Search NASA⌕ Search

SEARCH · Search NASA

Results for “convolution”

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 685 records · Page 38

Super resolution for root imaging

Premise High‐resolution cameras are very helpful for plant phenotyping as their images enable tasks such as target vs. background discrimination and the measurement and analysis of fine above‐ground plant attributes. However, the acquisition of high‐resolution images of plant roots is more challenging than above‐ground data collection. An effective super‐resolution (SR) algorithm is therefore needed for overcoming the resolution limitations of sensors, reducing storage space requirements, and boosting the performance of subsequent analyses. Methods We propose an SR framework for enhancing images of plant roots using convolutional neural networks. We compare three alternatives for training the SR model: (i) training with non‐plant‐root images, (ii) training with plant‐root images, and (iii) pretraining the model with non‐plant‐root images and fine‐tuning with plant‐root images. The architectures of the SR models were based on two state‐of‐the‐art deep learning approaches: a fast SR convolutional neural network and an SR generative adversarial network. Results In our experiments, we observed that the SR models improved the quality of low‐resolution images of plant roots in an unseen data set in terms of the signal‐to‐noise ratio. We used a collection of publicly available data sets to demonstrate that the SR models outperform the basic bicubic interpolation, even when trained with non‐root data sets. Discussion The incorporation of a deep learning–based SR model in the imaging process enhances the quality of low‐resolution images of plant roots. We demonstrate that SR preprocessing boosts the performance of a machine learning system trained to separate plant roots from their background. Our segmentation experiments also show that high performance on this task can be achieved independently of the signal‐to‐noise ratio. We therefore conclude that the quality of the image enhancement depends on the desired application.

Ruiz‐Munoz, Jose F.↗

Compendium of photovoltaic degradation rates

Abstract Published data on photovoltaic (PV) degradation measurements were aggregated and re‐examined. The subject has seen an increased interest in recent years resulting in more than 11 000 degradation rates in almost 200 studies from 40 different countries. As studies have grown in number and size, we found an impact from sampling bias attributable to size and accuracy. Because of the correlational nature of this study we examined the data in several ways to minimize this bias. We found median degradation for x‐Si technologies in the 0.5–0.6%/year range with the mean in the 0.8–0.9%/year range. Hetero‐interface technology (HIT) and microcrystalline silicon (µc‐Si) technologies, although not as plentiful, exhibit degradation around 1%/year and resemble thin‐film products more closely than x‐Si. Several studies showing low degradation for copper indium gallium selenide (CIGS) have emerged. Higher degradation for cadmium telluride (CdTe) has been reported, but these findings could reflect a convolution of less accurate studies and longer stabilization periods for some products. Significant deviations for beginning‐of‐life measurements with respect to nameplate rating have been documented over the last 35 years. Therefore, degradation rates that use nameplate rating as reference may be significantly impacted. Studies that used nameplate rating as reference but used solar simulators showed less variation than similar studies using outdoor measurements, even when accounting for different climates. This could be associated with confounding effects of measurement uncertainty and soiling that take place outdoors. Hotter climates and mounting configurations that lead to sustained higher temperatures may lead to higher degradation in some, but not all, products. Wear‐out non‐linearities for the worst performing modules have been documented in a few select studies that took multiple measurements of an ensemble of modules during the lifetime of the system. However, the majority of these modules exhibit a fairly linear decline. Modeling these non‐linearities, whether they occur at the beginning‐of‐life or end‐of‐life in the PV life cycle, has an important impact on the levelized cost of energy. Copyright © 2016 John Wiley & Sons, Ltd.

