SEARCH · Search NASA

Results for “higher accuracy”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 361 records · Page 20

All sources NASA NTRS DOE OSTI

DOE OSTI2022

DeepAdversaries: examining the robustness of deep learning models for galaxy morphology classification

With increased adoption of supervised deep learning methods for work with cosmological survey data, the assessment of data perturbation effects (that can naturally occur in the data processing and analysis pipelines) and the development of methods that increase model robustness are increasingly important. In the context of morphological classification of galaxies, we study the effects of perturbations in imaging data. In particular, we examine the consequences of using neural networks when training on baseline data and testing on perturbed data. We consider perturbations associated with two primary sources: (a) increased observational noise as represented by higher levels of Poisson noise and (b) data processing noise incurred by steps such as image compression or telescope errors as represented by one-pixel adversarial attacks. We also test the efficacy of domain adaptation techniques in mitigating the perturbation-driven errors. We use classification accuracy, latent space visualizations, and latent space distance to assess model robustness in the face of these perturbations. For deep learning models without domain adaptation, we find that processing pixel-level errors easily flip the classification into an incorrect class and that higher observational noise makes the model trained on low-noise data unable to classify galaxy morphologies. On the other hand, we show that training with domain adaptation improves model robustness and mitigates the effects of these perturbations, improving the classification accuracy up to 23% on data with higher observational noise. Domain adaptation also increases up to a factor of ${\approx}2.3$ the latent space distance between the baseline and the incorrectly classified one-pixel perturbed image, making the model more robust to inadvertent perturbations. Successful development and implementation of methods that increase model robustness in astronomical survey pipelines will help pave the way for many more uses of deep learning for astronomy.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2020

Towards improved speed and accuracy of laser powder bed fusion simulations via multiscale spatial representations

Due to the growing popularity of laser powder bed fusion (LPBF) as a metal additive manufacturing technique, there is a strong need to be able to accurately predict build outcomes. Full fidelity simulations of this process are not feasible due to the vast range of length and time scales inherent to it. While part-scale codes for simulating residual stress and distortion have shown reasonable predictive capability, they often neglect many aspects of the process occurring over smaller length/time scales, and thus are unable to capture effects of process parameter adjustments or the behavior of fine features. One way of capturing aspects at more refined length scales is through the use of adaptive mesh refinement (AMR). AMR allows for the process to be simulated at scales approaching the physical spatial dimensions without drastically increasing the total degrees of freedom in the simulation. This manuscript describes the implementation of an AMR algorithm within a multiphysics, parallelized finite element code, and its application to the LPBF problem. In this work, part-scale examples are provided where the use of AMR has allowed for higher fidelity thermal and thermomechanical simulations, as compared to experimental measurements. Results from these higher resolution simulations show that while AMR is a necessary component for increased accuracy in a computationally efficient manner, other improvements are also necessary, including handling of the multiple time scales inherent to the problem and the need for improved AM-specific material models.

42 ENGINEERING↗

DOE OSTI2025

Novel Deep Learning Transformer Model for Short to Sub‐Seasonal Streamflow Forecast

Accurate short-to-subseasonal streamflow forecasts are becoming crucial for effective water management in an increasingly variable climate. However, streamflow forecast remains challenging over extended lead times, uncertainty in meteorological inputs, and increased frequency and variability in extreme weather and climate events. We implemented a Future Time Series Transformer (FutureTST) model for streamflow forecasting that separately integrates past meteorological and streamflow data while incorporating future weather conditions. FutureTST achieves a mean Nash-Sutcliffe Efficiency (NSE) of 0.82 to 0.67 for 1- to 30-day streamflow forecasts. Incorporating upstream streamflow information improved forecast accuracy by up to 10%. During real-time forecast, FutureTST maintains higher forecast skills of 9.03 for 1-day and 5.74 for 14-day forecasts. In contrast, calibrated process-based hydrological model forecasts become unreliable beyond a 4-day lead time. Our findings demonstrate the potential of FutureTST as a reliable streamflow forecasting tool that offers a valuable addition to operational flood monitoring systems and climate-resilient decision-making.

