Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “embedding model”

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 145 records · Page 8
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Molecular property prediction for very large databases with natural language processing: a case study in ionic liquid design

The prospect of using artificial intelligence (AI) to accurately screen very large databases of compounds for multiple properties has yet to be realized. Here, we explore this possibility using ionic liquids (ILs) which offer unique physicochemical properties and excellent tunability, making them highly versatile solvents for various research applications. Screening millions of potential ILs for the best perfomance for use in specific tasks with experimental methods alone however, is impractical. Further, traditional’ physics-based computational chemistry is hindered by high computational cost. To address this challenge, we leverage a natural language processing (NLP)-based molecular embedding technique with advanced machine learning (ML) models to predict seven key IL properties: viscosity, density, ionic conductivity, surface tension, melting temperature, toxicity, and water solubility. Comprehensive datasets for these properties are obtained, then NLP featurization with Mol2vec is compared with other featurization techniques such as 2D Morgan fingerprints, and 3D quantum chemistry-derived sigma profiles. NLP-based featurization exhibited the best predictive performance, achieving the highest R 2 and lowest RMSE values for all the studied IL properties. Further, we present case studies of how ILs might be screened using combined property criteria for practical cases – lignocellulosic biomass processing, CO 2 capture, and optimal electrolytes for batteries – screening a novel database of ∼10.6 million generated feasible ILs. The results introduce NLP as a powerful tool for engineering many designer solvents with desirable properties for task specific applications.

Mohan, Mood [Oak Ridge National Laboratory (ORNL),↗
DOE OSTI2024

Stau pairs from natural SUSY at high luminosity LHC

Natural supersymmetry (SUSY) with light Higgsinos is perhaps the most plausible of all weak scale SUSY models while a variety of motivations point to (right) tau sleptons as the lightest of all the sleptons. We examine a SUSY model line with rather light right staus embedded within natural SUSY. For light τ ˜ 1 of a few hundred GeV, the decays τ ˜ 1 → τ χ ˜ 1 , 2 0 and ν τ χ ˜ 1 − occur at comparable rates where the (Higgsino-like) χ ˜ 1 ± and χ ˜ 2 0 release only small visible energy: in this case, the expected τ + τ − + E T signature is diminished from the usual expectations due to the presence of the nearly invisible decay mode τ ˜ 1 → ν τ χ ˜ 1 − . However, once m τ ˜ 1 ≳ m ( b i n o ) , decays to binos such as τ ˜ 1 → τ χ ˜ 3 0 open up where χ ˜ 3 0 decays to Higgsinos plus W ± , Z 0 , and h at comparable rates. For these heavier staus, the stau pair production gives rise to diboson + E T events, which may contain 0, 1, or 2 additional hard τ leptons. From these considerations, we examine the potential for future discovery of tau-slepton pair production at a high-luminosity LHC. While we do not find a 5 σ HL-LHC discovery reach for 3000 fb − 1 , we do find a 95% CL exclusion reach, ranging between m τ ˜ 1 : 100 – 450 GeV for m χ ˜ 1 0 ∼ 100 GeV . This latter reach disappears for m χ ˜ 1 0 ≳ 200 GeV . Published by the American Physical Society 2024

Astronomy & Astrophysics↗
DOE OSTI2024

Dynamic Graph Sequence Data from Simulated Neutron Reflectometry Measurements

This dataset comprises dynamic graph sequences derived from simulated in-situ neutron reflectometry measurements, capturing the gradual evolution of a layer structure over time. Each graph sequence represents a synthetic sample, with node features detailing the scattering vector and corresponding reflectivity measurements, while adjacency matrices have corresponding reference material parameters attached as metadata. The dataset spans multiple sets, each with a different number of sequences, offering a comprehensive basis for training models that handle dynamic input sequences with embedded physics. This dataset is particularly suited for tackling inverse problems with hidden physical states that evolve over time, challenges that are typically difficult to address using conventional iterative fitting methods.