Jordan, Dirk C.↗

8D likelihood effective Higgs couplings extraction framework in h → 4ℓ

We present an overview of a comprehensive analysis framework aimed at performing direct extraction of all possible effective Higgs couplings to neutral electroweak gauge bosons in the decay to electrons and muons, the so called ‘golden channel’. Our framework is based primarily on a maximum likelihood method constructed from analytic expressions of the fully differential cross sections for h → 4l and for the dominant irreducible $ q\overline{q} $ → 4l background, where 4l = 2e2μ, 4e, 4μ. Detector effects are included by an explicit convolution of these analytic expressions with the appropriate transfer function over all center of mass variables. Utilizing the full set of observables, we construct an unbinned detector-level likelihood which is continuous in the effective couplings. We consider possible ZZ, Zγ, and γγ couplings simultaneously, allowing for general CP odd/even admixtures. A broad overview is given of how the convolution is performed and we discuss the principles and theoretical basis of the framework. This framework can be used in a variety of ways to study Higgs couplings in the golden channel using data obtained at the LHC and other future colliders.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Revisiting single inclusive jet production: timelike factorization and reciprocity

Factorization theorems for single inclusive jet production play a crucial role in the study of jets and their substructure. In the case of small radius jets, the dynamics of the jet clustering can be factorized from both the hard production dynamics, and the dynamics of the low scale jet substructure measurement, and is described by a matching coefficient that can be computed in perturbative Quantum Chromodynamics (QCD). A proposed factorization formula describing this process has been previously presented in the literature, and is referred to as the semi-inclusive, or fragmenting jets formalism. By performing an explicit two-loop calculation, we show the inconsistency of this factorization formula, in agreement with another recent result in the literature. Building on recent progress in the factorization of single logarithmic observables, and the understanding of reciprocity, we then derive a new all-order factorization theorem for inclusive jet production. The use of a jet algorithm, being only a modification of the infrared structure of the measurement, modifies the structure of convolutions in the factorization theorem, as compared to inclusive fragmentation, but maintains the universality of the inclusive hard function and its associated Dokshitzer-Gribov-Lipatov-Altarelli-Parisi (DGLAP) evolution, which are ultraviolet properties. However, the non-trivial structure of convolutions in the factorization theorem implies that the jet functions exhibit a modified evolution. We perform an explicit two-loop calculation of the jet function in both N = 4 super Yang-Mills (SYM), and for all color channels in QCD, finding exact agreement with the structure derived from our renormalization group equations. In addition, we derive several new results, including an extension of our factorization formula to jet substructure observables, a jet algorithm definition of a generating function for the energy correlators, and new results for exclusive jet functions. Our results are a key ingredient for achieving precision jet substructure at colliders.

Effective Field Theories↗

Next-to-leading power endpoint factorization and resummation for off-diagonal “gluon” thrust

The lack of convergence of the convolution integrals appearing in next-to-leading-power (NLP) factorization theorems prevents the applications of existing methods to resum power-suppressed large logarithmic corrections in collider physics. We consider thrust distribution in the two-jet region for the flavour-nonsinglet off-diagonal contribution, where a gluon-initiated jet recoils against a quark-antiquark pair, which is power-suppressed. With the help of operatorial endpoint factorization conditions, we obtain a factorization formula, where the individual terms are free from endpoint divergences in convolutions and can be expressed in terms of renormalized hard, soft and collinear functions in four dimensions. This allows us to perform the first resummation of the endpoint-divergent SCETI observables at the leading logarithmic accuracy using exclusively renormalization-group methods. The presented approach relies on universal properties of the soft and collinear limits and may serve as a paradigm for the systematic NLP resummation for other 1 → 2 and 2 → 1 collider physics processes.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Multi-Regge kinematics and the moduli space of Riemann spheres with marked points

We show that scattering amplitudes in planar N = 4 Super Yang-Mills in multi-Regge kinematics can naturally be expressed in terms of single-valued iterated integrals on the moduli space of Riemann spheres with marked points. As a consequence, scattering amplitudes in this limit can be expressed as convolutions that can easily be computed using Stokes’ theorem. We apply this framework to MHV amplitudes to leading-logarithmic accuracy (LLA), and we prove that at L loops all MHV amplitudes are determined by amplitudes with up to L + 4 external legs. We also investigate non-MHV amplitudes, and we show that they can be obtained by convoluting the MHV results with a certain helicity flip kernel. We classify all leading singularities that appear at LLA in the Regge limit for arbitrary helicity configurations and any number of external legs. In conclusion, we use our new framework to obtain explicit analytic results at LLA for all MHV amplitudes up to five loops and all non-MHV amplitudes with up to eight external legs and four loops.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Bessel-weighted asymmetries in semi-inclusive deep inelastic scattering

