Search NASA⌕ Search

SEARCH · Search NASA

Results for “Linear Regression”

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 829 records · Page 46

Radiological results for samples collected on paired glass- and cellulose-fiber filters at the Sandia complex, Tonopah Test Range, Nevada

Airborne particulates are collected at U.S. Department of Energy sites that exhibit radiological contamination on the soil surface to help assess the potential for wind to transport radionuclides from the contamination sites. Collecting these samples was originally accomplished by drawing air through a cellulose-fiber filter. These filters were replaced with glass-fiber filters in March 2011. Airborne particulates were collected side by side on the two filter materials between May 2013 and May 2014. Comparisons of the sample mass and the radioactivity determinations for the side-by-side samples were undertaken to determine if the change in the filter medium produced significant results. The differences in the results obtained using the two filter types were assessed visually by evaluating the time series and correlation plots and statistically by conducting a nonparametric matched-pair sign test. Generally, the glass-fiber filters collect larger samples of particulates and produce higher radioactivity values for the gross alpha, gross beta, and gamma spectroscopy analyses. However, the correlation between the radioanalytical results for the glass-fiber filters and the cellulose-fiber filters was not strong enough to generate a linear regression function to estimate the glass-fiber filter sample results from the cellulose-fiber filter sample results.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Comparison of Two Surface Contamination Sampling Techniques Conducted for the Characterization of Two Pajarito Site Manhattan Project National Historic Park Properties

Technical Area-18 (TA-18), also known as Pajarito Site, is located on Los Alamos National Laboratory property and has historic buildings that will be included in the Manhattan Project National Historic Park. Characterization studies of metal contamination were needed in two of the four buildings that are on the historic registry in this area, a “battleship” bunker building (TA-18-0002) and the Pond cabin (TA-18-0029). However, these two buildings have been exposed to the elements, are decades old, and have porous and rough surfaces (wood and concrete). Due to these conditions, it was questioned whether standard wipe sampling would be adequate to detect surface dust metal contamination in these buildings. Thus, micro-vacuum and surface wet wipe sampling techniques were performed side-by-side at both buildings and results were compared statistically. A two-tail paired t-test revealed that the micro-vacuum and wet wipe techniques were statistically different for both buildings. Further mathematical analysis revealed that the wet wipe technique picked up more metals from the surface than the microvacuum technique. Wet wipes revealed concentrations of beryllium and lead above internal housekeeping limits; however, using an yttrium normalization method with linear regression analysis between beryllium and yttrium revealed a correlation indicating that the beryllium levels were likely due to background and not operational contamination. PPE and administrative controls were implemented for National Park Service (NPS) and Department of Energy (DOE) tours as a result of this study. Overall, this study indicates that the micro-vacuum technique may not be an efficient technique to sample for metal dust contamination.

47 OTHER INSTRUMENTATION↗

Optimizing the Performance of Radionuclide Identification Software in the Hunt for Nuclear Security Threats

The Radionuclide Analysis Kit (RNAK), my team’s most recent nuclide identification software, is entering the testing phase. A question arises: will removing rare nuclides from the software’s library improve its overall performance? An affirmative response indicates fundamental errors in the software’s framework, while a negative response confirms the effectiveness of the software’s key machine learning algorithms. After thorough testing, I found that the performance of RNAK cannot be improved with the library choice effect, thus verifying the effectiveness of RNAK’s algorithms—multiple linear regression, Bayesian network using the Viterbi algorithm, and branch and bound search.

