SEARCH · Search NASA

Results for “Reliability model reduction”

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 73 records · Page 4

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Bayesian Optimized Deep Ensemble for Uncertainty Quantification of Deep Neural Networks: a System Safety Case Study on Sodium Fast Reactor Thermal Stratification Modeling

Deep neural networks (DNNs) are increasingly important to scientific computing and engineering system simulations. Accurate uncertainty quantification (UQ) for DNNs is critical in safety-sensitive engineering domains. Traditional Deep Ensemble (DE) methods, while easy to implement, frequently suffer from poorly calibrated uncertainty estimates and limited predictive accuracy due to reliance on fixed architectures with varied weight initializations. To address these issues, we introduce a workflow that combines Bayesian Optimization (BO) and DE. The workflow is modular, scalable, and integrates parallel BO initialized with Sobol sequences to individually optimize the hyperparameters of each ensemble member. This method enhances ensemble diversity, improves predictive accuracy, and provides reliable uncertainty estimates. We evaluate the proposed BODE approach in a sodium fast reactor thermal stratification modeling case study, where we used a densely connected convolutional neural network to predict turbulent viscosity during the reactor transient with consideration of data noise. We benchmark its performance against several optimization approaches, including baseline deep ensemble, evolutionary algorithm-optimized ensemble, ensemble formed via random search combined with greedy selection, and a BO ensemble using random initialization. Here, our results demonstrate superior performance of the developed BODE approach. In noise-free scenarios, BODE notably reduces incorrect aleatoric uncertainty and significantly enhances predictive accuracy. Under conditions of 5% and 10% Gaussian noise, BODE adaptively quantifies uncertainty proportional to data noise, achieving up to an 80% reduction in root mean square error compared to baseline methods and producing well-calibrated prediction intervals.

Bayesian optimization↗

DOE OSTI2026

Demand response event simulator and risk-aware bidding tool for industrial customers

Incentive Based Demand Response (IBDR) program participation delivers financial benefits to the consumers and resiliency benefits to the electricity grid. Effectively participating in these programs as an industrial consumer requires bidding strategies that balance financial risk with operational constraints. Existing bidding tools tend not to fully incorporate stochastic IBDR event modeling, program specific baseline and payment/penalty calculations, or demand reduction process control schemes that account for the cascading impacts of shutdown in complex facilities. Here, this work presents an IBDR event simulator and risk-aware bidding framework tool integrating three key components: a flexible, parameterized demand response event generator that rigorously accounts for program structures and stochasticity, a demand response operational simulation model that generates explicit control strategies for load reduction, and a Monte Carlo simulator to evaluate financial risk for varied capacity bids. A case study at a wastewater treatment plant participating in PG&E's Capacity Bidding Program demonstrates the framework's utility. In the peak capacity price month of August, optimal bidding by the wastewater treatment plant nets a mean IBDR benefit of $101,000 (67% of the August electricity bill) with 0.4% probability of a financial loss. This framework enables industrial operators to make informed bidding decisions, negotiate better program terms with demand response load aggregators, and analyze energy flexibility investments at their facilities. Ultimately, this work reduces participation barriers in IBDR programs and supports the broader goal of enhancing grid reliability and renewable energy integration.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

DOE OSTI2024

Power generation-cooling water Nexus: Impacts of cooling water shortage on power system operation - a simulation case study in Illinois, U.S

Cooling water shortage, frequently attributed to drought and heat waves, poses a significant threat to the operations of thermoelectric power plants and further poses a challenge for the entire power system and environmental stakeholders. Recognizing the critical nexus between power generation and cooling water availability and the potential ability of power generations to adjust generation schedules during cooling water shortages, this paper introduces a security-constrained unit commitment and economic dispatch model considering water-energy nexus. In specific, the model is augmented with a unit-level cooling water requirement (CWR) model and multi-level cooling water availability (CWA) constraints. The unit-level CWR model quantifies the cooling water withdrawal per MWh of power generation, taking into account factors such as thermoelectric generation technologies, cooling system technologies, and environmental parameters. The multi-level CWA constraints incorporate pump-level, plant-level, watershed-level, and forced minimum power constraints, utilizing data derived from actual-based cooling water shortage scenarios. Using a simulation case study in Illinois, United States, this research examines the reliability, economic, and environmental implications of cooling water shortages on power system operations. The results show that Illinois may experience 10-15% daily load curtailment and severe congestion between certain regions from the east to central during cooling water shortages, while once-through and wet-tower units experience a 52% and 17% reduction in power generation. In conclusion, overall cooling water withdrawal decreases by 24-38% as severity intensifies.

