Search NASA⌕ Search

SEARCH · Search NASA

Results for “Variance to Mean”

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 235 records · Page 13
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Hanford Waste Treatment Plant Effluent Management Facility Stack Effluent Monitoring: Sampling Probe Location Qualification Evaluation

The Hanford Tank Waste Treatment and Immobilization Plant Effluent Management Facility (EMF) stack monitor location was qualified using a combination of scale model stacks to mitigate the risk of identifying that the sampling location does not meet the qualification criteria on the full-scale stack. The LV-S1 scale model stack was used as a baseline, augmented by the LB-S1 and LV-S2 scale model stacks to address the Direct Feed Low Activity Waste Effluent Management Facility Vessel Vent Process (DVP) injection into the main Active Confinement Ventilation (ACV) system duct. As required by the American National Standards Institute/Health Physics Society (ANSI/HPS) N13.1-1999 standard, the scale model and its sampling locations were geometrically similar to the actual stack, and the Reynolds numbers for both the actual and model stacks were >10,000. An additional criterion is that the product of the hydraulic diameter and mean velocity (DV) of the full-scale stack must be between 1/6 DV and 6 DV of the scale model stack tests. The LV-S1, LB-S1, and LV-S2 scale model stack tests have met the criteria of the ANSI/HPS N13.1-1999 standard to demonstrate the stack sampling locations are well mixed. Verification tests of the EMF stack were performed at normal operating conditions. The minimum 1/6 DV value and the maximum 6 DV value from the scale model testing determine the range of stack flow rates for which the full-scale stack may be operated while remaining in compliance with the stack verification criterion. A practical range for the full-scale stack qualification uses the average DV through 6 DV from the scale model tests to compute the corresponding flow rates. Table S1 lists the operating flow rate along with the average and maximum qualified stack flow rate based on the LV-S1 scale model DV values. The operating flow is below the maximum qualified stack flow, which means that the scale model test results are acceptable for stack qualification. The remaining criteria for the stack verification to be considered valid involve the flow angle and velocity uniformity results. First, the flow angle at the full-scale stack must be ≤20°. Second, the velocity uniformity at the full-scale stack must be ≤20% coefficient of variance (COV). Finally, the velocity uniformity results for the actual and scale model stack tests must agree within 5% COV. These criteria were met through the full-scale stack test at the EMF. Flow angle results were <5°; all flow angle results were within the ≤20° criterion. The velocity uniformity results for each test condition ranged between 2.2% COV and 4.3% COV, all of which were within the range of the target % COV values from the scale model tests on the LV-S1, LB-S1, and LV S2 scale models. Based on these stack verification test results, the EMF filtered exhaust stack sampling location meets the qualification criteria provided in the ANSI/HPS N13.1-1999 standard for all planned fan operating configurations. This includes each combination of ACV fans with DVP exhausters. Further changes to the system configuration or operating conditions that are outside the qualified flow rates described in this report may require additional tests or analyses to determine compliance with the standard.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2020

Data Collection and Analysis Challenges and Mitigation Strategies for Quantitative Human Factors Research Studies in Nuclear Power Plant Modernization

The United States (U.S.) Department of Energy (DOE) Light Water Reactor Sustainability (LWRS) program Plant Modernization Pathway is conducting targeted research and development (R&D) to address aging and reliability concerns with the legacy instrumentation and control and related information systems of the U.S. LWR fleet. In this effort, the application of human factors engineering (HFE) provides an important role in ensuring new digital plant technologies enable broad innovation and business improvement with continued operational safety. Evaluation is a key activity in HFE, which often occurs iteratively through the system design lifecycle. While qualitative methods are important in collecting information of how users perform tasks through observations, quantitative methods are equally important in assessing system design based on performance. In collecting and analyzing this quantitative data, there are notable challenges in the nuclear HFE domain that may threaten the validity and reliability of the inferences made in these studies. Notable challenges include small sample size and limited resources, large error variance and small effect size, an “adding test features to losing degrees of freedom” dilemma, non-normal distribution, and heterogeneity of variance. The results in control room usability studies are often statistically non-significant, which makes it hard to interpret. This work discusses these challenges across different scientific viewpoints and provides real-world examples of these challenges in practice. Collectively, the objective of this work is to position these challenges to the larger data science community as a means of identifying future opportunities to address these issues.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2019