Ambika, Anukesh Krishnankutty [Oak Ridge National ↗

DOE OSTI2021

Pionless effective field theory evaluation of nuclear polarizability in muonic deuterium

In this paper, we calculate the longitudinal structure function of the deuteron up through next-to-next-to-leading order in the framework of pionless effective field theory. We use these results to compute the two-photon polarizability contribution to Lamb shift in muonic deuterium, which can be utilized to extract the nuclear charge radius of the deuteron. We present analytical expressions order-by-order for the relevant transition matrix elements and the longitudinal structure function, and we give numerical results for the corresponding contributions to the Lamb shift. We also discuss the impact of relativistic and other higher-order effects. We find agreement with previous calculations and explain the accuracy of our calculation.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2026

DP-TwoLevel: two-stage gradient subspace learning for differentially private federated learning

Federated learning (FL) enables collaborative model training across distributed data sources without sharing raw data, but faces fundamental challenges in communication efficiency and privacy. Differentially private (DP) training mitigates information leakage but introduces noise that degrades model performance, especially in high-dimensional settings. We propose DP-TwoLevel, a hierarchical gradient projection method that improves utility under fixed DP constraints by exploiting low-dimensional structure in model updates. Our approach learns a two-level PCA-based representation of gradients and applies DP noise in a reduced-dimensional subspace, thereby lowering the effective noise magnitude while preserving dominant signal components. We evaluate the method across three datasets (MNIST, Fashion-MNIST, CIFAR-10) and three privacy regimes (ϵ∈0.5, 1.0, 2.0). Across nine experimental settings, DP-TwoLevel consistently outperforms DP-FedAvg, achieving an average accuracy improvement of 9.44%, with larger gains observed in lower ϵ(higher-noise) regimes (up to +22.31%). We further analyze scalability across models ranging from 100K to 1.49M parameters and identify a variance-based success criterion: performance remains strong when the projection preserves more than 75% of gradient variance, degrades in a marginal regime (65–75%), and fails below this threshold. Our results demonstrate that structure-aware dimensionality reduction can significantly improve the privacy–utility tradeoff in FL without modifying formal privacy guarantees. We also provide empirical evidence of scaling limitations for global projections and motivate per-layer extensions for larger models.

Kotevska, Olivera [ORNL] (ORCID:0000000316772243)↗

DOE OSTI2023

Mixed-Precision S/DGEMM Using the TF32 and TF64 Frameworks on Low-Precision AI Tensor Cores

Using NVIDIA graphics processing units (GPUs) equipped with Tensor Cores has enabled the significant acceleration of general matrix multiplication (GEMM) for applications in machine learning (ML) and artificial intelligence (AI) and in high-performance computing (HPC) generally. The use of such power-efficient, specialized accelerators can provide a performance increase between 8 × and 20 ×, albeit with a loss in precision. However, a high level of precision is required in many large scientific and HPC applications, and computing in single or double precision is still necessary for many of these applications to maintain accuracy. Fortunately, mixed-precision methods can be employed to maintain a higher level of numerical precision while also taking advantage of the performance increases from computing with lower-precision AI cores. With this in mind, we extend the state of the art by using NVIDIA’s new TF32 framework. This new framework not only burdens some constraints of the previous frameworks, such as costly 32 16-bit castings but also provides an equivalent precision and performance by using a much simpler approach. We also propose a new framework called TF64 that attempts double-precision arithmetic with low-precision Tensor Cores. Although this framework does not exist yet, we validated the correctness of this idea and achieved an equivalent of 64-bit precision on 32-bit hardware.