Cooling water shortage↗

DOE OSTI2025

Build-To-Replace Strategy to Reduce O&M Costs of Advanced Reactors

This paper targets the goal of reducing fixed operations and maintenance (O&M) costs to $2/MWh for advanced reactor (AR) designs, an order of magnitude reduction from current nuclear fleets. Such goal will be severely constrained if the underlying assumptions of O&M approaches and practices are not questioned and reexamined, especially today when changes can be implemented effectively and efficiently. Achieving a 90% reduction in AR O&M costs requires a completely new way of thinking – a paradigm shift – not through incremental, technology-focused approaches alone. Here, we address this challenge by evaluating the impact of moving to shorter design life for major structures, systems and components (SSCs) and shorter, more predictable refurbishment cycles as modeled by the commercial airline industry. This paper provides a brief overview of this different mindset to O&M applied to ARs: the build-to-replace approach. We provide a direct example of analysis of a build-to-replace scenario by identifying and evaluating scenarios for reduced system and component lifetimes and associated replacement and refurbishment schedules to evaluate impacts on O&M costs and other lifecycle elements such as SSC reliability.

97 - MATHEMATICS AND COMPUTING↗

DOE OSTI2026

Electrochemical behavior of SnCl 2 and influence of Cu and Ni ions in molten LiCl−KCl−CaCl 2 eutectic

Reliable transport and thermodynamic data for multivalent ions in complex molten salts are scarce, limiting model fidelity for electrorefining and impurity control. Here, we report a comprehensive electrochemical characterization of SnCl₂ in LiCl–KCl–CaCl₂ (50.5–44.2–5.3 mol%) at 685 K, including the effects of Ni 2+ and Cu + impurities. Using cyclic voltammetry (CV), chronoamperometry (CA), and chronopotentiometry (CP), we quantified Sn 2+ and Ni 2+ diffusion with exceptional agreement across methods: Sn 2+ averaged (1.03 ± 0.10) × 10 −5 cm 2 s −1 , and Ni 2+ averaged (0.75 ± 0.19) × 10 −5 cm 2 s −1 . The tight confidence-interval overlap across CV, CA, and CP strengthens confidence in these values and is uncommon in molten chloride studies. Open-circuit-potential measurements provided standard apparent reduction potentials that closely match LiCl–KCl literature, indicating minimal shift with CaCl₂ present. The Sn 2+ /Sn couple behaves as a reversible two-electron soluble–insoluble process at 685 K; the Sn 4+ /Sn 2+ couple transitions to soluble–soluble behavior near 788 K, which may correlate with the decomposition of surface bound chlorostannates, though direct characterization remains to be established. In mixed systems, Cu+/Cu overlaps Sn 2+ /Sn, limiting Cusingle bondSn electroseparation, whereas the larger potential gap between Ni 2+ /Ni and Sn 2+ /Sn supports selective Ni removal. These internally consistent transport and thermodynamic data establish a validated basis for process modeling and optimization of Sn electrorefining and impurity management in LiCl–KCl–CaCl₂.

Berzins-Delahay↗

DOE OSTI2025

On the optimal sizing of power take-off systems for wave energy converters

The power take-off (PTO) system in a wave energy converter (WEC) is the means by which the energy in an ocean wave is converted into useful energy. There has been a general lack of technology convergence in the wave energy industry and PTO design has been largely device-specific with the design process determined by the WEC developer. An optimized PTO design can significantly affect the efficiency, reliability, performance, and overall cost and viability of a device. There remains a need in the marine energy industry for best practices and efficient design processes to be identified as it could greatly benefit developers and researchers in designing an optimal WEC PTO. This research utilizes the open-source Matlab-Simulink based software, WEC-Sim, which was developed by the National Renewable Energy Laboratory, and the Eagle, a high-performance computing system, to model and optimize two WECs of different archetypes and then investigates if any trends exist that can be exploited for greater efficiency in optimal PTO design. The results from this research indicate that there may be an optimal sizing for WEC PTOs for both power and PTO force rating and that substantial reductions can be made in the cost of a device without significant loss in the energy produced. Furthermore this optimal sizing may be independent of WEC type and deployment location. With an optimal, or near optimal, approach to WEC PTO sizing demonstrated, a methodology is proposed to address the challenge of nameplate ratings in the wave energy industry.