97 MATHEMATICS AND COMPUTING↗

Quantitative analysis of concrete using portable x-ray fluorescence: Method development and validation

During Decommissioning and Demolition (D&D) activities at SRS, it is important that the building be screened for radionuclides and heavy metals to ensure that the proper safety and disposal metrics are in place. A major source of contamination at DOE facilities is the accumulation of mercury contamination, from nuclear material processing and Liquid Waste System (LWS). This buildup of mercury could possibly cause harm to any demolition crew or the environment should this material be released. The current standard method is to take core samples in various places in the facility and use X-ray fluorescence (XRF) to detect the contamination. This standard method comes with a high financial value due to the security levels of these sample facilities with unknown contamination levels. Here in we propose the use of portable XRF units to detect for this contamination on-site. To validate this method, the instrument has to be calibrated to detect the heavy metal contamination, be both precise with the known elemental concentrations and consistent with its actual results of a sample concrete and pristine contaminant, and be able to detect changes in the sample concrete’s composition. After receiving the various concrete samples with their compositions found by a XRF wave-dispersive method, the calibration factor’s linear regressions were adjusted to give the baseline concentration of the concrete with no contamination. Samples of both concrete and concrete/flyash were evaluated; their standard deviations revealed that the measurements were consistent with the known composition. Finally, the samples were contaminated with different concentrations of sodium tungsten dihydrate, allowed to air dry, and measured. When the contaminated samples were analyzed, the heavy metal contamination was seen within the spectrum of the instrument, but there was not a trend of quantification based on the concentration of the solution.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Overhead longwave infrared hyperspectral material identification using radiometric models

Material detection algorithms used in hyperspectral data processing are computationally efficient but can produce relatively high numbers of false positives. Material identification performed as a secondary processing step on detected pixels can help separate true and false positives. This paper presents a material identification processing chain for longwave infrared hyperspectral data of solid materials collected from airborne platforms. The algorithms utilize unwhitened radiance data and an iterative algorithm that determines the temperature, humidity, and ozone of the atmospheric profile. Pixel unmixing is done using constrained linear regression and Bayesian Information Criteria for model selection. The resulting product includes an optimal atmospheric profile and full radiance material model that includes material temperature, abundance values, and several fit statistics. A logistic regression method utilizing all model parameters to improve identification is also presented. This paper details the processing chain and provides justification for the algorithms used. Several examples are provided using modeled data at different noise levels.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Distribution System Secondary Circuit Parameter Estimation for Model Calibration

To analyze and coordinate the operation of distribution systems with rapidly increasing amounts of PV, more accurate distribution system models are required, especially for the distribution system secondary (low-voltage) circuits down to the point of common coupling for distributed PV. There is a growing need for automated procedures to calibrate the distribution system secondary circuit models that are typically either not modeled at all or are modeled with a lower level of detail than the better modeled medium-voltage systems. This report presents an accurate, flexible, and computationally efficient method to use measurement data to estimate secondary circuit series impedance parameters in existing utility feeder models. The parameter estimation method assumes well-modeled primary circuit models, known secondary circuit topologies, and AMI active power, and reactive power measurements at all the loads in the secondary circuit. The method also requires AMI voltage measurement at most of the loads in the secondary circuit but can handle loads that do not have voltage measurements. No existing secondary circuit model information is needed, except for topology. The method is based on the well-known linearized voltage drop approximation and linear regression. The performance of the method is demonstrated on a three-phase test circuit with ten different secondary circuit topologies and on the Georgia Tech campus distribution system with AMI data. The developed method can be utilized to improve existing utility feeder models for more accurate analysis and operation with ubiquitous distributed PV interconnected on the low-voltage circuits.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Advanced Models for Nondestructive Evaluation of Aging Nuclear Power Plant Cables