Valero Lara, Pedro↗

DOE OSTI2026

Scalability Analysis of Quantum Models for Stress and Emotion Detection

Stress and emotion detection from high-dimensional physiological signals is a challenging task, particularly when aiming for accurate classification across diverse behavioral states. Quantum machine learning (QML) is promising for modeling such high-dimensional data, but scalability is limited by qubit resources and the exponential cost of classical statevector simulation. This work studies the scalability of quantum support vector machines (QSVMs) for binary stress detection and three-class emotion recognition (Negative/Neutral/Positive) under varying qubit counts and angle-encoding strategies. We also present a comparison study with one-feature-per-qubit (1:1) and two-features-per-qubit (2:1) mappings. Experiments are executed on HPC infrastructure using NVIDIA CUDA-Q to evaluate performance, variance, and class-dependent separability at higher-qubit setups. Results show that larger Hilbert spaces can improve peak accuracy but may increase instability. At the same time, dense 2:1 encoding yields more consistent stress detection performance. For emotion recognition, scaling improves discrimination for classes like Negative and Positive more than Neutral. We find that effective QML scaling is task-dependent and benefits more from encoding design than simply increasing qubit count.

Onim, Md. Saif Hassan [University of Tennessee, Kn↗

DOE OSTI2025

Heat load measurements for the PIP-II pHB650 cryomodule

This study presents a brief overview of the 1st and 2nd phases and an in-depth analysis of the 3rd phase heat load testing performed on the pHB650 (prototype High Beta 650 MHz) cryomodule at PIP2IT (PIP-II Injector Test Facility), with a focus on both the results and the methodological advancements that have improved testing efficiency and accuracy. A key challenge identified in the testing campaign is the higher-than-expected heat loads observed in the first PIP-II (Proton Improvement Plan II) prototype cryomodules (pSSR1 and pHB650) tested at PIP2IT. Elevated heat loads are concerning given the fixed capacity of the PIP-II cryoplant that is currently being installed at Fermilab. However, understanding the sources of these elevated heat loads offers a critical opportunity to implement effective heat load mitigations on upcoming PIP-II cryomodules to stay within the available capacity of the PIP-II cryoplant. The study includes a summary of test results, descriptions of measurement procedures, and key observations on parameters directly and indirectly related to heat load measurements. Direct observations include measured heat loads and the effectiveness of JT heat exchanger under varying conditions, while indirect observation analyze factors such as the temperature distribution on the two-phase pipe and relief piping under varying conditions. Thermal acoustic oscillations (TAO) were identified during testing, which was mitigated by replacing the original G10 stem with a stainless steel stem equipped with wipers for the cryomodule cooldown valve. A major innovation during pHB650 Phase 3 testing was the development of an automated Python script to streamline data acquisition, analysis, and reporting of heat load results. This script automatically retrieved data from ACNET (Accelerator Control Network), performed heat load calculations, and generated detailed reports featuring plots and tables. This advancement significantly reduced manual labor and enhanced the thoroughness of data analysis compared to earlier campaigns. The heat load test reports were promptly uploaded to the electronic logbook shortly after each test, enabling rapid feedback and collaboration between the SRF and cryogenic teams. The heat load measurements included various components: HTTS (high-temperature thermal shield), LTTS (low-temperature thermal shield), 2K isothermal and non-isothermal heat loads. Results were recorded both within the cryomodule and between the bayonet can supply and return. Measurements were conducted under different operating conditions such as "standard", "linac", and "simulated dynamic". Additionally, HTTS and LTTS heat loads were calculated in real time, allowing for the tracking of thermal stability and identification of changes during testing, both in steady-state and transient conditions. The results of this testing campaign not only provide valuable insights into the performance of the pHB650 cryomodule but also highlight best practices and lessons learned that will inform future cryomodule testing at PIP2IT. These include adopting automated tools for data analysis, refining real-time measurement capabilities, and emphasizing detailed pre-test planning. The framework established in this campaign aims to set an improved standard for cryomodule testing and heat load reporting in future cryomodule test campaigns.