13 HYDRO ENERGY↗

DOE OSTI2024

Neutron-producing gas puff Z-pinch experiments on a fast, low-impedance, 0.5 MA linear transformer driver

A study on the neutron production from single and double gas puff Z-pinches on the CESZAR linear transformer driver with ~0.45 MA current and 170 ns rise time is presented. Total neutron yield measurements made with a LaBr activation detector are compared for three configurations, using a double nozzle setup. When a single, hollow, deuterium gas shell was used, reliable implosions could only be attained at higher load mass than the optimal value to match implosion time with the driver rise time, with neutron yields of ~10 6 per pulse. The use of a double gas puff configuration with a deuterium center jet allowed a reduction in the shell density and operation closer to machine-matched conditions, recording up to (4.1 ± 0.3) × 10 7 neutrons/pulse when either Kr or D 2 was used in the shell. For a comparable mass and implosion time, using a higher atomic-number gas in the outer shell results in more unstable plasma surface and smaller plasma radius at the location of instability bubbles, which, however, do not seem to consistently correlate with a higher neutron yield. Comparing implosion dynamics with models and neutron yields with literature scaling suggests that the machine current is not well coupled to the plasma during the final stages of compression. Optimizing current and energy coupling to the pinched plasma is critical to improving performance, particularly in low-impedance drivers.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2025

Robustness of Deep Learning Classification to Adversarial Input on GPUs: Asynchronous Parallel Accumulation Is a Source of Vulnerability

The ability of machine learning (ML) classification models to resist small, targeted input perturbations—known as adversarial attacks—is a key measure of their safety and reliability. We show that floating-point non associativity (FPNA) coupled with asynchronous parallel programming on GPUs is sufficient to result in misclassification, without any perturbation to the input. Additionally, we show that this misclassification is particularly significant for inputs close to the decision boundary and that standard adversarial robustness results may be overestimated up to 4.6 when not considering machine-level details. We first study a linear classifier, before focusing on standard Graph Neural Network (GNN) architectures and datasets used in robustness assessments. We develop a novel black-box attack using Bayesian optimization to discover external workloads that can change the instruction scheduling which bias the output of reductions on GPUs and reliably lead to misclassification. Motivated by these results, we present a new learnable permutation (LP) gradient-based approach to learning floating-point operation orderings that lead to misclassifications. The LP approach provides a worst-case estimate in a computationally efficient manner, avoiding the need to run identical experiments tens of thousands of times over a potentially large set of possible GPU states or architectures. Finally, using instrumentation-based testing, we investigate parallel reduction ordering across different GPU architectures under external background workloads, when utilizing multi-GPU virtualization, and when applying power capping. Our results demonstrate that parallel reduction ordering varies significantly across architectures under the first two conditions, substantially increasing the search space required to fully test the effects of this parallel scheduler-based vulnerability. These results and the methods developed here can help to include machine-level considerations into adversarial robustness assessments, which can make a difference in safety and mission critical applications.

Shanmugavelu, Sanjif [Maxeler Technologies, a Groq↗

DOE OSTI2025

Early Fault Detection in Nuclear Systems: A Digital Engineering Approach

Nuclear energy systems present unique challenges in terms of ensuring safety, reliability, and efficiency during their design and operation. Early fault detection is critical for mitigating risks and fostering system resilience. However, current methods often fall short at identifying faults during early stages, potentially leading to costly delays and safety risks. The present work proposes a comprehensive digital engineering approach that leverages digital twins, digital threads, model-based systems engineering, artificial intelligence, and immersive extended reality to support early fault detection in nuclear systems. Through a series of case studies, we highlight specific gaps in the fault detection mechanisms of traditional nuclear design and operation processes, then demonstrate a suite of solutions we are working to implement to address these shortcomings in similar projects. Our findings suggest that a digital engineering approach to design and operation can significantly improve fault detection, ultimately leading to reductions in risk.