Relative Importance of Vital Rates to Population Dynamics of Wood Ducks

The wood duck (Aix sponsa) is a common and important cavity-nesting duck in North America; however, we know very little about how changes in vital rates influence population growth rate (λ). We used estimates of fertility and survival of female wood ducks from our nest-box studies in South Carolina, Alabama, and Georgia, USA, to create a stage-based matrix population model. We conducted perturbation analyses and ranked elasticity values to examine the relative importance of 17 component vital rates to λ. Female survival is influenced by nest success, so we recognized this female heterogeneity in our analyses. Four vital rates showed the greatest importance to λ. Analytic elasticities were greatest for breeding season and nonbreeding season survival of females that nested successfully, followed by nest success and female recruitment to the breeding population. Differences in female quality were important to λ. Next, we used process variation of vital rates and conducted life-stage simulation analyses (LSA) followed by variance decomposition to determine the amount of variation in λ explained by each vital rate. Female recruitment to the breeding population explained 57.7% of the variation in λ followed by nest success (11.4%), and breeding and nonbreeding season survival of females that nested successfully (9.3% and 9.4%, respectively). Together these 4 vital rates explained 88% of the variation in λ. Mean asymptotic population growth rate (λ = 0.80 ± 0.08 [SD]) from LSA revealed a declining population. Recruitment of females hatched from nest boxes was insufficient to sustain the nest-box population. Furthermore, including yearling (SY) females that were produced outside of nest boxes (i.e., immigrants) increased recruitment rates 1.5 to 2 times more than when only SY females recruited from nest boxes were included. Future research that examines how emigration and immigration interact with survival and reproduction to influence local population dynamics of wood ducks will be important for identifying the value of nest-box programs to wood duck conservation and management.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2022

Reconstructing cosmic polarization rotation with ResUNet-CMB

Cosmic polarization rotation, which may result from parity-violating new physics or the presence of primordial magnetic fields, converts E-mode polarization of the cosmic microwave background (CMB) into B-mode polarization. Anisotropic cosmic polarization rotation leads to statistical anisotropy in CMB polarization and can be reconstructed with quadratic estimator techniques similar to those designed for gravitational lensing of the CMB. At the sensitivity of upcoming CMB surveys, lensing-induced B-mode polarization will act as a limiting factor in the search for anisotropic cosmic polarization rotation, meaning that an analysis which incorporates some form of delensing will be required to improve constraints on the effect with future surveys. In this study we extend the ResUNet-CMB convolutional neural network to reconstruct anisotropic cosmic polarization rotation in the presence of gravitational lensing and patchy reionization, and we show that the network simultaneously reconstructs all three effects with variance that is lower than that from the standard quadratic estimator nearly matching the performance of an iterative reconstruction method.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2022

Westward-Propagating Moisture Mode Over the Tropical Western Hemisphere

A westward-propagating Rossby-like wave signal is found to explain a large fraction of the intraseasonal variance in cloud brightness over the Western Hemisphere. A series of diagnostic criteria suggest that this wave is a moisture mode: its moisture anomalies dominate the distribution of moist static energy (MSE) and are in phase with the precipitation anomalies; and the thermodynamic equation obeys the weak temperature gradient approximation. Further, the wave propagates westward due to zonal moisture advection by the mean flow and is maintained by radiative heating and meridional moisture advection. These properties compare favorably with the westward propagating Rossby mode in an equatorial beta-plane model with prognostic moisture, mean meridional moisture gradient, and mean zonal wind. These results underscore the importance of water vapor in the dynamics of slowly evolving tropical systems, and the limitations of dry shallow water theory that rely on a “reduced equivalent depth” to represent moist dynamics.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Inferring Plant Acclimation and Improving Model Generalizability With Differentiable Physics‐Informed Machine Learning of Photosynthesis