Porwisiak, D. [Fermilab; Wroclaw Tech. U.]↗

DOE OSTI2026

Improving the Precision of First-Principles Calculation of Parton Physics from Lattice Quantum Chromodynamics

Large momentum effective theory (LaMET) provides a general framework for computing the multi-dimensional partonic structure of the proton from first principles using lattice quantum chromodynamics (QCD). In this effective field theory approach, LaMET predicts parton distributions through a power expansion and perturbative matching of a class of Euclidean observables—quasi-distributions—evaluated at large proton momenta. Recent advances in lattice renormalization, such as the hybrid scheme with leading renormalon resummation, together with improved matching kernel that incorporates higher-loop corrections and resummations, have enhanced both the perturbative and power accuracy of LaMET, enabling a reliable quantification of theoretical uncertainties. Moreover, the Coulomb-gauge correlator approach further simplifies lattice analyses and improves the precision of transverse-momentum-dependent structures, particularly in the non-perturbative region. State-of-the-art LaMET calculations have already yielded certain parton observables with important phenomenological impact. In addition, the recently proposed kinematically enhanced lattice interpolation operators promise access to unprecedented proton momenta with greatly improved signal-to-noise ratios, which will extend the range of LaMET prediction and further suppress the power corrections. The remaining challenges, such as controlling excited-state contamination in lattice matrix elements and extracting gluonic distributions, are expected to benefit from emerging lattice techniques for ground-state isolation and noise reduction. Thus, lattice QCD studies of parton physics have entered an exciting stage of precision control and systematic improvement, which will have a broader impact for nuclear and particle experiments.

Zhao, Yong [Argonne National Laboratory (ANL), Arg↗

DOE OSTI2022

Dynamic System Scaling Application to Accelerated Nuclear Fuel Testing

The development of nuclear fuel and materials requires a continuous effort to investigate the acute and prolonged effects of irradiation, thermal-material stress, chemical change, or other conceivable damage mechanics acquired during normal operation or accident scenarios throughout its lifetime. As in-core fuel property measurement techniques advance to support in the real-time, non-invasive, and enhanced accuracy realm, it is the future of fuel development to pursue to higher degrees of control, predictability of integral test behavior via separate effects test (SET), and shorter test time intervals. The fuel development life cycle from initial concept to commercial licensing is approximated to be 20 years and current literature suggests by optimizing fuel performance codes with SETs, the process could possibly be compressed to 5 to 10 years. Recently, a research group in the Idaho National Laboratory (INL) is testing reduced scale fuel rods and increased power density to accelerate evolution of fuel phenomena in metallic fuels. In support of nuclear fuel rod development, compressing fuel test process, and accelerating fuel phenomena, it was the purpose of this study to investigate nuclear fuel performance phenomena via literature review and effectively scale the initial conditions, boundary conditions, and geometric properties to describe the time-dependent response including to fuel burnup, thermal and mechanical stress, transmutations and inter-diffusion, and other relevant observed phenomena. The study was based on BISON simulations of historic EBR-II metallic fuel experiments and Dynamical System Scaling (DSS) method are utilized to assess effects of scaling activity including fuel phenomena acceleration and calculations of time-dependent distortions. The research successfully scaled metallic fuel phenomena, accelerated fuel testing, and assessed the distortions for each scaled case derived.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