The concept of weighted asymmetries is revisited for semi-inclusive deep inelastic scattering. We consider the cross section in Fourier space, conjugate to the outgoing hadron's transverse momentum, where convolutions of transverse momentum dependent parton distribution functions and fragmentation functions become simple products. Individual asymmetric terms in the cross section can be projected out by means of a generalized set of weights involving Bessel functions. Advantages of employing these Bessel weights are that they suppress (divergent) contributions from high transverse momentum and that soft factors cancel in (Bessel-) weighted asymmetries. Also, the resulting compact expressions immediately connect to previous work on evolution equations for transverse momentum dependent parton distribution and fragmentation functions and to quantities accessible in lattice QCD. Furthermore, bessel weighted asymmetries are thus model independent observables that augment the description and our understanding of correlations of spin and momentum in nucleon structure.

Deep Inelastic Scattering↗

The effective number of parameters in kernel density estimation

We devise a new formula for measuring the effective degrees of freedom (EDoF) in kernel density estimation (KDE). Starting from the orthogonal polynomial sequence (OPS) expansion for the ratio of the empirical to the oracle density, we show how convolution with the kernel leads to a new OPS with respect to which one may express the resulting KDE. The expansion coefficients of the two OPS systems can then be related via a kernel sensitivity matrix, which leads to a natural oracle definition of EDoF through the trace operator. Asymptotic properties of the (empirical) plug-in EDoF are worked out through influence functions, and connections with other empirical EDoFs are established. Minimization of Kullback-Leibler divergence is investigated as an alternative to integrated squared error based bandwidth selection rules, yielding a new normal scale rule. The methodology, which arises from a proper oracle formulation and is not restricted to convolution kernels, suggests the possibility of a new bandwidth selection rule based on an information criterion such as AIC.

bandwidth selection↗

Scalable training of trustworthy and energy-efficient predictive graph foundation models for atomistic materials modeling: a case study with HydraGNN

We present our work on developing and training scalable, trustworthy, and energy-efficient predictive graph foundation models (GFMs) using HydraGNN, a multi-headed graph convolutional neural network architecture. HydraGNN expands the boundaries of graph neural network (GNN) computations in both training scale and data diversity. It abstracts over message passing algorithms, allowing both reproduction of and comparison across algorithmic innovations that define nearest-neighbor convolution in GNNs. This work discusses a series of optimizations that have allowed scaling up the GFMs training to tens of thousands of GPUs on datasets consisting of hundreds of millions of graphs. Our GFMs use multitask learning (MTL) to simultaneously learn graph-level and node-level properties of atomistic structures, such as energy and atomic forces. Using over 154 million atomistic structures for training, we illustrate the performance of our approach along with the lessons learned on two state-of-the-art US Department of Energy (US-DOE) supercomputers, namely the Perlmutter petascale system at the National Energy Research Scientific Computing Center and the Frontier exascale system at Oak Ridge Leadership Computing Facility. The HydraGNN architecture enables the GFM to achieve near-linear strong scaling performance using more than 2000 GPUs on Perlmutter and 16,000 GPUs on Frontier.