Net photosynthesis (A N ) is a key component of the global carbon cycle influencing climate feedback over decadal scales. Although plant acclimation to environmental changes can modify A N , traditional vegetation models in Earth system models (ESMs) often rely on plant functional type (PFT)-specific parameterizations or simplified acclimation assumptions limiting generalizability across time, space, and PFTs. In this study, we developed a differentiable photosynthesis model to learn the environmental dependencies of V c,max25 (maximum carboxylation rate at 25°C, representing photosynthetic capacity), as this genre of hybrid physics-informed machine learning can seamlessly train neural networks and process-based equations together. Compared to PFT-specific parameterization of V c,max25 , learning the environment dependencies of key photosynthetic parameters improved model spatiotemporal generalizability. Applying environmental acclimation to V c,max25 led to substantial variations in global mean A N indicating the need to address acclimation in ESMs. The model effectively captured multivariate observations (V c,max25 , A N , and stomatal conductance (g s )) simultaneously with multivariate constraints, improving generalization across space and PFTs. It also learned sensible acclimation relationships of V c,max25 to different environmental conditions. The model explained more than 54%, 57%, and 62% of the variance of A N , g s , and V c,max25 , respectively, presenting a first global-scale spatial test benchmark of A N and g s . These results highlight the potential for differentiable modeling to enhance process-based modules in ESMs and effectively leverage information from large, multivariate data sets.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Evaluation of idealized large-eddy simulations performed with the Weather Research and Forecasting model using turbulence measurements from a 250 m meteorological mast

Abstract. We investigate the ability of the Weather Research and Forecasting model to perform large-eddy simulation of canonical flows. This is achieved through comparison of the simulation outputs with measurements from sonic anemometers on a 250 m meteorological mast located at Østerild, in northern Denmark. Østerild is on a flat and rough area, and for the predominant wind directions, the atmospheric flow can be considered to be close to homogeneous. The idealized simulated flows aim at representing atmospheric boundary layer turbulence under unstable, neutral, and stable stability conditions at the surface, which are statistically significant conditions observed at Østerild. We found that the resolved fields from the simulations appear to have the characteristics of the three stability regimes. Vertical profiles of observed mean wind speeds and direction are well reproduced by the simulations, with the largest differences under near-neutral conditions, where the effect of the subgrid-scale model is evident on the vertical wind shear close to the surface. Vertical profiles of observed eddy fluxes are also well reproduced by the simulations, with the largest differences for the three velocity component variances under stable stability conditions, although nearly always within the observed variability. With regards to turbulent kinetic energy, we find good agreement between observations and simulations at all vertical levels. Simulated and observed velocity spectra match very well and show very similar behavior with height and with atmospheric stability within the low-frequency interval; at the effective resolution, the simulated spectra show the typical drop-off of finite differences. Our findings demonstrate that these idealized simulations reproduce the characteristics of atmospheric stability regimes often observed at a high turbulent and flat site within a direction sector, where the air flows over nearly homogeneous land.

17 WIND ENERGY↗
DOE OSTI2022

Model study of the energy dependence of the correlation between anisotropic flow and the mean transverse momentum in Au + Au collisions

A hybrid model that employs the hadron-string transport model UrQMD and the (3+1)-dimensional relativistic viscous hydrodynamic code vHLLE, is used to investigate the beam energy dependence of the correlation coefficient ρ(v$^{2}_{2}$,[p T ]) between the average transverse momentum [p T ] of hadrons emitted in an event and the square of the anisotropic flow coefficient v$^{2}_{2}$. For Au+Au collisions, the model predicts characteristic patterns for the energy and event-shape dependence of the variances for [p T ] and v$^{2}_{n}$ [Var([p T ] and Var(v$^{2}_{2}$)], and the covariance of v$^{2}_{n}$ and [p T ][cov(v$^{2}_{2}$,[p T ])], consistent with the attenuation effects of the specific shear viscosity η/s. Here in contrast, ρ(v$^{2}_{2}$,[p T ]) is predicted to be insensitive to the beam energy but sensitive to the initial-state geometry of the collisions. These observations suggest that a precise set of measurements for Var([p T ]), Var(v$^{2}_{2}$), cov(v$^{2}_{2}$,[p T ]) and ρ(v$^{2}_{2}$,[p T ]) as a function of beam energy and event shape, could serve to constrain better the eccentricities and their fluctuations over the entire span of beam energies, and consequently aid precision extraction of the temperature and baryon chemical-potential dependence of η/s from the wealth of Au+Au data obtained in the RHIC beam energy scan.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2024

Stability-Constrained Learning for Frequency Regulation in Power Grids With Variable Inertia

