Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “value of data”

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 127 records · Page 7

MCTP system model based on linear programming optimization of apertures obtained from sequencing patient image data maps

Purpose: The authors present a hybrid direct multileaf collimator (MLC) aperture optimization model exclusively based on sequencing of patient imaging data to be implemented on a Monte Carlo treatment planning system (MC-TPS) to allow the explicit radiation transport simulation of advanced radiotherapy treatments with optimal results in efficient times for clinical practice. Methods: The planning system (called CARMEN) is a full MC-TPS, controlled through aMATLAB interface, which is based on the sequencing of a novel map, called “biophysical” map, which is generated from enhanced image data of patients to achieve a set of segments actually deliverable. In order to reduce the required computation time, the conventional fluence map has been replaced by the biophysical map which is sequenced to provide direct apertures that will later be weighted by means of an optimization algorithm based on linear programming. A ray-casting algorithm throughout the patient CT assembles information about the found structures, the mass thickness crossed, as well as PET values. Data are recorded to generate a biophysical map for each gantry angle. These maps are the input files for a home-made sequencer developed to take into account the interactions of photons and electrons with the MLC. For each linac (Axesse of Elekta and Primus of Siemens) and energy beam studied (6, 9, 12, 15 MeV and 6 MV), phase space files were simulated with the EGSnrc/BEAMnrc code. The dose calculation in patient was carried out with the BEAMDOSE code. This code is a modified version of EGSnrc/DOSXYZnrc able to calculate the beamlet dose in order to combine them with different weights during the optimization process. Results: Three complex radiotherapy treatments were selected to check the reliability of CARMEN in situations where the MC calculation can offer an added value: A head-and-neck case (Case I) with three targets delineated on PET/CT images and a demanding dose-escalation; a partial breast irradiation case (Case II) solved with photon and electron modulated beams (IMRT + MERT); and a prostatic bed case (Case III) with a pronounced concave-shaped PTV by using volumetric modulated arc therapy. In the three cases, the required target prescription doses and constraints on organs at risk were fulfilled in a short enough time to allow routine clinical implementation. The quality assurance protocol followed to check CARMEN system showed a high agreement with the experimental measurements. Conclusions: A Monte Carlo treatment planning model exclusively based on maps performed from patient imaging data has been presented. The sequencing of these maps allows obtaining deliverable apertures which are weighted for modulation under a linear programming formulation. The model is able to solve complex radiotherapy treatments with high accuracy in an efficient computation time.

60 APPLIED LIFE SCIENCES↗

Probabilistic Data-Driven Sampling via Multi-Criteria Importance Analysis

Although supercomputers are becoming increasingly powerful, their components have thus far not scaled proportionately. Compute power is growing enormously and is enabling finely resolved simulations that produce never-before-seen features. However, I/O capabilities lag by orders of magnitude, which means only a fraction of the simulation data can be stored for post hoc analysis. Prespecified plans for saving features and quantities of interest do not work for features that have not been seen before. Data-driven intelligent sampling schemes are needed to detect and save important parts of the simulation while it is running. Here, we propose a novel sampling scheme that reduces the size of the data by orders-of-magnitude while still preserving important regions. The approach we develop selects points with unusual data values and high gradients. Finally, we demonstrate that our approach outperforms traditional sampling schemes on a number of tasks.

97 MATHEMATICS AND COMPUTING↗

ARM Trajectories Data Set Value-Added Product Report

The U.S. Department of Energy Atmospheric Radiation Measurement (ARM) user facility’s ARM Trajectories Data Set (ARMTRAJ) Value-Added Product (VAP) provides trajectory data sets initialized at ARM deployment coordinates and configured using ARM data sets. The four trajectory data sets support aerosol, cloud, and planetary boundary-layer research. Trajectory calculations use the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model informed by the European Centre for Medium-Range Weather Forecasts (ECMWF) fifth-generation atmospheric reanalysis (ERA5) data set at its highest spatial resolution (~31 km). HYSPLIT also runs at multiple initial starting locations surrounding ARM deployments (in latitude/longitude and/or vertical coordinates), facilitating an ensemble for each sample in the data sets. The ensemble mean and variability reported in ARMTRAJ improve the fidelity and provide uncertainty estimates of trajectory coordinates, thermodynamic properties, and other output fields.