42 - ENGINEERING↗

DOE OSTI2025

Explainable AI for Multivariate Time Series Pattern Exploration: Latent Space Visual Analytics With Temporal Fusion Transformer and Variational Autoencoders in Power Grid Event Diagnosis

Detecting and analyzing complex patterns in multivariate time-series data is crucial for decision-making in urban and environmental system operations. However, challenges arise from the high dimensionality, intricate complexity, and interconnected nature of complex patterns, which hinder the understanding of their underlying physical processes. Existing AI methods often face limitations in interpretability, computational efficiency, and scalability, reducing their applicability in real-world scenarios. This paper proposes a novel visual analytics framework that integrates two generative AI models, Temporal Fusion Transformer (TFT) and Variational Autoencoders (VAEs), to reduce complex patterns into lower-dimensional latent spaces and visualize them in 2D using dimensionality reduction techniques such as PCA, t-SNE, and UMAP with DBSCAN. These visualizations, presented through coordinated and interactive views and tailored glyphs, enable intuitive exploration of complex multivariate temporal patterns, identifying patterns’ similarities and uncover their potential correlations for a better interpretability of the AI outputs. The framework is demonstrated through a case study on power grid signal data, where it identifies multi-label grid event signatures, including faults and anomalies with diverse root causes. Additionally, novel metrics and visualizations are introduced to validate the models and assess the performance, efficiency, and consistency of latent maps generated by VAE, which have been utilized in prior studies for latent space cartography and used as a benchmark in this study, and the emerging TFT architecture under various configurations. These analyses provide actionable insights for model parameter tuning and reliability improvements. Comparative results highlight that TFT achieves shorter run times and superior scalability to diverse time-series data shapes compared to VAE. This work advances fault diagnosis in multivariate time series, fostering explainable AI to support critical system operations.

Explainable AI↗

DOE OSTI2026

Toward real-time optimization through model reduction and model discrepancy sensitivities

Optimization problems arise in a range of scenarios, from optimal control to model parameter estimation. In many applications, such as the development of digital twins, it is essential to solve these optimization problems within wall-clock-time limitations. However, this is often unattainable for complex systems, such as those modeled by nonlinear partial differential equations. One strategy for mitigating this issue is to construct a reduced-order model (ROM) that enables more rapid optimization. In particular, the use of nonintrusive ROMs—those that do not require access to the full-order model at evaluation time—is popular because they facilitate the computation of optimization solutions within the wall-clock time requirements. However, the optimization solution will be unreliable if the iterates move outside the ROM training data. This article proposes the use of hyper-differential sensitivity analysis with respect to model discrepancy (HDSA-MD) as a computationally efficient tool to augment ROM-constrained optimization and improve its reliability. The proposed approach consists of two phases: (i) an offline phase where several full-order model evaluations are computed to train the ROM, and (ii) an online phase where a ROM-constrained optimization problem is solved, a limited number of full-order model evaluations are computed, and HDSA-MD is used to enhance the optimization solution. Numerical results are demonstrated for two examples, atmospheric contaminant control and wildfire ignition location estimation, in which a ROM is trained offline using inaccurate atmospheric data. In conclusion, the HDSA-MD update yields a significant improvement in the ROM-constrained optimization solution using only one full-order model evaluation online with corrected atmospheric data.

PDE-constrained optimization↗

DOE OSTI2024

Multi-scale impacts of climate change on hydropower for long-term water-energy planning in the contiguous United States

Climate change impacts on watersheds can potentially exacerbate water scarcity issues where water serves multiple purposes including hydropower. The long-term management of water and energy resources is still mostly approached in a siloed manner at different basins or watersheds, failing to consider the potential impacts that may concurrently affect many regions at once. There is a need for a large-scale hydropower modeling framework that can examine climate impacts across adjoining river basins and balancing authorities (BAs) and provide a periodic assessment at regional to national scales. Expanding from our prior assessment only for the United States (US) federal hydropower plants, we enhance and extend two regional hydropower models to cover over 85% of the total hydropower nameplate capacity and present the first contiguous US-wide assessment of future hydropower production under Coupled Model Intercomparison Project phase 6’s high-end Shared Socioeconomic Pathway 5-8.5 emission scenario using an uncertainty-aware multi-model ensemble approach. We present regional hydropower projections, using both BA regions and US Hydrologic Subregions (HUC4s), to consistently inform the energy and water communities for two future periods—the near-term (2020–2039) and the mid-term (2040–2059) relative to a historical baseline period (1980–2019). We find that the median projected changes in annual hydropower generation are typically positive—approximately 5% in the near-term, and 10% in the mid-term. However, since the risk of regional droughts is also projected to increase, future planning cannot overly rely on the ensemble median, as the potential of severe hydropower reductions could be overlooked. The assessment offers an ensemble of future hydropower generation projections, providing regional utilities and power system operators with consistent data to develop drought scenarios, design long duration storage and evaluate energy infrastructure reliability under intensified inter-annual and seasonal variability.