The increasing penetration of converter-based renewable generation has resulted in faster frequency dynamics, and low and variable inertia. As a result, there is a need for frequency control methods that are able to stabilize a disturbance in the power system at timescales comparable to the fast converter dynamics. This paper proposes a combined linear and neural network controller for inverter-based primary frequency control that is stable at time-varying levels of inertia. We model the time-variance in inertia via a switched affine hybrid system model. We derive stability certificates for the proposed controller via a quadratic candidate Lyapunov function. We test the proposed control on a 12-bus 3-area test network, and compare its performance with a base case linear controller, optimized linear controller, and finite-horizon Linear Quadratic Regulator (LQR). Our proposed controller achieves faster mean settling time and over 50% reduction in average control cost across 100 inertia scenarios compared to the optimized linear controller. Unlike LQR which requires complete knowledge of the inertia trajectories and system dynamics over the entire control time horizon, our proposed controller is real-time tractable, and achieves comparable performance to LQR.

data-driven control↗
DOE OSTI2022

Development of Short-Term Forecasting Models Using Plant Asset Data and Feature Selection

Nuclear power plants collect and store large volumes of heterogeneous data from various components and systems. With recent advances in machine learning (ML) techniques, these data can be leveraged to develop diagnostic and short-term forecasting models to better predict future equipment condition. Maintenance operations can then be planned in advance whenever degraded performance is predicted, thus resulting in fewer unplanned outages and the optimization of maintenance activities. This enables lower maintenance costs and improves the overall economics of nuclear power. This paper focuses on developing a short-term forecasting process that leverages a feature selection process to distill large volumes of heterogeneous data and predict specific equipment parameters. A variety of feature selection methods, including Shapley Additive Explanations (SHAP) and variance inflation factor (VIF), were used to select the optimal features as inputs for three ML methods: long short-term memory (LSTM) networks, support vector regression (SVR), and random forest (RF). Each combination of model and input features was used to predict a pump bearing temperature both 1 and 24 hours in advance, based on actual plant system data. The optimal inputs for the LSTM and SVR were selected using the SHAP values, while the optimal input for the RF consisted solely of the response variable itself. Each model produced similar 1-hour-ahead predictions, with root mean square errors (RMSEs) of roughly 0.006. For the 24-hour-ahead predictions, differences could be seen between LSTM, SVR, and RF, as reflected by model performances of 0.036 +- 0.014, 0.0026 +- 0, and 0.063 +- 0.004 RMSE, respectively. As big data and continuous online monitoring become more widely available, the proposed feature selection process can be used for many applications beyond the prediction of process parameters within nuclear infrastructure.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2025

Introducing a multiscale feature integration network for inpainting with applications to enhanced CMB map reconstruction

We introduce a novel neural network, SkyReconNet, which combines the expanded receptive fields of dilated convolutional layers along with standard convolutions, to capture both the global and local features for reconstructing the missing information in an image. We implement our network to inpaint the masked regions in a full-sky cosmic microwave background (CMB) map. Inpainting CMB maps is a particularly formidable challenge when dealing with extensive and irregular masks, such as galactic masks which can obscure substantial fractions of the sky. The hybrid design of SkyReconNet leverages the strengths of standard and dilated convolutions to accurately predict CMB fluctuations in the masked regions by effectively utilizing the information from surrounding unmasked areas. During training, the network optimizes its weights by minimizing a composite loss function that combines the structural similarity index measure (SSIM) and mean squared error (MSE). SSIM preserves the essential structural features of the CMB, ensuring an accurate and coherent reconstruction of the missing CMB fluctuations, while MSE minimizes the pixelwise deviations, thus enhancing the overall accuracy of the predictions. The predicted CMB maps and their corresponding angular power spectra align closely with the targets, achieving the performance limited only by the fundamental uncertainty of cosmic variance. The network’s generic architecture enables application to other physics-based challenges involving data with missing or defective pixels, systematic artifacts, etc. In conclusion, our results demonstrate its effectiveness in addressing the challenges posed by large irregular masks, offering a significant inpainting tool not only for CMB analyses but also for image-based experiments across disciplines where such data imperfections are prevalent.