54 ENVIRONMENTAL SCIENCES↗

ARM Trajectories Data Set Value-Added Product Report

The U.S. Department of Energy Atmospheric Radiation Measurement (ARM) user facility’s ARM Trajectories Data Set (ARMTRAJ) Value-Added Product (VAP) provides trajectory data sets initialized at ARM deployment coordinates and configured using ARM data sets. The six trajectory data sets support aerosol, cloud, planetary boundary layer, and related research (aerosol-cloud interactions, etc.), as well as studies using ARM Aerial Facility (AAF) and tethered balloon system (TBS) measurements. Trajectory calculations use the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model informed by the European Centre for Medium-Range Weather Forecasts (ECMWF) fifth-generation atmospheric reanalysis (ERA5) data set at its highest spatial resolution (~31 km). HYSPLIT also runs at multiple initial starting locations surrounding ARM deployments (in latitude/longitude and/or vertical coordinates), facilitating an ensemble for each sample in the data sets. The ensemble mean and variability reported in ARMTRAJ improve the fidelity and provide uncertainty estimates of trajectory coordinates, thermodynamic properties, and other output fields.

54 ENVIRONMENTAL SCIENCES↗

Algorithms for Non-Negative Matrix Factorization on Noisy Data With Negative Values

Non-negative matrix factorization (NMF) is a dimensionality reduction technique that has shown promise for analyzing noisy data, especially astronomical data. For these datasets, the observed data may contain negative values due to noise even when the true underlying physical signal is strictly positive. Prior NMF work has not treated negative data in a statistically consistent manner, which becomes problematic for low signal-to-noise data with many negative values. In this paper we present two algorithms, Shift-NMF and Nearly-NMF, that can handle both the noisiness of the input data and also any introduced negativity. Both of these algorithms use the negative data space without clipping or masking and recover non-negative signals without any introduced positive offset that occurs when clipping or masking negative data. We demonstrate this numerically on both simple and more realistic examples, and prove that both algorithms have monotonically decreasing update rules.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Gradient-based Optimization for Regression in the Functional Tensor-Train Format

We consider the task of low-multilinear-rank functional regression, i.e., learning a low-rank parametric representation of functions from scattered real-valued data. Our first contribution is the development and analysis of gradient-based optimization procedures, including stochastic gradient descent and quasi-Newton methods, for learning the parameters of a functional tensor-train (FT) that yields improved accuracy over standard alternating least squares methods. The functional tensor-train uses the tensor-train (TT) representation of low-rank arrays as an ansatz for a class of low-multilinear-rank functions. The FT is represented by a set of matrix-valued functions that contain a set of univariate functions, and the regression task is to learn the parameters of these univariate functions. Our second contribution demonstrates that using nonlinearly parameterized univariate functions, e.g., symmetric kernels with moving centers, within each core can outperform the standard approach of using a linear expansion of basis functions. Our final contributions are new rank adaptation and group-sparsity regularization procedures to minimize overfitting. We use several benchmark problems to demonstrate at least an order of magnitude lower accuracy with gradient-based optimization methods than ALS in the low-sample number regime. We also demonstrate an order of magnitude reduction in accuracy on a test problem resulting from nonlinear parameterizations over linear parameterizations. Finally we compare regression performance with 22 other nonparametric and parametric regression methods on 10 real-world data sets. We achieve top-five accuracy for seven of the data sets and best accuracy for two of the data sets. These rankings are the best amongst parametric models and competetive with the best non-parametric methods.