Background: Nuclear power plant (NPP) cables must function correctly for proper power supply and control of the nuclear reactor. This is especially important for public and environmental safety at critical times such as during reactor shut-down following loss of coolant. Over extended periods of service, the cable insulation and jacket polymer materials suffer from degradation due to exposure to environmental stressors such as heat and ionizing radiation. The focus of this project has been the development of detailed descriptions of the way in which insulation polymers degrade, from a materials science perspective, and the development of new electrical techniques to monitor cable polymer condition nondestructively. Objectives: The objectives of this project have been to i) perform detailed characterization experiments upon cable insulation polymers that have experienced thermal and radiation exposure mimicking that received during NPP operation, including the exploration of novel cable nondestructive evaluation methods; capacitive, THz and infrared methods; ii) identify the most sensitive indicators of aging in these materials, and iii) develop advanced, validated models describing microstructural and chemical changes that are observed in cable insulation polymers due to thermal and radiation exposure. Methods: Accelerated aging was performed on the two most common United States NPP cable insulation polymers (cross-linked polyethylene (XLPE) and ethylene propylene rubber (EPR)) under various conditions of elevated temperature and gamma radiation dose. Aging mechanisms were elucidated through extensive thermal, mechanical and microscopic characterization of the aged samples, and materials modeling. Dielectric and IR spectra were measured on pristine and aged samples. Data was analyzed statistically, via multiple linear regression, partial least squares and probabilistic neural network approaches. Impact: Extant cable monitoring techniques do not deal successfully with all polymer types. This research has laid a foundation upon which new electrical nondestructive test methods for cable polymer condition monitoring may be developed, hence launching the development of new test methods for cable insulation monitoring based on measuring electrical rather than mechanical parameters. Collaboration: This project has benefitted from collaboration between experts in materials science and engineering, materials modeling, electrical engineering, and nondestructive evaluation. The majority of the characterization measurements were carried out at Iowa State University. Materials modeling was carried out primarily at Washington State University. Irradiation of samples took place at Pacific Northwest National Laboratory (PNNL), where a student intern from Iowa State was hosted for one summer, setting up the aging experiments. A research student from the University of Bologna, Italy, joined the project for their thesis research at Iowa State University with samples provided by PNNL. The project benefitted from regular involvement of senior engineers in US-based industry partners. Project investigators presented findings at various national and international conferences and workshops.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

An Evaluation of Bang-Time Measurements from a Multichannel, Triaxial, nTOF Diagnostic for MagLIF Experiments at the Z facility

Neutron bang times for a series of MagLIF (Magnetic Liner Inertial Fusion) experiments with D 2 -filled targets have been measured at the Z facility. The emitted neutrons were detected as current-mode pulses in a multichannel, neutron time-of-flight (nTOF) diagnostic with conventional, scintillator-photomultiplier-tube (PMT) detectors. In these experiments, the detectors were fielded at known, fixed distances L (690-2510 cm) from the target, and on three, non-coplanar (but convergent) lines-of-sight (LOS). The primary goal of this diagnostic was to estimate a fiducial time (bang time) relative to an externally generated time-base for synchronizing all the diagnostics in an experiment. Recorded arrival times (A7) of the pulses were characterized experimentally by three numerical methods: a first-moment estimate (centroid) and two nodal measures — Savitzky-Golay (SG) smoothing and a single point peak estimate of the raw data. These times were corrected for internal detector time delays (transit and impulse-response function) — an adjustment that linked the recorded ATs to the corresponding arrival of uncollided neutrons at each detector. The bang time was then estimated by linearly regressing the arrival times against the associated distances to the source; t bang (on the system timescale) was taken as the temporal intercept of the regression equation at distance L = 0. This article reports the analysis for a representative shot #2584 for which (a) the recorded ATs — even without detector corrections — agreed by method in each channel to within 1-2 ns; (b) internal corrections were each ~3 — 5 ns; and (c) a 95% uncertainty (confidence) interval for t bang in this shot was estimated at ±3 ns with 4 degrees of freedom. A secondary goal for this diagnostic was to check that the bang time measurements corresponded to neutrons emitted by the D(d,n) 3 He reaction in a thermalized DD plasma. According to the theoretical studies by Brysk, such neutrons should be emitted with an isotropic Gaussian distribution of mean kinetic energy $ \overline{E}$ of 2.449 MeV; this energy translates to a mean neutron speed $ \overline{u}$ of 2.160 cm/ns [D. H. Munro, Nuclear Fusion, 56(3) 036001 (2016)]. In the MagLIF series of shots there was no evidence of spatial asymmetry in the time-distance regressions, and it was possible to extract the mean neutron speed from the slope of these fits. In shot 2584 $ \overline{u}$ was estimated at 2.152 cm/ns ± 0.010 cm/ns [95 % confidence, 4 dof] and the mean kinetic energy $ \overline{E}$ (with relativistic corrections) was 2.431 MeV ± 0.022 MeV [95 % confidence, 4 dof] — results supporting the assumption that D-D neutrons were, in fact, measured.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Quantification of MagLIF Morphology using the Mallat Scattering Transformation

