Search NASA⌕ Search

SEARCH · Search NASA

Results for “Training Analysis”

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

Documented Safety Analysis Addendum for the Neutron Radiography Reactor Facility Core Conversion

The Neutron Radiography Reactor Facility (NRAD) is a Training, Research, Isotope Production, General Atomics (TRIGA) reactor which was installed in the Idaho National Laboratory (INL) Hot Fuels Examination Facility (HFEF) at the Materials and Fuels Complex (MFC) in the mid 1970s. The facility provides researchers the capability to examine both irradiated and non-irradiated materials in support of reactor fuel and components programs through non-destructive neutron radiography examination. The facility has been used in the past as one facet of a suite of reactor fuels and component examination facilities available to researchers at the INL and throughout the DOE complex. The facility has also served various commercial research activities in addition to the DOE research and development support. The reactor was initially constructed using Fuel Lifetime Improvement Program (FLIP)- type highly enriched uranium (HEU) fuel obtained from the dismantled Puerto Rico Nuclear Center (PRNC) reactor. In accordance with international non-proliferation agreements, the NRAD core will be converted to a low enriched uranium (LEU) fuel and will continue to utilize the PRNC control rods, control rod drives, startup source, and instrument console as was previously used with the HEU core. The existing NRAD Safety Analysis Report (SAR) was created and maintained in the preferred format of the day, combining sections of both DOE-STD-3009 and Nuclear Regulatory Commission Regulatory Guide 1.70. An addendum was developed to cover the refueling and reactor operation with the LEU core. This addendum follows the existing SAR format combining required formats from both the DOE and NRC. This paper discusses the project to successfully write a compliant and approved addendum to the existing safety basis documents.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Analysis of two-color photoelectron spectroscopy for attosecond metrology at seeded free-electron lasers

The generation of attosecond pulse trains at free-electron lasers opens new opportunities in ultrafast science, as it gives access, for the first time, to reproducible, programmable, extreme ultraviolet (XUV) waveforms with high intensity. In this work, we present a detailed analysis of the theoretical model underlying the temporal characterization of the attosecond pulse trains recently generated at the free-electron laser FERMI. In particular, the validity of the approximations used for the correlated analysis of the photoelectron spectra generated in the two-color photoionization experiments are thoroughly discussed. The ranges of validity of the assumptions, in connection with the main experimental parameters, are derived.

74 ATOMIC AND MOLECULAR PHYSICS↗

Line Faults Classification Using Machine Learning on Three Phase Voltages Extracted from Large Dataset of PMU Measurements

An end-to-end supervised learning method is developed to classify transmission line faults in a twoyear field-recorded dataset that includes synchronized measurements of three-phase voltages recorded by 38 Phasor Measurement Units (PMU) sparsely located in in the US Western Grid interconnection. Statistical analysis is performed to extract features from this large dataset to train Support Vector Machine (SVM), Random Forest (RF), and eXtreme Gradient Boosting (XGBoost) classifiers initially. The training further leverages a simulated dataset from a synthetic grid with 12 PMUs to increase the number of faults of types infrequently seen in the field-recorded dataset. Training the classification models with the combined dataset resulted in a classification accuracy of 97.7%. This is a significant improvement over 89.7% to 92.5% accuracy obtained by relying on the field-recorded dataset alone.

47 OTHER INSTRUMENTATION↗

Temporal analysis of multispectral scanner data.

Multispectral scanner reflectance data were sampled for bare soil, cotton, sorghum, corn, and citrus at four dates during a growing season (April, May, June, and July 1969) to develop a time-dependent signature for crop and soil discrimination. Discrimination tests were conducted for single-date and multidate formats using training and test data sets. For classifications containing several crops, the multidate or temporal approach improved discrimination compared with the single-date approach. The multidate approach also preserved recognition accuracy better in going from training fields to test fields than the single-date analysis. The spectral distinctiveness of bare soil versus vegetation resulted in essentially equal discrimination using single-date versus multidate data for those two categories.

Richardson, A. J.↗

Audacity of huge: overcoming challenges of data scarcity and data quality for machine learning in computational materials discovery

Machine learning (ML)-accelerated discovery requires large amounts of high-fidelity data to reveal predictive structure–property relationships. For many properties of interest in materials discovery, the challenging nature and high cost of data generation has resulted in a data landscape that is both scarcely populated and of dubious quality. Data-driven techniques starting to overcome these limitations include the use of consensus across functionals in density functional theory, the development of new functionals or accelerated electronic structure theories, and the detection of where computationally demanding methods are most necessary. When properties cannot be reliably simulated, large experimental data sets can be used to train ML models. In the absence of manual curation, increasingly sophisticated natural language processing and automated image analysis are making it possible to learn structure–property relationships from the literature. Finally, models trained on these data sets will improve as they incorporate community feedback.