97 MATHEMATICS AND COMPUTING↗

Deep anomaly detection for industrial systems: a case study

We explore the use of deep neural networks for anomaly detection of industrial systems where the data are multivariate time series measurements. We formulate the problem as a self-supervised learning where data under normal operation are used to train a deep neural network autoregressive model, i.e., use a window of time series data to predict future data values. The aim of such a model is to learn to represent the system dynamic behavior under normal conditions, while expect higher model vs. measurement discrepancies under faulty conditions. In real world applications, many control settings are discrete in nature. In this paper, vector embedding and joint losses are employed to deal with such situations. Both LSTM and CNN based deep neural network backbones are studied on the Secure Water Treatment (SWaT) testbed datasets. Also, Support Vector Data Description (SVDD) method is adapted to such anomaly detection settings with deep neural networks. Evaluation methods and results are discussed based on the SWaT dataset along with potential pitfalls.

anomaly detection, deep neural network↗

Quenching of single-particle strength inferred from nucleon-removal transfer reactions on 15 C

The difference in the proton and neutron separation energies (ΔS) of the weakly bound 15 C ground state is -19.86 MeV, an extreme value. Data from intermediate-energy heavy-ion induced (HI-induced) knockout reactions on nuclei spanning -20 ≲ ΔS ≲ +20 MeV, suggest that the degree to which single-particle strength is quenched, R s , has a negative correlation with ΔS, decreasing from unity around -20 MeV to around 0.2 at +20 MeV. For the 15 C ground state (R s = 0.96 (4) in HI-induced knockout), contrasting results have recently been obtained via the neutron-adding transfer reaction, which reveal a value of R s = 0.64 (15), similar to the value observed at modest and more extreme values of ΔS with reaction probes other than HI knockout. In order to explore the any potential differences between adding and removing processes in transfer reactions at extreme ΔS, single-neutron removal transfer reactions on 15 C were performed at 7.1 MeV/u in inverse kinematics. The removal of a valence neutron in 2s 1/2 orbit using both (p, d) and (d, t) reactions shows consistent quenching factors and agrees with those from the neutron-adding reaction. The present results, which can be compared with neutron knockout reaction, suggest that correlations, represented by the quenching factor, show limited dependence on neutron-proton asymmetry under the most extreme asymmetry conditions so far achieved in transfer reactions.

quenching factor↗

Scalable analysis tools for sensitivity analysis and UQ (3160) results.

The 9/30/2009 ASC Level 2 Scalable Analysis Tools for Sensitivity Analysis and UQ (Milestone 3160) contains feature recognition capability required by the user community for certain verification and validation tasks focused around sensitivity analysis and uncertainty quantification (UQ). These feature recognition capabilities include crater detection, characterization, and analysis from CTH simulation data; the ability to call fragment and crater identification code from within a CTH simulation; and the ability to output fragments in a geometric format that includes data values over the fragments. The feature recognition capabilities were tested extensively on sample and actual simulations. In addition, a number of stretch criteria were met including the ability to visualize CTH tracer particles and the ability to visualize output from within an S3D simulation.

97 MATHEMATICS AND COMPUTING↗

A review of coal heating value correlations with application to coal char, tar, and other fuels

Accurately measuring a fuel’s heating value is one of the first steps in the classification of a new fuel. Heating values are widely used in coal combustion research and are becoming more useful in other fuel types as well. Many different empirical correlations to predict heating values based on primary organic (CHONS) elemental composition are found in the literature, many of which were originally created to predict heating values of parent coal, and some of which have been extended for biomass. However, no correlations exist for heating values of coal chars and tars. Thirteen literature heating value correlations (10 model forms with 13 unique sets of coefficients) were evaluated for their predictive accuracy of a wide variety of fuel types, including 353 coals. This comparative analysis showed that the predictive capabilities of many of these literature models are very similar for coal-based fuels, with good correlation of coal char heating values. Correlation of coal tar heating values was hampered by the lack of heating value data. The choice of which model to use becomes more important for biomass fuels as well as for more unique fuel types like propellants and explosives, municipal solid waste, etc. Finally, the best model forms for each fuel type are suggested based on several statistical measures of fit.