97 MATHEMATICS AND COMPUTING↗

Online thermal profile prediction for large format additive manufacturing: A hybrid CNN-LSTM based approach

Large format additive manufacturing (LFAM) is an advanced 3D printing technique that efficiently fabricates large-scale components through a layer-by-layer extrusion and deposition process. Accurate surface layer temperature monitoring is essential to prevent manufacturing failures and ensure final product quality. Traditional physics-based offline approaches for simulating thermal behavior are often inefficient and complex, posing challenges on real-time, in-situ monitoring. Here, to address this, we propose a data-driven hybrid CNN-LSTM model to predict sequential thermal images of arbitrary length using real-time infrared thermal imaging. In this approach, a Convolutional Neural Networks (CNN) is trained offline to capture spatial features, reduce dimensional complexity, and enhance time efficiency, while a stacked Long Short-Term Memory (LSTM) is applied online to capture temporal information for improved prediction of future thermal behavior in subsequent printing layers. Model performance is evaluated using MSE, SSIM, and PSNR metrics and is benchmarked against stacked LSTM and convolutional LSTM models, demonstrating superior accuracy and applicability. Additionally, to mitigate noise from moving extruders and gantry backgrounds in thermal images, a fine-tuned semantic segmentation model is implemented offline to extract printing geometry, enabling precise temperature tracking along the tool path for further thermal analysis. The frameworks developed in this study significantly advance temperature monitoring, thermal analysis, and in-situ manufacturing control for LFAM, bridging the gap between theoretical modeling and practical application.

Geometry extraction↗

Nonparametric, data-based kernel interpolation for particle-tracking simulations and kernel density estimation

Traditional interpolation techniques for particle tracking include binning and convolutional formulas that use pre-determined (i.e., closed-form, parameteric) kernels. In many instances, the particles are introduced as point sources in time and space, so the cloud of particles (either in space or time) is a discrete representation of the Green’s function of an underlying PDE. As such, each particle is a sample from the Green’s function; therefore, each particle should be distributed according to the Green’s function. In short, the kernel of a convolutional interpolation of the particle sample “cloud” should be a replica of the cloud itself. This idea gives rise to an iterative method by which the form of the kernel may be discerned in the process of interpolating the Green’s function. When the Green’s function is a density, this method is broadly applicable to interpolating a kernel density estimate based on random data drawn from a single distribution. We formulate and construct the algorithm and demonstrate its ability to perform kernel density estimation of skewed and/or heavy-tailed data including breakthrough curves.

42 ENGINEERING↗

Machine learning the dynamics of quantum kicked rotor

Highlights: • Quantum kicked rotor. • Quantum phase transition. • Machine learning. • LSTM network. • Convolutional neural network. Using the multilayer convolutional neural network (CNN), we can detect the quantum phases in random electron systems, and phase diagrams of two and higher dimensional Anderson transitions and quantum percolations as well as disordered topological systems have been obtained. Here, instead of using CNN to analyze the wave functions, we analyze the dynamics of wave packets via long short-term memory network (LSTM). We adopt the quasi-periodic quantum kicked rotors, which simulate the three and four dimensional Anderson transitions. By supervised training, we let LSTM extract the features of the time series of wave packet displacements in localized and delocalized phases. We then simulate the wave packets in unknown phases and let LSTM classify the time series to localized and delocalized phases. We compare the phase diagrams obtained by LSTM and those obtained by CNN.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

HNF-1B specifically regulates the transcription of the {gamma}a-subunit of the Na{sup +}/K{sup +}-ATPase