DOE OSTI2022

A deep potential model with long-range electrostatic interactions

Machine learning models for the potential energy of multi-atomic systems, such as the deep potential (DP) model, make molecular simulations with the accuracy of quantum mechanical density functional theory possible at a cost only moderately higher than that of empirical force fields. However, the majority of these models lack explicit long-range interactions and fail to describe properties that derive from the Coulombic tail of the forces. To overcome this limitation, we extend the DP model by approximating the long-range electrostatic interaction between ions (nuclei + core electrons) and valence electrons with that of distributions of spherical Gaussian charges located at ionic and electronic sites. The latter are rigorously defined in terms of the centers of the maximally localized Wannier distributions, whose dependence on the local atomic environment is modeled accurately by a deep neural network. In the DP long-range (DPLR) model, the electrostatic energy of the Gaussian charge system is added to short-range interactions that are represented as in the standard DP model. The resulting potential energy surface is smooth and possesses analytical forces and virial. Missing effects in the standard DP scheme are recovered, improving on accuracy and predictive power. By including long-range electrostatics, DPLR correctly extrapolates to large systems the potential energy surface learned from quantum mechanical calculations on smaller systems. We illustrate the approach with three examples: the potential energy profile of the water dimer, the free energy of interaction of a water molecule with a liquid water slab, and the phonon dispersion curves of the NaCl crystal.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2026

Deployment of Traditional and Hybrid Machine Learning for Critical Heat Flux Prediction in the CTF Thermal-Hydraulics Code

Critical heat flux (CHF) marks the transition from nucleate to film boiling, where heat transfer to the working fluid can rapidly deteriorate. Accurate CHF prediction is essential for efficiency, safety, and preventing equipment damage, particularly in nuclear reactors. Although widely used, empirical correlations frequently exhibit discrepancies when compared to experimental data, limiting their reliability in diverse operational conditions. Traditional machine learning (ML) approaches have demonstrated potential for CHF prediction but often suffer from limited interpretability, data scarcity, and insufficient knowledge of physical principles. Hybrid model approaches, which combine data-driven ML with base models, mitigate these concerns by incorporating prior knowledge of the domain. This study integrates an externally trained purely data-driven ML model and two hybrid models (using the Biasi and Bowring CHF correlations) within the CTF subchannel code via a custom Fortran framework. Performance was evaluated using two validation cases: a subset of the Nuclear Regulatory Commission (NRC) CHF database and the Bennett dryout experiments. In both cases, the hybrid models demonstrated significantly lower error metrics compared to conventional empirical correlations, with the best models often reducing relative error by about 5 percentage points. The pure ML model achieved comparable accuracy, outperforming the hybrid Biasi model in the NRC test case (3.3% versus 5.5% relative error) but exhibiting slightly higher error against the hybrid Bowring model in the Bennett test case (7.7% versus 6.1%). Trend analysis of error parity indicated that ML-based models reduced the tendency for CHF overprediction, improving overall accuracy. These results demonstrate that ML-based CHF models can be effectively integrated into subchannel codes and could potentially increase performance compared to conventional methods.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2025

Dynamical Sketching for Enhanced Communication Efficiency in Federated Learning

Federated learning (FL) has revolutionized distributed machine learning by enabling collaborative model training without sharing local data. However, communication efficiency and privacy guarantees remain significant challenges. This paper introduces a dynamic sketching mechanism in FL, optimizing the trade-off between communication efficiency and model accuracy. By dynamically selecting the sketch matrix size, our approach adapts to the evolving characteristics of the data and the model, ensuring optimal performance across diverse scenarios. We leverage Bayesian optimization to systematically tune the sketch parameters, achieving an effective balance between resource efficiency and model performance. Experimental results on the MNIST dataset using a convolutional neural network (CNN) architecture validate the proposed method's efficiency and scalability. Our dynamic sketching approach significantly outperforms fixed-size sketching techniques, achieving higher compression ratios (up to 62x) and providing better privacy guarantees while maintaining high model accuracy. These findings highlight the robustness and versatility of our approach and make it a valuable solution for privacy-preserving, communication-efficient federated learning.