01 COAL, LIGNITE, AND PEAT↗

Poisson Log-Normal Process for Count Data Prediction

Modeling count data is important in physics and other scientific disciplines, where measurements often involve discrete, non-negative quantities such as photon or neutrino detection events. Traditional parametric approaches can be trained to generate integer-count predictions but may struggle with capturing complex, non-linear dependencies often observed in the data. Gaussian process (GP) regression provides a robust non-parametric alternative to modeling continuous data; however, it cannot generate integer outputs. We propose the Poisson Log-Normal (PoLoN) process, a framework that employs GP to model Poisson log-rates. As in GP regression, our approach relies on the correlations between data points captured via GP kernel structure rather than explicit functional parameterizations. We demonstrate that the PoLoN predictive distribution is Poisson-LogNormal and provide an algorithm for optimizing kernel hyperparameters. Furthermore, we adapt the PoLoN approach to the problem of detecting weak localized signals superimposed on a smoothly varying background - a task of considerable interest in many areas of science and engineering. Our framework allows us to predict the strength, location and width of the detected signals. We evaluate PoLoN's performance using both synthetic and real-world datasets, including the open dataset from CERN which was used to detect the Higgs boson at the Large Hadron Collider. Our results indicate that the PoLoN process can be used as a non-parametric alternative for analyzing, predicting, and extracting signals from integer-valued data.

Saha, Anushka [Rutgers U., Piscataway]↗

Evaluation of performance indicators applied to a material recovery facility fed by mixed packaging waste

Highlights: • Material Flows assessment of a Material Recovery Facility has been realized. • Performance Indicators for a generic MRF have been proposed. • A Dashboard has been utilized to measure the facility performance. • Experimental values of incoming waste and output products composition are reported. - Abstract: Most of the integrated systems for municipal solid waste management aim to increase the recycling of secondary materials by means of physical processes including sorting, shredding and reprocessing. Several restrictions prevent from reaching a very high material recycling efficiency: the variability of the composition of new-marketed materials used for packaging production and its shape and complexity are critical issues. The packaging goods are in fact made of different materials (aluminium, polymers, paper, etc.), possibly assembled, having different shape (flat, cylindrical, one-dimensional, etc.), density, colours, optical properties and so on. These aspects limit the effectiveness and efficiency of the sorting and reprocessing plants. The scope of this study was to evaluate the performance of a large scale Material Recovery Facility (MRF) by utilizing data collected during a long period of monitoring. The database resulted from the measured data has been organized in four sections: (1) data related to the amount and type of inlet waste; (2) amount and composition of output products and waste; (3) operating data (such as worked hours for shift, planned and unscheduled maintenance time, setting parameters of the equipment, and energy consumption for shift); (4) economic data (value of each product, disposal price for the produced waste, penalty for non-compliance of products and waste, etc.). A part of this database has been utilized to build an executive dashboard composed by a set of performance indicators suitable to measure the effectiveness and the efficiency of the MRF operations. The dashboard revealed itself as a powerful tool to support managers and engineers in their decisions in respect to the market demand or compliance regulation variation as well as in the designing of the lay-out improvements. The results indicated that the 40% of the input waste was recovered as valuable products and that a large part of these (88%) complied with the standards of the recycling companies. The evaluation of the indicators led to the decision to modify the layout to improve the interception of some polymers for which the performance indicators were poor. In particular, two additional optical sorters have to be inserted to increase the yield indicator and to the overall performance of the facility. Definitely, the results of the work allowed to: increase the yield and purity of products’ flows; ensure the compliance of waste flows; increase the workability.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Accelerating Relative-error Bounded Lossy Compression for HPC datasets with Precomputation-Based Mechanisms