36 MATERIALS SCIENCE↗
DOE OSTI2026

XMark: Reliable Multi-Bit Watermarking for LLM-Generated Texts

Multi-bit watermarking has emerged as a promising solution for embedding imperceptible binary messages into Large Language Model (LLM)-generated text, enabling reliable attribution and tracing of malicious usage of LLMs. Despite recent progress, existing methods still face key limitations: some become computationally infeasible for large messages, while others suffer from a poor trade-off between text quality and decoding accuracy. Moreover, the decoding accuracy of existing methods drops significantly when the number of tokens in the generated text is limited, a condition that frequently arises in practical usage. To address these challenges, we propose XMark, a novel method for encoding and decoding binary messages in LLM-generated texts. The unique design of XMark’s encoder produces a less distorted logit distribution for watermarked token generation, preserving text quality, and also enables its tailored decoder to reliably recover the encoded message with limited tokens. Extensive experiments across diverse downstream tasks show that XMark significantly improves decoding accuracy while preserving the quality of watermarked text, outperforming prior methods. The code will be made publicly available upon acceptance.

Xu, Jiahao [University of Nevada, Reno]↗
DOE OSTI2024

MULTI-LEADER: MULTI-source LEarning-Accelerated Design of high-Efficiency multi-stage compRessor (Final Technical Report)

The objective of MULTI-LEADER is to cut design costs by 80% while generating more energy-efficient designs of multi-stage compressors by developing and implementing novel machine learning (ML) techniques, which enable faster and fewer design iterations, improved solver performance, and concurrent multi-disciplinary design. Current industrial practices for the design of multi-stage compressors involve simulation-based design optimization with successive levels of model fidelity, iteratively evaluated between distinct disciplines, one stage at a time to tackle the high dimensional design variations. This project addresses these key design challenges: (1) concurrent optimization of multiple stages under many non-linear constraints; (2) multitude of evaluation of high-fidelity and expensive solvers and their gradients during optimization convergence in high-dimensional design; (3) multi-disciplinary design to maximize aerodynamic performance while guaranteeing structural integrity and additive manufacturability; (4) utilization of multiple fidelity of solvers with disparate parameterization and modeling assumptions. MULTI-LEADER achieved more than 5x speed up in detailed design of more energy-efficient compressors via these machine learning (ML) innovations: (i) rapid design surrogates by multi-source learning from diverse fidelities across multiple disciplines, (ii) physics-constrained data-augmented modeling for improved empiricism, (iii) generative manifold embedding for high dimensional concurrent design without gradient information; (iv) budget-constrained fidelity-adaptive sampling towards fewer design iterations.

33 ADVANCED PROPULSION SYSTEMS↗
DOE OSTI2024

Axion domain walls, small instantons, and non-invertible symmetry breaking

Non-invertible global symmetry often predicts degeneracy in axion potentials and carries important information about the global form of the gauge group. When these symmetries are spontaneously broken they can lead to the formation of stable axion domain wall networks which support topological degrees of freedom on their worldvolume. Such non-invertible symmetries can be broken by embedding into appropriate larger UV gauge groups where small instanton contributions lift the vacuum degeneracy, and provide a possible solution to the domain wall problem. We explain these ideas in simple illustrative examples and then apply them to the Standard Model, whose gauge algebra and matter content are consistent with several possible global structures. Each possible global structure leads to different selection rules on the axion couplings, and various UV completions of the Standard Model lead to more specific relations. As a proof of principle, we also present an example of a UV embedding of the Standard Model which can solve the axion domain wall problem. The formation and annihilation of the long-lived axion domain walls can lead to observables, such as gravitational wave signals. Observing such signals, in combination with the axion coupling measurements, can provide valuable insight into the global structure of the Standard Model, as well as its UV completion.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2025

Modeling Offshore Wind Farm Performance in Coastal Low-Level Jets Using Coupled Mesoscale-Microscale Large Eddy Simulations