Cosmic microwave background↗
DOE OSTI2024

Variance-Reduced Accelerated First-Order Methods: Central Limit Theorems and Confidence Statements

In this paper, we consider a strongly convex stochastic optimization problem and propose three classes of variable sample-size stochastic first-order methods: (i) the standard stochastic gradient descent method, (ii) its accelerated variant, and (iii) the stochastic heavy-ball method. In each scheme, the exact gradients are approximated by averaging across an increasing batch size of sampled gradients. We prove that when the sample size increases at a geometric rate, the generated estimates converge in mean to the optimal solution at an analogous geometric rate for schemes (i)–(iii). Based on this result, we provide central limit statements, whereby it is shown that the rescaled estimation errors converge in distribution to a normal distribution with the associated covariance matrix dependent on the Hessian matrix, the covariance of the gradient noise, and the step length. If the sample size increases at a polynomial rate, we show that the estimation errors decay at a corresponding polynomial rate and establish the associated central limit theorems (CLTs). Under certain conditions, we discuss how both the algorithms and the associated limit theorems may be extended to constrained and nonsmooth regimes. As a result, we provide an avenue to construct confidence regions for the optimal solution based on the established CLTs and test the theoretical findings on a stochastic parameter estimation problem.

Lei, Jinlong↗
DOE OSTI2026

An Analysis of Future Wind Energy Resources and Cost Uncertainties Across the United States

Wind power is a growing source of energy generation that relies on complex global atmospheric and earth system processes. There has been evidence of reductions in average wind speeds over land in North America since the 1980s, and several models project that average wind speeds will continue to decrease. Concurrently, the cost of wind energy systems in the United States has been decreasing since around 2010, a trend also projected to continue. There is considerable uncertainty in these future projections, with quantitative estimates of future wind resource and system costs varying widely. To study this, we run wind energy models with possible future system costs, turbine designs, and meteorological inputs from multiple downscaled earth system models over the contiguous United States. Changes in mean annual energy production from 2000-2019 to 2040-2059 can be as high as +10% in South Texas or as low as -20% in Iowa. Larger turbines and moderate reductions in system costs can offset the largest projected decreases in wind resource, but much uncertainty remains in the extent to which wind resources will change and to what extent system costs can be reduced. An analysis of variance shows, in several states in the Midwest, uncertainty in future wind resources can influence the cost of wind energy nearly as much as uncertainty in future system costs.

17 WIND ENERGY↗
DOE OSTI2021

Temporal deposition of copper and zinc in the sediments of metal removal constructed wetlands

The objective of this study was to explore the effects of time, seasons, and total carbon (TC) on Copper (Cu) and Zinc (Zn) deposition in the surface sediments. This study was performed at the H-02 constructed wetland on the Savannah River Site (Aiken, SC, USA). Covering both warm (April-September) and cool (October-March) seasons, several sediment cores were collected twice a year from the H-02 constructed wetland cells from 2007 to 2013. Total concentrations of Cu and Zn were measured in the sediments. Concentrations of Cu and Zn (mean ± standard deviation) in the surface sediments over 7 years of operation increased from 6.0 ± 2.8 and 14.6 ± 4.5 mg kg -1 to 139.6 ± 87.7 and 279.3 ± 202.9 mg kg -1 dry weight, respectively. The linear regression model explained the behavior and the variability of Cu deposition in the sediments. On the other hand, using the generalized least squares extension with the linear regression model allowed for unequal variance and thus produced a model that explained the variance properly, and as a result, was more successful in explaining the pattern of Zn deposition. Total carbon significantly affected both Cu ( p = 0.047) and Zn ( p < 0.001). Time effect on Cu deposition was statistically significant ( p = 0.013), whereas Zn was significantly affected by the season ( p = 0.009).

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Vegetation type is an important predictor of the arctic summer land surface energy budget

Despite the importance of high-latitude surface energy budgets (SEBs) for land-climate interactions in the rapidly changing Arctic, uncertainties in their prediction persist. Here, we harmonize SEB observations across a network of vegetated and glaciated sites at circumpolar scale (1994–2021). Our variance-partitioning analysis identifies vegetation type as an important predictor for SEB-components during Arctic summer (June-August), compared to other SEB-drivers including climate, latitude and permafrost characteristics. Differences among vegetation types can be of similar magnitude as between vegetation and glacier surfaces and are especially high for summer sensible and latent heat fluxes. The timing of SEB-flux summer-regimes (when daily mean values exceed 0 Wm –2 ) relative to snow-free and -onset dates varies substantially depending on vegetation type, implying vegetation controls on snow-cover and SEB-flux seasonality. Our results indicate complex shifts in surface energy fluxes with land-cover transitions and a lengthening summer season, and highlight the potential for improving future Earth system models via a refined representation of Arctic vegetation types.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Analytical Model for the Higher Order Moments of Midlatitude Atmospheric Temperature Distributions