13 HYDRO ENERGY↗

DOE OSTI2025

Navigating the Noise: Bringing Clarity to ML Parameterization Design With O $\boldsymbol{\mathcal{O}}$(100) Ensembles

Abstract Machine‐learning (ML) parameterizations of subgrid processes (here of turbulence, convection, and radiation) may one day replace conventional parameterizations by emulating high‐resolution physics without the cost of explicit simulation. However, uncertainty about the relationship between offline and online performance (i.e., when integrated with a large‐scale general circulation model) hinders their development. Much of this uncertainty stems from limited sampling of the noisy, emergent effects of upstream ML design decisions on downstream online hybrid simulation. Our work rectifies the sampling issue via the construction of a semi‐automated, end‐to‐end pipeline for size ensembles of hybrid simulations, revealing important nuances in how systematic reductions in offline error manifest in changes to online error and online stability. For example, removing dropout and switching from a Mean Squared Error to a Mean Absolute Error loss both reduce offline error, but they have opposite effects on online error and online stability. Other design decisions, like incorporating memory, converting moisture input from specific humidity to relative humidity, using batch normalization, and training on multiple climates do not come with any such compromises. Finally, we show that ensemble sizes of may be necessary to reliably detect causally relevant differences online. By enabling rapid online experimentation at scale, we can empirically settle debates regarding subgrid ML parameterization design that would have otherwise remained unresolved in the noise.

Lin, Jerry [Department of Earth System Sciences Un↗

DOE OSTI2024

An Evaluation of The Dynamic Physical Security Risk Assessment Methodology for Fleet-Wide Applications

The requirements for U.S. nuclear power plants to maintain a large onsite physical security force contribute to their high operational costs. The cost of maintaining the current physical security posture is approximately 10% of the overall operation and maintenance budget for commercial nuclear power plants. The goal of the Light Water Reactor Sustainability (LWRS) program’s physical security pathway is to develop tools, methods, and technologies and provide the technical basis for an optimized physical security posture. The conservatisms built into current security postures may be analyzed and minimized to reduce security costs while still ensuring adequate security and operational safety. The research performed at Idaho National Laboratory within LWRS program’s physical security pathway has successfully developed a dynamic force-on-force modeling framework using various computer simulation tools and integrating them with the dynamic assessment Event Modeling Risk Assessment using Linked Diagrams (EMRALD) tool. This integrated process for physical security analysis is named Modeling and Analysis for Safety Security using Dynamic EMRALD Framework (MASS-DEF). This document provides an update on the progress in applying the MASS-DEF process to an operating commercial nuclear power plant as well as additional industry feedback regarding use of the tool for other physical security risk-informed topics. This report is only a summary of the progress and does not contain specific modeling results as those contain sensitive security information. Previous reports described how a user could integrate their plant-specific force-on-force models with the dynamic simulation tool EMRALD, model operator actions, and integrate with probabilistic risk assessment tools, such as CAFTA (Computer Aided Fault Tree Analysis System) or SAPHIRE (Systems Analysis Programs for Hands-on Integrated Reliability Evaluations), and with thermal-hydraulic tools, such as RELAP-5 or MAAP. Previous reports applied various combinations of available simulations codes with EMRALD using generic plant models to demonstrate how to perform the analysis. This report is an update the progress of applying the dynamic computational framework to an actual nuclear facility using their security scenarios and timelines. This report also provides an update to the procedural guidance for the MASS-DEF process and an overview of the generic models available for use by utilities. This report does not contain any plant’s sensitive information and/or safeguards information. This study’s purpose was to verify that the results achieved using generic models are similar to actual plant results and refine our guidance on the use of the framework. This assessment enables further analysis, such as what-if scenarios and staff-reduction evaluation, thereby optimizing physical security at plants.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2025