Scientific simulations in high-performance computing (HPC) environments are producing vast volume of data, which may cause a severe I/O bottleneck at runtime and a huge burden on storage space for post-analysis. Unlike the traditional data reduction schemes (such as deduplication or lossless compression), not only can error-controlled lossy compression significantly reduce the data size but it can also hold the promise to satisfy user demand on error control. Point-wise relative error bounds (i.e., compression errors depends on the data values) are widely used by many scientific applications in the lossy compression, since error control can adapt to the precision in the dataset automatically. Point-wise relative error bounded compression is complicated and time consuming. In this work, we develop efficient precomputation-based mechanisms in the SZ lossy compression framework. Our mechanisms can avoid costly logarithmic transformation and identify quantization factor values via a fast table lookup, greatly accelerating the relative-error bounded compression with excellent compression ratios. In addition, our mechanisms also help reduce traversing operations for Huffman decoding, and thus significantly accelerate the decompression process in SZ. Experiments with four well-known real-world scientific simulation datasets show that our solution can improve the compression rate by about 30% and decompression rate by about 70% in most of cases, making our designed lossy compression strategy the best choice in class in most cases.

compression rate↗

Performance Optimization for Relative-Error-Bounded Lossy Compression on Scientific Data

Scientific simulations in high-performance computing (HPC) environments generate vast volume of data, which may cause a severe I/O bottleneck at runtime and a huge burden on storage space for postanalysis. Unlike traditional data reduction schemes such as deduplication or lossless compression, not only can error-controlled lossy compression significantly reduce the data size but it also holds the promise to satisfy user demand on error control. Pointwise relative error bounds (i.e., compression errors depends on the data values) are widely used by many scientific applications with lossy compression since error control can adapt to the error bound in the dataset automatically. Pointwise relative-error-bounded compression is complicated and time consuming. In this article, we develop efficient precomputation-based mechanisms based on the SZ lossy compression framework. Our mechanisms can avoid costly logarithmic transformation and identify quantization factor values via a fast table lookup, greatly accelerating the relative-error-bounded compression with excellent compression ratios. In addition, we reduce traversing operations for Huffman decoding, significantly accelerating the decompression process in SZ. Experiments with eight well-known real-world scientific simulation datasets show that our solution can improve the compression and decompression rates (i.e., the speed) by about 40 and 80 p, respectively, in most of cases, making our designed lossy compression strategy the best-in-class solution in most cases.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Discrete beta dose kernel matrices for nuclides applied in targeted radionuclide therapy (TRT) calculated with MCNP5

Purpose: Radiopharmaceuticals administered in targeted radionuclide therapy (TRT) rely to a great extent not only on beta-emitting nuclides but also on emitters of monoenergetic electrons. Recent advances like combined PET/CT devices, the consequential coregistration of both data, the concept of using beta couples for diagnosis and therapy, respectively, as well as the development of voxel models offer a great potential for developing TRT dose calculation systems similar to those available for external beam treatment planning. The deterministic algorithms in question for this task are based on the convolution of three-dimensional matrices, one representing the activity distribution and the other the dose point kernel. This study aims to report on three-dimensional kernel matrices for various nuclides used in TRT. Methods: The Monte Carlo code MCNP5 was used to calculate discrete dose kernels of beta particles including the contributions from their respective secondary radiation in soft tissue for the following nuclides: {sup 32}P, {sup 33}P, {sup 67}Cu, {sup 89}Sr, {sup 90}Y, {sup 103}Rh{sup m}, {sup 131}I, {sup 177}Lu, {sup 186}Re, and {sup 188}Re. For each nuclide a kernel cube of 10x10x10 mm{sup 3} was calculated, the dimensions of a voxel being 1 mm{sup 3}. Additional kernels with voxel sizes of 3x3x3 mm{sup 3} were simulated. Results: Comparison with the S-value data regarding {sup 32}P, {sup 89}Sr, {sup 90}Y, and {sup 131}I of the MIRD committee which were calculated with the EGS4 code showed a very good agreement, the secondary particle transport of {sup 90}Y being the only exception. Documented analytical kernels on the other side show deviations very close and very far to the source. Conclusions: The good accordance with the only discrete dose kernels published up to date justifies the method chosen. Together with the additional six nuclides, this report provides a considerable database for three-dimensional kernel matrices with regard to beta radionuclides applied in TRT. In contrast to analytical dose point kernels, the discrete kernels elude the problem of overestimation near the source and take energy depositions into account, which occur beyond the range of the continuous-slowing-down approximation (csda range). Recalculation of the 1x1x1 mm{sup 3} kernels to other dose kernels with varying voxel dimensions, cubic or noncubic, is shown to be easily manageable and thereby provides a resolution-independent system of dose calculation.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Quantum theory as a description of robust experiments: Derivation of the Pauli equation