The morphology of the stagnated plasma resulting from Magnetized Liner Inertial Fusion (MagLIF) is measured by imaging the self-emission x-rays coming from the multi-keV plasma, and the evolution of the imploding liner is measured by radiographs. Equivalent diagnostic response can be derived from integrated rad-MHD simulations from programs such as Hydra and Gorgon. There have been only limited quantitative ways to compare the image morphology, that is the texture, of simulations and experiments. We have developed a metric of image morphology based on the Mallat Scattering Transformation (MST), a transformation that has proved to be effective at distinguishing textures, sounds, and written characters. This metric has demonstrated excellent performance in classifying ensembles of synthetic stagnation images. We use this metric to quantitatively compare simulations to experimental images, cross experimental images, and to estimate the parameters of the images with uncertainty via a linear regression of the synthetic images to the parameter used to generate them. This coordinate space has proved very adept at doing a sophisticated relative back-ground subtraction in the MST space. This was needed to compare the experimental self emission images to the rad-MHD simulation images. We have also developed theory that connects the transformation to the causal dynamics of physical systems. This has been done from the classical kinetic perspective and from the field theory perspective, where the MST is the generalized Green's function, or S-matrix of the field theory in the scale basis. From both perspectives the first order MST is the current state of the system, and the second order MST are the transition rates from one state to another. An efficient, GPU accelerated, Python implementation of the MST was developed. Future applications are discussed.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Detector Response Function and Directional Gamma-Ray Source Calculations for Polaris

A Directional Unfolded Source Term (DUST) method was developed to compute directionally resolved gamma-ray source terms based on back-projection spectra synthesized by Compton Cameras. Spectral features in the unprocessed spectra are indistinct primarily because the rotational angles for the conical projections cannot be determined, so probability distributions are constructed from overlapping cones. The DUST method uses an angular response function to compute a covariance matrix, which is used to process count rates in back-projection spectra by linear regression to partition the gamma-rays among several spatial regions. This method was applied to analyze data collected by the Polaris detector during an evaluation that was conducted at Oak Ridge National Laboratory (ORNL). The evaluation includes measurements of calibration sources with angular separations ranging from 1° to more than 50°. Measurements were also performed for cylindrical depleted uranium castings and a 137 Cs source inside a large polyethylene sphere. The DUST algorithm was able to differentiate gamma-rays emitted by 137 Cs and 60 Co when the sources were separated by less than 2°, but separation greater than 10° was required to isolate the 133 Ba emission from gamma-rays emitted by the other sources. The computed source terms were consistent with emission profiles from the calibration sources and from models of the spatially-extended sources. Methods for viewing radiation profiles were also evaluated because user input is required to select spatial regions of interest.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Verification of Triso Fuel Burnup Using Machine Learning Algorithms

Pebble Bed Reactors are fueled with fuel pebbles that are circulated multiple times through the reactor vessel before discharge. During the normal operation of a PBR, ejected pebbles are returned to the reactor or discharged depending on the fuel burnup and physical condition of the pebbles. The burnup measurement is usually based on detected radiation signatures of fission products accumulated in the pebble fuel over burnup. Previous research has shown that height of photopeaks of fission products, such as 134 Cs, 137 Cs, 154 Eu, etc., can be used independently or in combination to infer or predict the level of burnup in the fuel. However, it remains challenging to measure such complex sources due to self-shielding effects, strong radiation background and intervening materials. Another operational challenge is the required high throughput of burnup measurement, which necessitates limited measurement time and thus impacts quality of measured gamma-ray spectra. Hence, advanced spectral analysis methods are needed to analyze the noisy gamma spectra and predict the burnup values. We propose to use machine learning (ML) method to interpret gamma-ray spectra and predict the burnup values of the pebbles. ML has achieved widespread success and adoption across a few domains that require pattern recognition and analysis in varied data types. In this work, we apply three proven ML approaches - multilayer perceptrons, convolutional neural networks, and transformers - to the task of predicting fuel burnup from measured gamma spectra, and compile a dataset of simulated spectra for training and validation of the ML models. In this paper, we will discuss the network architecture of these three ML approaches and compare the performance of the simplest of these (MLP) to a standard linear regression.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Using machine learning to improve efficiency and accuracy of burnup measurements at PBR reactors [Slides]