Accurately predicting wind farm reliability under complex offshore atmospheric conditions remains a key challenge, particularly during noncanonical meteorological events such as coastal low-level jets (LLJs). LLJs, characterized by strong nonmonotonic vertical shear and directional veer, depart significantly from the simplified inflow assumptions embedded in conventional design standards, low-fidelity engineering models, and microscale large eddy simulations of the atmospheric boundary layer. In this work, we use the virtual wind farm framework—an exascale, graphics processing unit–accelerated large eddy simulation platform coupled with high-fidelity aeroservoelastic turbine models and advanced mesoscale-microscale coupling via the ExaWind software stack—to investigate turbine responses under realistic LLJ forcing. Simulations are performed over the U.S. North Atlantic offshore domain with the use of meteorological inputs from New York State Energy Research and Development Authority buoy data, focusing on a representative LLJ case impacting the International Energy Agency 15 MW reference turbine. Our results show that LLJs can cause up to 50% power deficits in downstream turbine rows and significantly amplify low-speed shaft and tower loads through nonlinear coupling between complex inflow characteristics and turbine structural dynamics. Two primary mechanisms drive these load amplifications: (1) unique LLJ inflow features—including veer and vertical/lateral shear—and (2) the downstream evolution of the flow under stable thermal stratification, which suppresses turbulence mixing and alters wake recovery. These mechanisms produce streamwise variations in turbine loading not captured by standard hub height–based metrics or existing design load case (DLC) definitions. This study highlights the critical role of rotor-scale flow gradients in driving fatigue and system-level aeroelastic responses, challenging current DLC and control strategies. We advocate the integration of full-flow field, environment-aware wind inputs into load modeling and control algorithms. By leveraging exascale computing to resolve mesoscale-microscale coupling, this work lays the groundwork for next-generation offshore wind turbine design and operation in meteorologically complex marine environments.

17 WIND ENERGY↗
DOE OSTI2025

Transfer learning nonlinear plasma dynamic transitions in low dimensional embeddings via deep neural networks

Deep learning algorithms provide a new paradigm to study high-dimensional dynamical behaviors, such as those in fusion plasma systems. Development of novel, data-driven model reduction methods, coupled with detection of abnormal modes with plasma physics, opens a unique opportunity to identify plasma instabilities through automated construction of parsimonious models that can be tuned to balance accuracy and cost. Our fusion transfer learning (FTL) model demonstrates success in rapidly reconstructing nonlinear kink mode structures by learning from a limited amount of nonlinear simulation data. The knowledge transfer process leverages a pre-trained neural encoder–decoder network, initially trained on linear simulations, to effectively capture nonlinear dynamics. The low-dimensional embeddings extract the coherent structures of interest, while preserving the inherent dynamics of the complex system. Experimental results highlight FTL’s capacity to capture transitional behaviors and dynamical features in plasma dynamics—a task often challenging for conventional methods. The model developed in this study is generalizable and can be extended broadly through transfer learning to address various magnetohydrodynamics modes.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2026

Finite deformation implementation of a mixed-mode single-integral type cohesive zone with reorienting surfaces of separation

To model material ductile failure and crack propagation, cohesive zone elements can be embedded along potential fracture paths in a finite element simulation. When damage criteria are met, elements in the mesh decohere, simulating the formation and propagation of a crack. In this paper, we present a novel computational algorithm based on finite deformation theory, essential to modeling crack initiation and growth in solids undergoing large deformations. This new algorithm was formulated within a Lagrangian frame of reference to extend previous cohesive zone algorithms to include modeling crack growth in finite deformation contexts. The local coordinate system, necessary for defining an embedded cohesive zone, is constructed based upon the current configuration and is updated within the nonlinear iteration process, thereby resulting in the convergence of the solution for a growing crack in a large deformation quasi-static setting. The model’s accuracy was demonstrated by comparing finite element model simulation results with the analytic case of a constant surface separation, as shown in the verification examples. The power and efficacy of the algorithm to capture large deformations during crack growth were then demonstrated with a double cantilever beam example case. It indicates that the model can be applied to a variety of physical circumstances for predicting crack initiation and growth with delamination and fracture.