Afrose, Sharmin [ORNL]↗

DOE OSTI2020

Accuracy of self‐reported weight, height and body mass index among older people in Japan

Aim Weight and height are usually self‐reported in population‐based epidemiological surveys. While the accuracy of self‐reports has been extensively studied in younger populations, less is known in older populations. We investigated the accuracy of self‐reported weight, height and body mass (BMI) in an older cohort in Japan, where overweight/obesity and underweight coexist. Methods We used data from older Japanese adults (≥65 years) participating in the Japan Gerontological Evaluation Study in 2016 to 2017 (7357 men and 9271 women). Self‐report data were linked to objective data obtained from clinical examinations. Results The mean ± standard deviation (SD) age was 74.5 ± 5.8 years, mean ± SD weight, height and BMI were 55.7 ± 10.1 kg, 156.0 ± 8.9 cm and 22.8 ± 3.1, respectively. Results showed high intraclass correlation coefficients for self‐reported and measured values (0.97 for weight; 0.96 for height). While weight/height were overestimated among men (weight by 0.096 kg; height by 0.27 cm) and women (weight by 0.18 kg; height by 0.27 cm), BMI tended to be slightly underestimated (−0.034 kg/m 2 for men; −0.037 kg/m 2 for women). However, the absolute differences between self‐reported and measured values were not negligible; people had a higher risk for both under‐ and overestimation of their BMI category with increasing age. Lower education predicted BMI overestimation, whereas lower income predicted BMI underestimation. Conclusions Overall accuracy of self‐reported body habitus was higher in this cohort of older Japanese compared with previous reports. Nevertheless, misclassification of BMI due to the misreporting of their weight/height was more common among the oldest‐old, as well as those with lower education and income. Geriatr Gerontol Int 2020; 20: 803–810 .

Yazawa, Aki↗

DOE OSTI2024

Modeling the effect of wind speed and direction shear on utility‐scale wind turbine power production

Abstract Wind speed and direction variations across the rotor affect power production. As utility‐scale turbines extend higher into the atmospheric boundary layer (ABL) with larger rotor diameters and hub heights, they increasingly encounter more complex wind speed and direction variations. We assess three models for power production that account for wind speed and direction shear. Two are based on actuator disc representations, and the third is a blade element representation. We also evaluate the predictions from a standard power curve model that has no knowledge of wind shear. The predictions from each model, driven by wind profile measurements from a profiling LiDAR, are compared to concurrent power measurements from an adjacent utility‐scale wind turbine. In the field measurements of the utility‐scale turbine, discrete combinations of speed and direction shear induce changes in power production of −19% to +34% relative to the turbine power curve for a given hub height wind speed. Positive speed shear generally corresponds to over‐performance and increasing magnitudes of direction shear to greater under‐performance, relative to the power curve. Overall, the blade element model produces both higher correlation and lower error relative to the other models, but its quantitative accuracy depends on induction and controller sub‐models. To further assess the influence of complex, non‐monotonic wind profiles, we also drive the models with best‐fit power law wind speed profiles and linear wind direction profiles. These idealized inputs produce qualitative and quantitative differences in power predictions from each model, demonstrating that time‐varying, non‐monotonic wind shear affects wind power production.

Energy & Fuels↗

DOE OSTI2020

Modeling and analysis of heat emissions from buildings to ambient air

Heat emissions from buildings is a significant source of anthropogenic heat influencing the urban microclimate; however, they are usually oversimplified in urban climate and microclimate modeling. This study developed a bottom-up physics-based approach to calculate heat emissions from buildings to the ambient air and implemented the approach in EnergyPlus. In this work, a simple result verification was conducted by comparing the EnergyPlus simulated results against the spreadsheet calculations. Simulations covering 16 commercial building types, four climates, and two energy efficiency levels were conducted to understand and evaluate the building heat emissions and their temporal patterns as well as three major components: (1) building envelope (convective heat transfer to ambient air), (2) zones (air exfiltration and exhaust air), and (3) HVAC systems (relief air and heat rejection from condensers or cooling towers). The main findings are: (1) heat emissions are usually higher than the site energy use (about 2.5 times), and their dynamics should be considered; (2) building characteristics and their energy systems lead to differences in heat emission contributions from the three components, and their dynamics, for example, in the warehouse models, the envelope component accounts for 90.4%, while it is 12.7% for the large office models; (3) for most building typologies, the climate has a strong impact on heat emissions, for example, buildings with dominant heat emissions from the zone exhaust air and/or the HVAC reject heat, a general decrease in heat emissions in hotter climates is observed, while envelope-dominated buildings show the opposite; and (4) building technologies that reduce energy use in buildings may perform differently in reducing heat emissions. The developed heat emissions calculation method can be adopted in EnergyPlus and most other building energy modeling programs. It can provide dynamic building heat emissions as an input to urban climate computational fluid dynamics (CFD) models at a higher spatial and temporal resolution than is currently available, to improve the simulation accuracy of the urban microclimate and capture the urban heat island effect and urban overheating.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