The outline of the slides include: Motivations of the work; Modeling and simulation; Machine learning model; Results and comparison study with linear regression; and Conclusions. This work was done to help PBR designers and operators understand the burnup measurement better. We look forward to discussing the results in detail with industrial collaborators.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Physics-Infused AI/ML Based Digital-Twin Framework for Flow-Induced-Vibration Damage Prediction in a Nuclear Reactor Heat Exchanger

This report summarizes some of the ongoing work related to the development of an expert-elicitation-digital-twin framework for real time damage state prediction in heat exchanger components of a nuclear reactor. The framework is targeted towards predicting damage associated with coupled low cycle fatigue (associated with regular heat-up, cool-down and power operation transients) and high cycle fatigue (associated with flow induced vibration transients). The overall framework will be based on a NoSQL based database, physics-infused-geometry-dependent virtual-sensor data, different AI/ML techniques-based data-driven-predictive-model applications (Apps) and real-time plant sensor measurements available through few existing sensors. Towards this overall goal, this report updates some of the ongoing work, such as on implementation of a NoSQL Database (such as MongoDB), FE based heat transfer analysis of a heat exchanger (e.g. of a PWR steam generator) for generating geometry-dependent virtual sensor data and evaluation of various AI/ML models such as based on multivariate linear regression, ensembled decision-tree based Random-Forest and Gradient-Boosting regression and high-dimensional-kernel-function-transformation based Support-Vector-Machine regression models. The AI/ML models were evaluated for predicting multi-time-series thermal states at thousands of 3D point-clouds

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Adapting Secure MultiParty Computation to Support Machine Learning in Radio Frequency Sensor Networks

In this project we developed and validated algorithms for privacy-preserving linear regression using a new variant of Secure Multiparty Computation (MPC) we call "Hybrid MPC" (hMPC). Our variant is intended to support low-power, unreliable networks of sensors with low-communication, fault-tolerant algorithms. In hMPC we do not share training data, even via secret sharing. Thus, agents are responsible for protecting their own local data. Only the machine learning (ML) model is protected with information-theoretic security guarantees against honest-but-curious agents. There are three primary advantages to this approach: (1) after setup, hMPC supports a communication-efficient matrix multiplication primitive, (2) organizations prevented by policy or technology from sharing any of their data can participate as agents in hMPC, and (3) large numbers of low-power agents can participate in hMPC. We have also created an open-source software library named "Cicada" to support hMPC applications with fault-tolerance. The fault-tolerance is important in our applications because the agents are vulnerable to failure or capture. We have demonstrated this capability at Sandia's Autonomy New Mexico laboratory through a simple machine-learning exercise with Raspberry Pi devices capturing and classifying images while flying on four drones.

42 ENGINEERING↗

Decay Curve Correction Analysis Report

The decay curve analysis that is done on the short-lived radionuclide gas samples is used to differentiate between gaseous radionuclides that have the same characteristic gamma decay energy, 511 kiloelectron-volts (keV). A sample of stack gas is isolated and the total counts in the 511 keV peak are counted repeatedly in 10-second intervals to evaluate the decay rate of the sample over time. Analysis of this decay data required a series of steps. First, a raw data report is generated by the gamma acquisition system, based on an analysis template within the acquisition software. The data report file was then loaded into Microsoft Word, and a macro was used to perform minor formatting (remove colons and insert tabs between data columns) to allow analysis within Excel. The file is then saved as a text file at this point. The text file is then uploaded into Excel and a series of macros are used to add labels, calculate radioactive decay constants, and analyze the gamma decay data using linear regression techniques. The analysis template has been used since 1998 for stack 53000303 (TA-53, building 0003, exhaust stack 03) and 2000 for stack 53000702. The overall process, including the gamma report format and the macros used in Word and Excel for processing the report, had remained unchanged until 2015. In October of 2015, staff made a change to the report template in the gamma acquisition software which resulted in an error in the calculations later performed by the Excel macro. This error was not caught until a more in-depth review of the analysis took place regarding 2020 data. This report covers a much more complete review of the issue that occurred regarding the decay curve analysis, a review of the calculations completed to correct the issue, a review of the updated decay curve analysis process, and recommendations for moving forward.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Determination of the uranium content of storage containers filled with waste from Fukushima using cosmic ray data