Research highlights: {yields} Defects in HNF-1B transcription factor affect Mg{sup 2+} handling in the distal kidney. {yields} {gamma}a- and {gamma}b- subunits of the Na{sup +}/K{sup +}-ATPase colocalize in the distal convoluted tubule of the nephron. {yields} HNF-1B specifically activates {gamma}a expression. {yields} HNF-1B mutants have a dominant negative effect on wild type HNF-1B activity. {yields} Defective transcription of {gamma}a may promote renal Mg{sup 2+} wasting. -- Abstract: Hepatocyte nuclear factor-1B (HNF-1B) is a transcription factor involved in embryonic development and tissue-specific gene expression in several organs, including the kidney. Recently heterozygous mutations in the HNF1B gene have been identified in patients with hypomagnesemia due to renal Mg{sup 2+} wasting. Interestingly, ChIP-chip data revealed HNF-1B binding sites in the FXYD2 gene, encoding the {gamma}-subunit of the Na{sup +}/K{sup +}-ATPase. The {gamma}-subunit has been described as one of the molecular players in the renal Mg{sup 2+} reabsorption in the distal convoluted tubule (DCT). Of note, the FXYD2 gene can be alternatively transcribed into two main variants, namely {gamma}a and {gamma}b. In the present study, we demonstrated via two different reporter gene assays that HNF-1B specifically acts as an activator of the {gamma}a-subunit, whereas the {gamma}b-subunit expression was not affected. Moreover, the HNF-1B mutations H69fsdelAC, H324S325fsdelCA, Y352finsA and K156E, previously identified in patients with hypomagnesemia, prevented transcription activation of {gamma}a-subunit via a dominant negative effect on wild type HNF1-B. By immunohistochemistry, it was shown that the {gamma}a- and {gamma}b-subunits colocalize at the basolateral membrane of the DCT segment of mouse kidney. On the basis of these data, we suggest that abnormalities involving the HNF-1B gene may impair the relative abundance of {gamma}a and {gamma}b, thus affecting the transcellular Mg{sup 2+} reabsorption in the DCT.

60 APPLIED LIFE SCIENCES↗

Automated single-ion peak fitting as an efficient approach for analyzing complex chromatographic data

Chromatography provides important detail on the composition of environmental samples and their chemical processing. However, the complexity of these samples and their tendency to contain many structurally and chemically similar compounds frequently results in convoluted or poorly resolved data. Data reduction from raw chromatograms of complex environmental data into integrated peak areas consequently often requires substantial operator interaction. This difficulty has led to a bottleneck in analysis that increases analysis time, decreases data quality, and will worsen as advances in field-based instrumentation multiply the quantity and informational density of data produced. In this work, we develop and validate an automated approach to fitting chromatographic data within a target retention time window with a combination of multiple idealized peaks (Gaussian peaks either with or without an exponential decay component). Herein we compare this single-ion peak fitting approach to drawn baseline integration methods of more than 70,000 peaks collected by field-based chromatographs spanning across a wide range of volatilities and functionalities. Accuracy of peak fitting under real-world conditions is found to be within 10%. The quantitative parameters describing the fit (e.g. coefficients, fit residuals, etc.) are found to provide valuable information to increase the efficiency of quality control and provide constraints to accurately integrate peaks that are significantly convoluted with neighboring peaks. Implementation of the peak fitting method is shown to yield accurate integration of peaks otherwise too poorly resolved to separate into individual compounds and improved quantitative metrics to determine the fidelity of the data reduction process, while substantially decreasing the time spent by operators on data reduction.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

A fast Fourier transform-based solver for elastic micropolar composites

This work presents a spectral micromechanical formulation for obtaining the full-field and homogenized response of elastic micropolar composites. The algorithm relies on a coupled set of convolution integral equations for the micropolar strains, where periodic Green’s operators associated with a linear homogeneous reference medium are convolved with functions of the Cauchy and couple stress fields that encode the material’s heterogeneity, as well as any potential material nonlinearity. Such convolution integral equations take an algebraic form in the reciprocal Fourier space that can be solved iteratively. In this vein, the fast Fourier transform (FFT) algorithm is leveraged to accelerate the numerical solution, resulting in a mesh-free formulation in which the periodic unit cell representing the heterogeneous material can be discretized by a regular grid of pixels in two dimensions (or voxels in three dimensions). For verification, the numerical solutions obtained with the micropolar FFT solver are compared with analytical solutions for a matrix with a dilute circular inclusion subjected to plane strain loading. The developed computational framework is then used to study length-scale effects and effective (micropolar) moduli of composites with various topological configurations.