A control-inspired approach for energy transition planning under uncertainty

As the global carbon footprint continues to grow, many countries are implementing carbon emission reduction policies which have incentivized the expansion of low-carbon and renewable technologies. However, the speed and scale of deployment falls short of that needed to meet climate goals. Energy system models serve as key tools for guiding investment decisions and helping policymakers evaluate the effects of various policies on the development of an energy system. This study focuses on the energy system of the United States and builds upon prior work by incorporating more geographic granularity to account for the trade of commodities and addresses transmission congestion through electricity price adjustments. Furthermore, real-world characteristics, such as delays in constructing new liquid fuel production and electricity generation facilities, are integrated using a sequential decision-making approach that better reflects how decisions can be updated as uncertainties unfold. Results demonstrate that stochastic programming combined with sequential decision-making produces energy transition pathways that are robust to multiple uncertain futures. Additionally, considering real-world characteristics significantly impacts the deployment of renewable technologies and the ability to meet carbon emission reduction goals while also reliably meeting demand. These findings highlight the importance of accounting for uncertainty and real-world characteristics to avoid overly optimistic projections in energy system planning.

energy systems↗

DOE OSTI2025

Digital Twin Framework for PIP-II Linac: AI-Driven Multi-Scale Modeling from Ion Source to 800 MeV

The PIP-II linac will enable >1.2 MW beam power for DUNE, requiring unprecedented operational reliability across its warm front-end (RFQ, MEBT) and five distinct SRF sections operating at 162.5/325/650 MHz. We present a comprehensive digital twin framework uniquely combining a fully differentiable fast beam transport code with neural network surrogates trained on high-fidelity PIC simulations, capturing space charge and nonlinear dynamics beyond traditional envelope codes while achieving 10⁴ speedup at <1% accuracy. End-to-end differentiability enables gradient-based optimization across 500+ parameters simultaneously previously impossible with conventional tools while the model incorporates static/dynamic errors and serves as a virtual commissioning platform for diverse hardware integration. The framework facilitates reinforcement learning for pulsed/CW mode transitions, predictive maintenance through anomaly detection, and autonomous tuning algorithm development with real-time execution capability. Validation against physics simulations shows excellent agreement for the front-end, with initial results demonstrating potential for 30% commissioning time reduction and proactive fault mitigation, providing a scalable blueprint for operating next-generation high-intensity accelerators.

Pathak, Abhishek [Fermilab] (ORCID:000000021704208↗

DOE OSTI2025

Digital Twin Framework for PIP-II Linac: AI-Driven Multi-Scale Modeling from Ion Source to 800 MeV

The PIP-II linac will enable >1.2 MW beam power for DUNE, requiring unprecedented operational reliability across its warm front-end (RFQ, MEBT) and five distinct SRF sections operating at 162.5/325/650 MHz. We present a comprehensive digital twin framework uniquely combining a fully differentiable fast beam transport code with neural network surrogates trained on high-fidelity PIC simulations, capturing space charge and nonlinear dynamics beyond traditional envelope codes while achieving 10⁴× speedup at <1% accuracy. End-to-end differentiability enables gradient-based optimization across 500+ parameters simultaneously—previously impossible with conventional tools—while the model incorporates static/dynamic errors and serves as a virtual commissioning platform for diverse hardware integration. The framework facilitates reinforcement learning for pulsed/CW mode transitions, predictive maintenance through anomaly detection, and autonomous tuning algorithm development with real-time execution capability. Validation against physics simulations shows excellent agreement for the front-end, with initial results demonstrating potential for 30% commissioning time reduction and proactive fault mitigation, providing a scalable blueprint for operating next-generation high-intensity accelerators.

Pathak, Abhishek [Fermilab] (ORCID:000000021704208↗

DOE OSTI2026

LAF-Net: A Deep Residual and Cross-Attention Framework for Day-Ahead Load Forecasting: Preprint

Accurate day-ahead load forecasting is essential for reliable power system operations and market efficiency. System operators such as the Midcontinent Independent System Operator (MISO) rely on forecasts from multiple vendors, yet combining them effectively remains a persistent challenge due to vendor-specific biases. This paper presents a novel LSTM-Attention Fusion Network with Error Representation (LAF-Net) that enhances day-ahead hourly load forecasting through deep residual learning and multi-modal cross-attention. The proposed model builds a historical error memory from past vendor performance and dynamically queries it with future hour context to generate adaptive, hour-specific trust weights for each vendor. A bounded residual correction further refines forecasts by mitigating systematic and temporally localized errors. Tested on real MISO LBA data with multi-vendor forecasts, LAF-Net consistently outperforms the best vendor baseline across all 38 LBAs, achieving more than a 40% reduction in system-level mean absolute error (MAE) during peak load hours relative to the best vendor baseline.

24 POWER TRANSMISSION AND DISTRIBUTION↗