This report presents the results of an experimental study of cosmic ray tomography aimed at determining the uranium contents of waste containers filled with waste from Fukushima. Data, taken using the Giant Muon Tracker, were obtained on a set of scenes constructed from blocks of polyethylene, concrete, iron, and lead. The data have been provided to Toshiba and have been used to evaluate the composition of scenes using multivariate linear regression.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Interpretable Net Load Forecasting Using Smooth Multiperiodic Features

We consider the problem of forecasting net load over a horizon such as one day, using a trailing window of past net load values as well as date and time. We focus on three variations on this problem: point forecasts, marginal quantile forecasts, and generating conditional samples of the future value. We propose a method that relies on linear regression using some custom engineered time-based features to capture multiple periodicities, such as daily, weekly, and seasonal, and their interactions. Our proposed models are readily interpretable, and rely on efficient and reliable convex optimization [1] to fit. We illustrate our method on four years worth of hourly net load data, comparing predictions made with various subsets of the features.

Ogut, Mehmet G↗

Protecting and Defending against Autonomous Control Systems and Digital Twin Cyber Attacks: Response Strategy for Hyperparameter attacks of Digital Twin Machine Learning Models in Nuclear Power Plants (Final)

Navigating through the complex tapestry of technological advancements, "Response Strategy for Hyperparameter attacks of Digital Twin Machine Learning Model in Nuclear Power Plants" stands at the intersection of cybersecurity and nuclear power plant operations, embarking on a journey through the intricacies of securing digital twins against malicious cyber activities. As nuclear power plants progressively integrate digital twin technology and machine learning models to optimize operations and ensure system reliability, they inadvertently expose themselves to a new spectrum of vulnerabilities, notably in the realm of hyperparameter attacks. Hyperparameters, integral in machine learning model tuning and optimal performance of digital twins, have emerged as a target for adversaries aiming to destabilize the predictive capabilities and therefore, the operational accuracy of these digital entities within critical infrastructures like nuclear plants. This paper, therefore, meticulously threads the needle through the development of a robust response strategy, poised to shield these digital reflections against calculated hyperparameter manipulations, ensuring that the digital twin can effectively and securely function as a reliable proxy for its physical counterpart. The ensuing sections delve into the orchestrated maelstrom of multi-rate time-changing intelligent coordinated hyperparameter attacks and the implementation of event-triggered predictive control, laying down a structured, predictive, and responsive framework that safeguards the nexus where the digital and physical realms of nuclear power plants coalesce. The operational integrity of digital twins in nuclear power plants depends critically on the security of machine learning hyperparameters. This study makes two different contributions. First, a decision-based idea known as a multi-rate time changing intelligent coordinated hyperparameter attack is put forth. In this attack, many hyperparameters are repeatedly changed using both random and intelligent optimal techniques by the attacker. These assaults introduce varied rates at different attack steps, compromise various amounts of hyperparameters, and improve stealth and flexibility. Second, a technique is developed for event triggered predictive control to rapidly respond to potential hyperparameter attacks. This control integrates a sliding window framework, retaining a history of previous data points and employing linear regression to predict the next data point from the current dataset. The control gain K is determined using the Lyapunov-Krasovskii method, and subsequently, an action is developed. Finally, the outcome of the simulation demonstrates the viability of the proposed method for defending nuclear power plant digital twins from hyperparameter attacks.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