97 MATHEMATICS AND COMPUTING↗

A Green’s function fast multipole method for computation of micromechanical fields in heterogeneous materials

Computation of micromechanical fields in heterogeneous materials is usually performed using either the finite element method or the Green’s function method based on FFTs. The finite element method allows for accurate discretization and for non-periodic boundary conditions but is computationally expensive. On the other hand, the FFT-based method is computationally efficient but requires discretization on a regular grid of hexahedral voxels. In this paper, a Green’s function method allowing for accurate discretization using tetrahedral elements and for non-periodic boundary conditions is proposed. The convolution is computed using the fast multipole method, which provides good accuracy even for low-order expansion due to the fast decay of interactions between elements. The proposed Green’s function fast multipole method is verified by comparison with analytical and FFT-based solutions. Furthermore, the computational time is analyzed and compared to the FFT-based method for non-periodic convolution. Finally, effective properties of an elastic polycrystalline microstructure containing thin intergranular cracks are computed and analyzed.

36 MATERIALS SCIENCE↗

Time series anomaly detection in power electronics signals with recurrent and ConvLSTM autoencoders

The anomalies in the high voltage converter modulator (HVCM) remain a major down time for the spallation neutron source facility, that delivers the most intense neutron beam in the world for scientific materials research. In this work, we propose neural network architectures based on Recurrent AutoEncoders (RAE) to detect anomalies ahead of time in the power signals coming from the HVCM. Bi-directional gated recurrent unit, bi-directional long-short term memory (LSTM), and convolutional LSTM (ConvLSTM) are developed, trained, and tested using real experimental signals from the HVCM module. The results show a good performance of the proposed RAE models, achieving precision up to 91%, recall up to 88%, false omission rate as low as 20% (i.e. 80% of the anomalies were detected), and area under the ROC curve up to 0.9. The three RAE models provide very comparable performance, with LSTM showing slightly better performance than GRU and ConvLSTM. The RAE models are benchmarked against other anomaly detection methods, including isolation forest, support vector machine, local outlier factor, feedforward and convolutional autoencoders, and others; showing a better performance. Here, the results of this study demonstrate the promising potential of RAE in anomaly detection for real-world power systems, and for increasing the reliability of the HVCM modules in the spallation neutron source.

42 ENGINEERING↗

Analyzing inference workloads for spatiotemporal modeling

Ensuring power grid resiliency, forecasting climate conditions, and optimization of transportation infrastructure are some of the many application areas where data is collected in both space and time. Spatiotemporal modeling is about modeling those patterns for forecasting future trends and carrying out critical decision-making by leveraging machine learning/deep learning. Once trained offline, field deployment of trained models for near real-time inference could be challenging because performance can vary significantly depending on the environment, available compute resources and tolerance to ambiguity in results. Users deploying spatiotemporal models for solving complex problems can benefit from analytical studies considering a plethora of system adaptations to understand the associated performance-quality trade-offs. To facilitate the co-design of next-generation hardware architectures for field deployment of trained models, it is critical to characterize the workloads of these deep learning (DL) applications during inference and assess their computational patterns at different levels of the execution stack. In this paper, we develop several variants of deep learning applications that use spatiotemporal data from dynamical systems. We study the associated computational patterns for inference workloads at different levels, considering relevant models (Long short-term Memory, Convolutional Neural Network and Spatio-Temporal Graph Convolution Network), DL frameworks (Tensorflow and PyTorch), precision (FP16, FP32, AMP, INT16 and INT8), inference runtime (ONNX and AI Template), post-training quantization (TensorRT) and platforms (Nvidia DGX A100 and Sambanova SN10 RDU). Overall, our findings indicate that although there is potential in mixed-precision models and post-training quantization for spatiotemporal modeling, extracting efficiency from contemporary GPU systems might be challenging. Instead, co-designing custom accelerators by leveraging optimized High Level Synthesis frameworks (such as SODA High-Level Synthesizer for customized FPGA/ASIC targets) can make workload-specific adjustments to enhance the efficiency.

97 MATHEMATICS AND COMPUTING↗