42 ENGINEERING↗
DOE OSTI2026

Cooperative effect of local active stresses on the macroscopic contractility of elastic fiber networks

The collective action of actively contractile units embedded in elastic biopolymer networks plays a crucial role in regulating the network's macroscopic mechanical response. Here, in this study, we investigate how the macroscopic boundary stress in model elastic fiber networks depends on the number and nature of embedded contractile units, each exerting an isotropic force dipole, as well as on the bending stiffness of fibers. We find that the macroscopic stress increases nonlinearly with the number of dipoles due to mutual stiffening of initially soft, bending-dominated networks. Using effective medium theory, we relate this enhanced contractility to an increase in the effective average network coordination number due to constraints imposed by the force dipoles. By comparing three distinct force dipole models that differ in their local structures, we demonstrate that the specific manner in which an active unit constrains the network strongly influences the onset and nature of the stiffening transition. Our results highlight that not only the quantity but also the local geometry of force-generating units critically determines the macroscopic mechanical behavior. This framework provides a physical basis for understanding how biological systems—such as molecular motors in the cytoskeleton, or adherent cells in the extracellular matrix—can modulate network-scale nonlinear elastic properties through local tuning of active force-generating units.

Biological and medical sciences↗
DOE OSTI2025

Data-Driven Analysis of Multipactor Dynamics via Dynamic Mode Decomposition

Multipactor effect is a performance-limiting kinetic plasma effect that can occur in high-power microwave and radio frequency (RF) devices. Multipactor effect is of special concern in vacuum or near-vacuum conditions such as those in particle accelerators and spaceborne devices. In this work, we present a data-driven reduced-order model (ROM) based on dynamic mode decomposition (DMD) for modeling of multipactor effects. We study multipactor effects and the resulting nonlinear harmonic generation by processing high-fidelity data generated from electromagnetic particle-in-cell (EMPIC) simulations using the DMD algorithm. We also investigate time-delay embedding extensions of DMD with improved generalizability and accuracy for modeling the electron plasma current density behavior. Here, the results show that DMD provides valuable insights into multipactor phenomena by extracting relevant modal spatiotemporal patterns and frequencies. In addition, DMD offers the potential to time extrapolate EMPIC simulations at a minimal cost, thereby reducing overall simulation time.

43 PARTICLE ACCELERATORS↗
DOE OSTI2026

Netload Range Cost Curves for Coordinated Transmission-Distribution Planning Under DER Growth Uncertainty

The increasing penetration of distributed energy resources (DERs) requires better coordination between transmission and distribution (T&D) planning to ensure system security and cost efficiency. However, misaligned planning horizons, computational burdens, and privacy concerns hinder effective coordination, leading to either underutilized resources caused by overinvestments or reliability risks due to underinvestment. To address this challenge, we introduce netload range cost curves (NRCCs), a novel approach for managing long-term DER growth uncertainty through T&D coordination, while preserving existing data-sharing and regulatory structures. NRCCs provide pairs of (i) peak substation netload guarantees and (ii) corresponding distribution upgrade options and costs, enabling their seamless integration into transmission planning workflows. To compute NRCCs efficiently, we develop a transmission-aware distribution network planning (TADNP), which is subsequently integrated to an iterative computation procedure. These NRCCs are then embedded into an NRCC-informed transmission planning model to enable resource-efficient coordination. We illustrate our proposed approach with a case study based on realistic distribution and transmission systems in the San Francisco Bay Area, California. Our results indicate the possibility of dramatic savings in transmission investments by incorporating the proposed NRCC-integrated T&D coordination framework.