It is shown that the Pauli equation and the concept of spin naturally emerge from logical inference applied to experiments on a charged particle under the conditions that (i) space is homogeneous (ii) the observed events are logically independent, and (iii) the observed frequency distributions are robust with respect to small changes in the conditions under which the experiment is carried out. The derivation does not take recourse to concepts of quantum theory and is based on the same principles which have already been shown to lead to e.g. the Schrödinger equation and the probability distributions of pairs of particles in the singlet or triplet state. Application to Stern–Gerlach experiments with chargeless, magnetic particles, provides additional support for the thesis that quantum theory follows from logical inference applied to a well-defined class of experiments. - Highlights: • The Pauli equation is obtained through logical inference applied to robust experiments on a charged particle. • The concept of spin appears as an inference resulting from the treatment of two-valued data. • The same reasoning yields the quantum theoretical description of neutral magnetic particles. • Logical inference provides a framework to establish a bridge between objective knowledge gathered through experiments and their description in terms of concepts.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Robust Computation of Morse-Smale Complexes of Bilinear Functions

The Morse-Smale (MS) complex has proven to be a useful tool in extracting and visualizing features from scalar-valued data. However, existing algorithms to compute the MS complex are restricted to either piecewise linear or discrete scalar fields. This paper presents a new combinatorial algorithm to compute MS complexes for two dimensional piecewise bilinear functions defined on quadrilateral meshes. We derive a new invariant of the gradient flow within a bilinear cell and use it to develop a provably correct computation which is unaffected by numerical instabilities. This includes a combinatorial algorithm to detect and classify critical points as well as a way to determine the asymptotes of cell-based saddles and their intersection with cell edges. Finally, we introduce a simple data structure to compute and store integral lines on quadrilateral meshes which by construction prevents intersections and enables us to enforce constraints on the gradient flow to preserve known invariants.

97 MATHEMATICS AND COMPUTING↗

Benchmarking sensor fusion capabilities of an integrated instrumentation package

Quantifying and mitigating environmental risks introduced by marine energy conversion systems requires a variety of sensors (active acoustic, passive acoustic, and optical). The operation of these sensors must satisfy three directives to be effective: (1) do not alter the environment through operation of sensors; (2) capture rare events; and (3) do not accrue unmanageable volumes of low-value data. This requires integrating sensors into a single package, rather than operating them independently. The Adaptable Monitoring Package is an integrated instrumentation package that combines a multibeam sonar, acoustic camera, current profiler, optical cameras, and an array of hydrophones. The capabilities and limitations of the AMP sensors were benchmarked using cooperative targets, and real-time target tracking and detection was used to detect opportunistic targets (e.g., diving birds, seals). During an initial deployment, automatic detection of opportunistic targets achieved a 58% true positive rate and a 99% true negative rate (100% corresponding to an ideal system in both cases). In post-processing, target tracking data were used to evaluate automatic target classification capabilities using a k-nearest neighbor algorithm. Findings suggest that real-time target classification should be possible and enable integrated instrumentation systems to meet the monitoring needs of marine energy deployments.

16 TIDAL AND WAVE POWER↗