36 MATERIALS SCIENCE↗

Design of Neural Networks for Fast Convergence and Accuracy: Dynamics and Control

A procedure for the design and training of artificial neural networks, used for rapid and efficient controls and dynamics design and analysis for flexible space systems, has been developed. Artificial neural networks are employed, such that once properly trained, they provide a means of evaluating the impact of design changes rapidly. Specifically, two-layer feedforward neural networks are designed to approximate the functional relationship between the component/spacecraft design changes and measures of its performance or nonlinear dynamics of the system/components. A training algorithm, based on statistical sampling theory, is presented, which guarantees that the trained networks provide a designer-specified degree of accuracy in mapping the functional relationship. Within each iteration of this statistical-based algorithm, a sequential design algorithm is used for the design and training of the feedforward network to provide rapid convergence to the network goals. Here, at each sequence a new network is trained to minimize the error of previous network. The proposed method should work for applications wherein an arbitrary large source of training data can be generated. Two numerical examples are performed on a spacecraft application in order to demonstrate the feasibility of the proposed approach.

Maghami, Peiman G.↗

Training Students to Analyze Spatial and Temporal Heterogeneities in Reservoir and Seal Petrology, Mineralogy, and Geochemistry: Implications for CO{sub 2} Sequestration Prediction, Simulation, and Monitoring

The objective of this project was to expose and train multiple students in geological tools that are essential to reservoir characterization and geologic sequestration including but not limited to advanced petrological methods, mineralogical methods, and geochemical methods; core analysis, and geophysical well-log interpretation. These efforts have included training of multiple students through geologically based curriculum and research using advanced petrological, mineralogical, and geochemical methods. In whole, over the last 3+ years, this award has supported 5,828 hours of student research, supporting the work of several graduate and undergraduate students. They have all received training directly related to ongoing CO{sub 2} sequestration demonstrations. The students have all conducted original scientific research on topics related to understanding the importance of lithological, textural, and compositional variability in formations that are being targeted as CO{sub 2} sequestration reservoirs and seals. This research was linked to the Mount Simon Sandstone reservoir and overlying Eau Claire Formation seal in the Illinois Basin- a system where over one million tons of CO{sub 2} are actively being injected with the first large-scale demonstration of anthropogenic CO{sub 2} storage in the U.S. Student projects focused specifically on 1) reservoir porosity characterization and evaluation, 2) petrographic, mineralogical, and geochemical evidence of fluid-related diagenesis in the caprock, 3) textural changes in reservoir samples exposed to experimental CO{sub 2} + brine conditions, 4) controls on spatial heterogeneity in composition and texture in both the reservoir and seal, 5) the implications of small-scale fractures within the reservoir, and 6) petrographic and stable isotope analyses of carbonates in the seal to understand the burial history of the system. The student-led research associated with this project provided real-time and hands-on experience with a relevant CO{sub 2} system, provided relevant information to the regional partnerships who are working within these formations, and provides more broadly applicable understanding and method development for other carbon capture and storage systems.

54 ENVIRONMENTAL SCIENCES↗

Specialty chemicals production case study: Economic analysis of modular chemical process intensification versus conventional stick‐built approaches

Abstract The beneficial synergies of chemical process intensification and plant modularization present a unique step‐wise advancement opportunity for many chemical manufacturers, but economic case studies are needed to raise awareness of such opportunities. The primary objective of this case study is to better understand the business case economics of a specialty chemical plant using modular chemical process intensification (MCPI), by comparing it with that of a conventional stick‐built (CSB) plant that produces the same product at the same production capacity. When comparing MCPI against CSB approaches for a plant project strategy decision, analysts should thoroughly understand and model the differences and similarities in scope bases. The MCPI approach for this case study benefitted from dramatic reductions in capital expenditures (CAPEX). A sizeable reduction in plant spatial volume likely explains some of these reductions. Sizeable operational expenditure (OPEX) reductions with the MCPI plant appear to be associated with the reduced operator staffing required from converting a labor‐intensive batch process into an automated, continuous flow process. Traditional project investment economic measures strongly favored the MCPI case for the design, construction, and operation of the specialty chemical plant. The net present value for the MCPI case was nearly twice that for the CSB case over a ten‐year period, and the payback period for the CSB case was nearly five times longer than that of the MCPI case. The opinion‐based perspective of the study participant identified significant contributors to superior MCPI economic performance for this case study. With the two conditions of low MCPI CAPEX and high product profit margin, economic analysis indicates that incorporation of a backup MCPI train, for operational redundancy when downtime occurs, is a very beneficial strategy.