Li, Yujia↗
DOE OSTI2026

Describing Point Defect Topology in 2D Energy Materials Through Computer Vision

Point defects such as vacancies and impurity atoms strongly impact the performance of 2D materials. Traditional efforts often rely on manual detection, a process that is time-intensive, prone to human error, and challenging to scale. Here we leverage machine learning (ML) methods to identify and quantify vacancies within 2D transition metal carbides (Ti3C2, MXenes), aiming to expedite detection while improving accuracy. MXenes exhibit valuable defect-defined electrochemical properties, but we currently lack statistical understanding of defect topology needed to fully harness these materials. Here we employ a convolutional neural network for semantic segmentation of experimental MXene images, opening an opportunity to conduct a rigorous statistical study on defect hierarchy while investigating local relaxation in the lattice. We show how the integration of ML can yield fundamental insight into point defects, providing a powerful tool that will play an increasingly crucial role in the future of materials science. ML is often not just a matter of straightforward application, and pretrained models proved ineffective in this case. Instead, we trained our own neural network (NN) and applied data augmentation techniques and fine-tuning to the training dataset. Since labeled microscopy data is often scarce, we developed training data from a previously published wide-frame MXene image, using customized Gaussian fitting to locate atomic positions. Our trained model was then applied to a large dataset of experimental images, enabling a statistical study of defect configurations across three samples prepared with different HF etchant concentrations (5%, 9.1%, and 12.5%), as shown in Fig. 1. This also allowed us to investigate local strain around vacancies, though we find that we are limited by the precision of measurements using high-angle annular dark field (HAADF) images, as shown in Fig. 2. This study demonstrates how ML enables large-scale, quantitative analysis of atomic defects - an otherwise infeasible task with traditional methods. While our NN was specialized for Ti3C2 MXenes, the pipeline we developed provides a foundation for future ML models tailored to other materials. Ultimately, we envision embedding the NN onto the microscope to give real-time feedback to the user. To make this a reality, continued work is necessary to fully understand the NN's capabilities and limitations. This study gets one step closer to our goals of automated experimentation moving away from traditional methods of manual labeling. As ML capabilities advance, we hope to continue adapting and applying these techniques in microscopy.

2D materials↗
DOE OSTI2025

DriveSense: A Noise-Resilient Framework for Driving Mode Identification

Accurate drive mode classification is essential for enhancing the reliability and predictive maintenance of heavy-duty electric trucks. This study proposes a novel fuzzy logic-based framework, DriveSense, for real-time drive mode classification, addressing key challenges such as sensor noise, transitional behaviors, and computational efficiency. The proposed approach integrates a two-stage filtering pipeline, combining adaptive outlier removal and a dynamic Kalman filter to enhance data quality. A fuzzy inference system with smoothened trapezoidal membership functions is then applied to classify driving modes into standstill, constant speed, acceleration, and deceleration while mitigating the effects of noise and edge cases. Performance evaluation using real-world and simulated drive cycles demonstrates significant improvements in classification accuracy (up to 97.8%), F1-score (up to 0.97), and robustness against noise, while reducing false positives. Comparative analysis against baseline models, demonstrates DriveSense’s superior accuracy and generalizability across diverse driving patterns. The framework’s lightweight and interpretable fuzzy inference engine operates with low computational latency, ensuring compatibility with real-time embedded systems typical of heavy-duty electric trucks. Moreover, DriveSense models transitional behaviors through overlapping fuzzy sets and adaptive borderline classification logic, enabling smooth identification of subtle shifts such as rolling stops or gradual deceleration. These results highlight DriveSense’s potential to enhance predictive maintenance strategies, reduce downtime, and support scalable, fleet-wide diagnostics.