DOE OSTI2023

Privacy-preserving federated learning: Application to behind-the-meter solar photovoltaic generation forecasting

Here, the growing usage of decentralized renewable energy sources has made accurate estimation of their aggregated generation crucial for maintaining grid flexibility and reliability. However, the majority of distributed photovoltaic (PV) systems are behind-the-meter (BTM) and invisible to utilities, leading to three challenges in obtaining an accurate forecast of their aggregated output. Firstly, traditional centralized prediction algorithms used in previous studies may not be appropriate due to privacy concerns. There is therefore a need for decentralized forecasting methods, such as federated learning (FL), to protect privacy. Secondly, there has been no comparison between localized, centralized, and decentralized forecasting methods for BTM PV production, and the trade-off between prediction accuracy and privacy has not been explored. Lastly, the computational time of data-driven prediction algorithms has not been examined. This article presents a FL power forecasting method for PVs, which uses federated learning as a decentralized collaborative modeling approach to train a single model on data from multiple BTM sites. The machine learning network used to design this FL-based BTM PV forecasting model is a multi-layered perceptron, which ensures privacy and security of the data. Comparing the suggested FL forecasting model to non-private centralized and entirely private localized models revealed that it has a high level of accuracy, with an RMSE that is 18.17% lower than localized models and 9.9% higher than centralized models.

14 SOLAR ENERGY↗

DOE OSTI2025

A hybrid surrogate modeling framework for the Digital Twin of a Fluoride-salt-cooled High-temperature Reactor (FHR)

While nuclear energy is a non-greenhouse-gas emitting energy source, expensive operational costs due to the high-level of safety requirements decreases their competitiveness in the sustainable energy market. Advanced reactor concepts paired with Digital Twins aim to increase the commercialization gains of nuclear energy by reducing operational costs, increasing reactor reliability and enhancing power generation. To support Digital Twin tasks such as real-time autonomous control, proactive maintenance monitoring or optimizing power demand operations, a fast and accurate virtual representation of the Nuclear Power Plant (NPP) is required. The computational cost of high-fidelity, physics-based models are unsuitable for real-time analysis or scalability. Here, in this work, a hybrid surrogate modeling framework is developed fora Fluoride-salt-cooled High-temperature Reactor (FHR) that leverages physics-inspired models for key reactor components and uses data-driven methods for rapid system state space prediction. The Xenon reactivity feedback model is integrated to inform the surrogate model about the reactor core and the homologous pump theory model is the basis for representing pump degradation. Using a detailed, two dimensional thermal hydraulics model to generate data on the FHR, we train a network of Vectorized Autoregressive Moving-Average with eXogenous input (VARMAX) models to predict the remaining state values. The result is a surrogate model that provides a detailed reactor state representation of 41 system states and a pump degradation analysis. The framework is applied to Load Follows profiles, yielding high accuracy and a speedup that is more than 4000x faster compared to the higher- fidelity thermal hydraulics model, enabling real-time operational intelligence and applications in long horizon predictions. While the surrogate model framework is demonstrated for the particular case of FHR, the hybrid physical/data-driven modeling approach including the network of surrogates and the underlying modularity has the potential to be applied to other physical asset systems.

Digital Twins↗