O'Connor, James T.↗

Virtual monoenergetic reconstruction of contrast-enhanced CT scans of the abdomen and pelvis at 40 keV improves the detection of peritoneal metastatic deposits

Purpose: To evaluate the role of virtual monoenergetic imaging (VMI) in the detection of peritoneal metastatic disease in contrast-enhanced computed tomography (CT) of the abdomen and pelvis and to compare this technique to the conventional 120 kV mixed dataset. Materials and methods: Institutional review board approval was obtained with no informed consent required for this retrospective analysis. 43 consecutive patients with histopathologically confirmed peritoneal disease were scanned using a standard protocol on a 128-section dual-source, dual-energy CT system (100/140 keV). Scans were retrospectively reconstructed at VMI energy levels from 40–110 keV in 10 keV increments and were analyzed both quantitatively and qualitatively. CNR values for peritoneal metastatic deposits were recorded using region of interest (ROI) analysis at each energy level for all VMI datasets. Subjective analysis was performed by two independent fellowship-trained readers with combined experience of greater than 15 years. Qualitative parameters included diagnostic acceptability, subjective noise, and contrast resolution and confidence. Results: The contrast-to-noise ratios (CNRs) for peritoneal metastatic deposits at the different VMI energy levels were compared using a one-way ANOVA with Tukey Post Test, and the optimal CNR was observed at 40 keV (p < 0.0001). Qualitative parameters were compared using a Paired T Test. Subjective noise, diagnostic acceptability, and contrast resolution was significantly better on the conventional images, but readers reported increased confidence on VMI at 40 keV (p < 0.001). Conclusion: VMI reconstruction of contrast-enhanced dual-energy CT scans of the abdomen and pelvis at 40 keV maximizes the conspicuity of metastatic peritoneal deposits and improves radiologists’ diagnostic confidence compared with conventional CT images. We recommend using virtual monoenergetic datasets at 40 keV as a tool for improving the detection of these lesions in routine clinical practice.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Opportunistic bone density screening for the abdominal radiologist using colored CT images: a pilot retrospective study

Purpose: The purpose of the study was to develop an accurate and reproducible method for detecting low spinal bone density on abdominal CT images. Methods: For this IRB-approved HIPAA-compliant single-center retrospective study, nonenhanced CT images of the lower abdomen were obtained in 631 African-American participants. Mean attenuation of L3/L4 was associated with quantitative CT bone density (QCT) in a randomly selected training cohort (N = 511), and receiver operating characteristics analysis was used to identify the optimal mean attenuation threshold for differentiating normal from low bone density. Custom image processing software was used to generate grayscale and colored CT images of the midline spine, with green for normal and red for low bone density. Five radiologists independently assessed bone density at L3/L4 in a validation cohort (N = 120) using various methods: QCT, visual assessment of sagittal grayscale images (Grayscale), quantitative measurement of mean attenuation on a midline sagittal image (Attenuation), and visual assessment of a midline sagittal colored image (Color). Accuracy was calculated using the average QCT bone density as a reference standard. Inter-observer agreement was assessed using intraclass correlation coefficient (ICC). Results: The optimal mean attenuation threshold for differentiating normal from low bone density at L3/L4 was 145 Hounsfield Units. The average accuracy of Grayscale, Attenuation, and Color methods was 58, 87, and 91% (p < 0.001), respectively. Inter-observer agreement was poor for Grayscale (ICC: 0.20; 95% CI 0.12, 0.28) and excellent for both Attenuation (ICC: 0.85; 95% CI 0.73, 0.91) and Color methods (ICC: 0.87; 95% CI 0.83, 0.90). Conclusion: Detection of low spinal bone density using colored abdominal CT images was highly accurate and reproducible.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Instance Segmentation for Direct Measurements of Satellites in Metal Powders and Automated Microstructural Characterization from Image Data

