective clouds Experiments in three dimensions

A three-dimensional numerical cloud model has been used to test a method for retrieving temperature and pressure deviation fields from detailed wind and water fields in deep convective clouds. A comparison of the retrieved fields with the output from the numerical model was used to test the validity of the theoretical treatment and accuracy of the programming. The local time derivatives of each of the velocity components are known to be potential problem sources in using Doppler radar data, and a test was done with this derivative estimated over a 4 min time span rather than 30 s, resulting in excellent agreement with the original solution for this data set. When the local derivative was eliminated, the solution was judged useful for general temperature patterns. Errors due to the inability to measure cloudwater mixing ratio and inaccuracies in rainwater mixing ratio were found to be significant, but not so severe as in the turbulence and steady-state sensitivity tests.

Hane, C. E.↗

Matrix-based, finite-difference algorithms for computational acoustics

A compact numerical algorithm is introduced for simulating multidimensional acoustic waves. The algorithm is expressed in terms of a set of matrix coefficients on a three-point spatial grid that approximates the acoustic wave equation with a discretization error of O(h exp 5). The method is based on tracking a local phase variable and its implementation suggests a convenient coordinate splitting along with natural intermediate boundary conditions. Results are presented for oblique plane waves and compared with other procedures. Preliminary computations of acoustic diffraction are also considered.

Davis, Sanford↗