Kumar, Praveen [Oak Ridge National Laboratory (ORN↗
DOE OSTI2026

Quantum computing approach for building surface sunlit in urban-scale energy modeling

Solar shadow calculations are needed in building energy modeling and performance simulation of PV systems installed on roofs or facades of buildings. We present a quantum computing approach for calculation of building surface sunlit fractions by recasting solar visibility as a binary optimization problem solved by quantum annealing. Each triangulated surface centroid is encoded as a binary qubit indicating sunlit or shaded status. Geometric visibility constraints are derived from the Möller-Trumbore intersection algorithm and converted into a constrained quadratic binary model compatible with contemporary quantum annealers. The coefficients were embedded to D-Wave quantum computer. To demonstrate feasibility, we conducted a case study in San Francisco for a target building with 52 triangles and roughly 2700 nearby triangles within 50 m evaluated at representative winter and summer solar positions. The results demonstrated that quantum annealing can reliably calculate and distinguish sunlit from shaded surfaces. Quantum samples achieved average accuracy exceeding 92.4 %, with the aggregate surface-level agreement approaching 99.9 %. The outputs of quantum computers agreed closely with classical algorithms, indicating practical feasibility and promising scalability. Finally, the hourly sunlit fractions of building surfaces can be obtained for urban energy modelling. This is the first study to apply quantum computing to the solar shadow and building surface sunlit calculation. It introduces a new paradigm that differs fundamentally from traditional approaches.

Deng, Zhipeng↗
DOE OSTI2026

Satellite Embedding-Based Population Imputation for Areas with Missing Building Footprint Data: A Computer Vision-Based Approach

High-resolution population modeling is important for supporting effective decision-making across diverse sectors. LandScan Mosaic generates population estimates at the level of individual buildings and aggregates them to 3 arc-second grids, and this approach performs well in regions where building footprint data are comprehensive and reliable. However, large portions of the globe still suffer from incomplete, sparse, or entirely missing building stock datasets, creating a structural limitation for strictly building-based population models. To address this research gap, this study proposes a computer vision-based framework that employs Google Earth Engine satellite embeddings and UNet, which allows us to directly impute grid-level population estimates in building-data-deficient areas. Applied to Taiwan as a case study, the framework achieved strong predictive performance with R$^{2}$ of 0.89, RMSE of 18.70, and MAE of 8.41, outperforming traditional machine learning approaches. Notably, the proposed framework effectively addressed building false-positive errors inherent in Global Human Settlement Layer (GHSL) data, correctly identifying uninhabited areas that were erroneously classified as populated. The framework also offers significant advantages for global population mapping, particularly in terms of scalability and temporal consistency, thereby extending the coverage and accuracy of high-resolution population products in data-scarce regions worldwide. Urban planners, decision makers, and related stakeholders can obtain granular population distributions to support more accurate and targeted infrastructure investment, service delivery, resource allocation, and risk assessment decisions.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

PRIME: An evaluation framework for protein representation inference and generalization in viral mutation space

Background Protein language models (PLMs) have revolutionized protein fitness prediction, yet their application to rapidly evolving viral pathogens is often confounded by extreme sequence homology. This homology leads to “data leakage” in standard random validation splits, yielding inflated performance metrics that fail to translate into real-world biosurveillance utility. Results We present Protein Representation Inference for Mutation Evaluation (PRIME), a framework that integrates domain-specific fine-tuning with a rigorous position-stratified validation protocol to evaluate viral threats. Using a dataset of 347,432 SARS-CoV-2 receptor binding domain (RBD) sequences, we demonstrate that while random training data split yields deceptive R 2 values (> 0.90), they fail to generalize to novel mutational sites. By benchmarking models up to 650 M parameters, we show that domain-specific fine-tuning of the ESM-C 600 M model with correctly stratified data provides an initial demonstration of predictive signal for binding affinity and expression at unseen mutational sites of binding affinity and expression on unseen sites (R 2 ~0.23), a significant advancement over base foundation models which exhibit no predictive power (R 2 <0). PRIME’s embedding-based clustering identified 3.03% of bat coronavirus sequences as candidates for further experimental prioritization based on their functional similarity to human-infective strains in embedding space, offering a perspective complementary to traditional phylogenetic methods. Conclusion PRIME establishes a new benchmark for the application of PLMs in pathogen surveillance. Our findings demonstrate that state-of-the-art models and fine-tuning, when paired with stratified validation, provide biologically meaningful insights into pathogen evolution and zoonotic risk.

59 BASIC BIOLOGICAL SCIENCES↗