In this work, we propose instance segmentation as a useful tool for image analysis in materials science. Instance segmentation is an advanced technique in computer vision which generates individual segmentation masks for every object of interest that is recognized in an image. Using an out-of-the-box implementation of Mask R-CNN, instance segmentation is applied to images of metal powder particles produced through gas atomization. Leveraging transfer learning allows for the analysis to be conducted with a very small training set of labeled images. As well as providing another method for measuring the particle size distribution, we demonstrate the first direct measurements of the satellite content in powder samples. After analyzing the results for the labeled data dataset, the trained model was used to generate measurements for a much larger set of unlabeled images. The resulting particle size measurements showed reasonable agreement with laser scattering measurements. The satellite measurements were self-consistent and showed good agreement with the expected trends for different samples. Finally, we present a small case study showing how instance segmentation can be used to measure spheroidite content in the UltraHigh Carbon Steel DataBase, demonstrating the flexibility of the technique.

36 MATERIALS SCIENCE↗

Subtleties in the trainability of quantum machine learning models

A new paradigm for data science has emerged, with quantum data, quantum models, and quantum computational devices. This field, called quantum machine learning (QML), aims to achieve a speedup over traditional machine learning for data analysis. However, its success usually hinges on efficiently training the parameters in quantum neural networks, and the field of QML is still lacking theoretical scaling results for their trainability. Some trainability results have been proven for a closely related field called variational quantum algorithms (VQAs). While both fields involve training a parametrized quantum circuit, there are crucial differences that make the results for one setting not readily applicable to the other. In this work, we bridge the two frameworks and show that gradient scaling results for VQAs can also be applied to study the gradient scaling of QML models. Our results indicate that features deemed detrimental for VQA trainability can also lead to issues such as barren plateaus in QML. Consequently, our work has implications for several QML proposals in the literature. In addition, we provide theoretical and numerical evidence that QML models exhibit further trainability issues not present in VQAs, arising from the use of a training dataset. We refer to these as dataset-induced barren plateaus. These results are most relevant when dealing with classical data, as here the choice of embedding scheme (i.e., the map between classical data and quantum states) can greatly affect the gradient scaling.

97 MATHEMATICS AND COMPUTING↗

Advancing Building Energy Modeling with Large Language Models: Exploration and Case Studies

The rapid progression in artificial intelligence has facilitated the emergence of large language models like ChatGPT, offering potential applications extending into specialized engineering modeling, especially physics-based building energy modeling. This paper investigates the innovative integration of large language models with building energy modeling software, focusing specifically on the fusion of ChatGPT with EnergyPlus. A literature review is first conducted to reveal a growing trend of incorporating large language models in engineering modeling, albeit limited research on their application in building energy modeling. We underscore the potential of large language models in addressing building energy modeling challenges and outline potential applications including simulation input generation, simulation output analysis and visualization, conducting error analysis, co-simulation, simulation knowledge extraction and training, and simulation optimization. Three case studies reveal the transformative potential of large language models in automating and optimizing building energy modeling tasks, underscoring the pivotal role of artificial intelligence in advancing sustainable building practices and energy efficiency. The case studies demonstrate that selecting the right large language model techniques is essential to enhance performance and reduce engineering efforts. The findings advocate a multidisciplinary approach in future artificial intelligence research, with implications extending beyond building energy modeling to other specialized engineering modeling.

building energy modeling↗

2D Dose Reconstruction by Artificial Neural Network for Pretreatment Verification of IMRT Fields

The use of intensity-modulated radiation therapy (IMRT) is developing rapidly in clinical routines. Because of the high complexity and uniqueness of IMRT treatment plans, patient-specific pretreatment quality assurance is generally considered a necessary prerequisite for patient treatment. In this work, we proposed a modified methodology of electronic portal imaging device (EPID)–based dose validation for pretreatment verification of IMRT fields by applying artificial neural networks (ANNs). The ANN must be trained and validated before use for pretreatment dose verification. For this purpose, 20 EPID fluence maps of IMRT prostate anterior-posterior fields were used as an input for ANN (feed forward type) and a dose map of those fluence maps that were predicted by treatment planning system as an output for ANN. After the training and validation of the neural network, the analysis of 10 IMRT prostate anterior-posterior fields showed excellent agreement between ANN output and dose map predicted by the treatment planning system. The average overall fields pass rate was 96.0% ± 0.1% with 3 mm/3% criteria. The results indicated that the ANN can be used as a low-cost, fast, and powerful tool for pretreatment dose verification, based on an EPID fluence map.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Efficient prediction of attosecond two-colour pulses from an X-ray free-electron laser with machine learning