Observed distributions of atmospheric temperature are non-Gaussian. Therefore, moments beyond variance are necessary in determining the frequency of extreme temperature events. Here we propose a simple kinematic model for atmospheric mid-latitude temperature variability based on symmetric advection from a non-symmetric background temperature profile. We then use this model to derive analytical expressions for the higher order moments of temperature distributions. Our results show that nonzero skewness and kurtosis arise due to the nonlinearity of the time-mean meridional temperature profile. The analytical model matches an idealized Held-Suarez atmospheric model, indicating nonlinearity of time-mean temperature in latitude is the dominant contribution to nonzero skewness and kurtosis in synoptic temperature variations. Model analysis further shows decrease in higher order moments due to climate change come roughly equally from changes in mixing length and changes in the background temperature profiles.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Uncertainty Quantification and Sensitivity Analysis of Non-Nuclear Advanced Controls Testbed Reactor Mockup

The research presented in this report describes our progress in applying stochastic methods and uncertainty quantification, parametric study, and variance-based sensitivity analysis (also known as Sobol sensitivity analysis) to a full-core model of a nuclear thermal propulsion (NTP) system simulated with Griffin, with the goal of developing a reduced order (surrogate) model which can be rapidly sampled while perturbing multiple input parameters. In this NTP system, reactivity and power feedback affect the rotation of control drums, which are controlled by a hybrid proportional, integral and derivative (PID) controller, actuated by the power demand and reactivity feedback from the numerical model. This model uses reactor kinetic feedback (mean generation time and $\beta$ from a transient Griffin simulation executed with the improved quasi-static method to provide the kinetic parameters) as inputs to functions which control the CD rotation angle. Using a number of stochastic method approaches, we developed a dual purpose training-surrogate model of the NTP system using polynomial regression. The trained model can be rapidly sampled while simultaneously perturbing various input parameters of the model, such as coefficients on the PID control, or temperature (directly affect the neutron cross section). The surrogate model delivers accurate results orders-of-magnitude faster (minutes, not days) than the base model. Once the base model has been trained, distributions of the uncertain parameters can be changed at will to investigate the effects of perturbing multiple inputs and their effect on the output. For example, coefficients used in the PID control system may vary due to some physical interference, or there may be uncertainty in the temperature of the neutron cross sections in various regions of the reactor. A distribution can be placed on these parameters and operational boundaries can be determined. The goal of this work is to support development of an advanced control system to operate CDs in a functioning NTP system.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2022

Analytics-at-scale of Sensor Data for Digital Monitoring in Nuclear Plants (3 rd Annual Report)

Nuclear power plants collect and store large volumes of heterogeneous data from various components and systems. With recent advances in machine learning (ML) techniques, these data can be leveraged to develop diagnostic and short-term forecasting models to better predict future equipment condition. Maintenance operations can then be planned in advance whenever degraded performance is predicted, thus resulting in fewer unplanned outages and the optimization of maintenance activities. This enables lower maintenance costs and improves the overall economics of nuclear power. This report primarily focuses on developing a short-term forecasting process that leverages a feature selection process to distill large volumes of heterogeneous data and predict specific equipment parameters. A variety of feature selection methods, including Shapley Additive Explanations (SHAP) and variance inflation factor (VIF), were used to select the optimal features as inputs for three ML methods: long short-term memory (LSTM) networks, support vector regression (SVR), and random forest (RF). Each combination of model and input features was used to predict a pump bearing temperature both 1 and 24 hours in advance, based on actual plant system data. The optimal inputs for the LSTM and SVR were selected using the SHAP values, while the optimal input for the RF consisted solely of the response variable itself. Each model produced similar 1-hour-ahead predictions, with root mean square errors (RMSEs) of roughly 0.006. For the 24-hour-ahead predictions, differences could be seen between LSTM, SVR, and RF, as reflected by model performances of 0.036 ± 0.014, 0.0026 ± 0, and 0.063 ± 0.004 RMSE, respectively. As big data and continuous online monitoring become more widely available, the proposed feature selection process can be used for many applications beyond the prediction of process parameters within nuclear infrastructure. This report summarizes the Fiscal Year 2021 research progress encompassing the (1) data cleaning and feature selection necessary for ML applications; (2) development of short-term forecasting models to predict future plant process parameters for both single and multiple time steps ahead; and (3) validation of the feature selection methods and short-term forecasting models given new data from different systems.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