Abstract X-ray free-electron lasers are sources of coherent, high-intensity X-rays with numerous applications in ultra-fast measurements and dynamic structural imaging. Due to the stochastic nature of the self-amplified spontaneous emission process and the difficulty in controlling injection of electrons, output pulses exhibit significant noise and limited temporal coherence. Standard measurement techniques used for characterizing two-coloured X-ray pulses are challenging, as they are either invasive or diagnostically expensive. In this work, we employ machine learning methods such as neural networks and decision trees to predict the central photon energies of pairs of attosecond fundamental and second harmonic pulses using parameters that are easily recorded at the high-repetition rate of a single shot. Using real experimental data, we apply a detailed feature analysis on the input parameters while optimizing the training time of the machine learning methods. Our predictive models are able to make predictions of central photon energy for one of the pulses without measuring the other pulse, thereby leveraging the use of the spectrometer without having to extend its detection window. We anticipate applications in X-ray spectroscopy using XFELs, such as in time-resolved X-ray absorption and photoemission spectroscopy, where improved measurement of input spectra will lead to better experimental outcomes.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Galaxy morphological classification in deep-wide surveys via unsupervised machine learning

ABSTRACT Galaxy morphology is a fundamental quantity, which is essential not only for the full spectrum of galaxy-evolution studies, but also for a plethora of science in observational cosmology (e.g. as a prior for photometric-redshift measurements and as contextual data for transient light-curve classifications). While a rich literature exists on morphological-classification techniques, the unprecedented data volumes, coupled, in some cases, with the short cadences of forthcoming ‘Big-Data’ surveys (e.g. from the LSST), present novel challenges for this field. Large data volumes make such data sets intractable for visual inspection (even via massively distributed platforms like Galaxy Zoo), while short cadences make it difficult to employ techniques like supervised machine learning, since it may be impractical to repeatedly produce training sets on short time-scales. Unsupervised machine learning, which does not require training sets, is ideally suited to the morphological analysis of new and forthcoming surveys. Here, we employ an algorithm that performs clustering of graph representations, in order to group image patches with similar visual properties and objects constructed from those patches, like galaxies. We implement the algorithm on the Hyper-Suprime-Cam Subaru-Strategic-Program Ultra-Deep survey, to autonomously reduce the galaxy population to a small number (160) of ‘morphological clusters’, populated by galaxies with similar morphologies, which are then benchmarked using visual inspection. The morphological classifications (which we release publicly) exhibit a high level of purity, and reproduce known trends in key galaxy properties as a function of morphological type at z < 1 (e.g. stellar-mass functions, rest-frame colours, and the position of galaxies on the star-formation main sequence). Our study demonstrates the power of unsupervised machine learning in performing accurate morphological analysis, which will become indispensable in this new era of deep-wide surveys.

Martin, G.↗

Disorder Identification in Hysteresis Data: Recognition Analysis of the Random-Bond-Random-Field Ising Model

An approach for the direct identification of disorder type and strength in physical systems based on recognition analysis of hysteresis loop shape is developed. A large number of theoretical examples uniformly distributed in the parameter space of the system is generated and is decorrelated using principal component analysis (PCA). The PCA components are used to train a feed-forward neural network using the model parameters as targets. The trained network is used to analyze hysteresis loops for the investigated system. The approach is demonstrated using a 2D random-bond-random-field Ising model, and polarization switching in polycrystalline ferroelectric capacitors.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Towards Low-Overhead Resilience for Data Parallel Deep Learning

Data parallel techniques have been widely adopted both in academia and industry as a tool to enable scalable training of deep learning models. At scale, DL training jobs can fail due to software or hardware bugs, may need to be preempted or terminated due to unexpected events, or may perform suboptimally because they were misconfigured. Under such circumstances, there is a need to recover and/or reconfigure data-parallel DL training jobs on-the-fly, while minimizing the impact on the accuracy of the DNN model and the runtime overhead. In this regard, state-of-art techniques adopted by the HPC community mostly rely on checkpoint-restart, which inevitably leads to loss of progress, thus increasing the runtime overhead. In this paper we explore alternative techniques that exploit the properties of modern deep learning frameworks (overlapping of gradient averaging and weight updates with local gradient computations through pipeline parallelism) to reduce the overhead of resilience/elasticity. To this end we introduce a failure simulation framework and two resilience strategies (immediate mini-batch rollback and lossy forward recovery), which we study compared with checkpoint-restart approaches in a variety of settings in order to understand the trade-offs between the accuracy loss of the DNN model and the runtime overhead.

data-parallel training↗