Search NASA⌕ Search

SEARCH · Search NASA

Results for “operator networks”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Physics-informed latent neural operator for real-time predictions of time-dependent parametric PDEs

Deep operator network (DeepONet) has shown significant promise as surrogate models for systems governed by partial differential equations (PDEs), enabling accurate mappings between infinite-dimensional function spaces. However, when applied to systems with high-dimensional input-output mappings arising from large numbers of spatial and temporal collocation points, these models often require heavily overparameterized networks, leading to long training times. Latent DeepONet addresses some of these challenges by introducing a two-step approach: first learning a reduced latent space using a separate model, followed by operator learning within this latent space. While efficient, this method is inherently data-driven and lacks mechanisms for incorporating physical laws, limiting its robustness and generalizability in data-scarce settings. Here, in this work, we propose PI-Latent-NO, a physics-informed latent neural operator framework that integrates governing physics directly into the learning process. Our architecture features two coupled DeepONets trained end-to-end: a Latent-DeepONet that learns a low-dimensional representation of the solution, and a Reconstruction-DeepONet that maps this latent representation back to the physical space. By embedding PDE constraints into the training via automatic differentiation, our method eliminates the need for labeled training data and ensures physics-consistent predictions. The proposed framework is both memory and compute-efficient, exhibiting near-constant scaling with problem size and demonstrating significant speedups over traditional physics-informed operator models. We validate our approach on a range of parametric PDEs, showcasing its accuracy, scalability, and suitability for real-time prediction in complex physical systems.

Latent representations↗
DOE OSTI2026

Throughput Estimation of Data Transport Networks From Digital Twin Measurements

Digital twins of networked infrastructures, known as Virtual Infrastructure Twins (VITs), are increasingly used for software development, pre-deployment testing, and design space exploration. While VITs avoid the costs and potential disruptions associated with experiments on operational networks, their throughput measurements are typically not sufficiently accurate for performance profiling of wide-area networks that they emulate. Here, machine learning (ML) methods are developed to transform these inaccurate VIT network throughput measurements to closely match in peak and overall profile of those from a physical testbed or production network. First, a micro kernel network reflecting a physical network is utilized to collect one-time measurements on a host to support this ML transformation. Then, a generic multi-modal ML method is developed to learn a map that transforms measurements from subsequent VITs on the same host to match past, current and follow-on testbed and cloud networks. ML generalization equations are derived to establish its correctness and probabilistically guarantee its generalization accuracy. Experimental results are presented for a variety of VIT hosts with target testbed and cloud networks; they include a case study of a four-site science ecosystem wherein inaccurate convex VIT measurement profiles are transformed into accurate concave profiles of target networks.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Efficient Reformulation and Optimization for SC-ACOPF with Line Switching

This project aims to develop efficient and robust computational methods for solving the security-constrained alternating current optimal power flow problem (SC-ACOPF). The SC-ACOPF problem is a central problem in operating the electric power grids in the United States. It determines the most economically efficient way to operate the generation and transmission system to meet daily electricity demand. The solution found by solving an SC-ACOPF problem must satisfy the physics of the alternating current (AC) power flows, various generator and network operational constraints, and must maintain secure operation under various contingency scenarios, where a generator, a transmission branch, or a transformer may unexpectedly trip offline.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Scale-bridging Optimization Framework for Desalination Integrated Produced Water Networks

Produced water (PW) management poses a major challenge to U.S. oil and gas development. Due to freshwater scarcity and PW injection curtailments, desalination may become a necessity in some parts of the country. Integrating rigorous desalination process models within a produced water network for co-optimization of desalination design and network operation can be challenging to scale-up due to computational complexity. In this work we use the trust region filter framework to decompose the integrated optimization problem into an optimization master problem consisting of the network model and simplified surrogates for the desalination units and an optimization sub problem for the rigorous desalination unit model. The approach is demonstrated on a multiperiod produced water network from the PARETO library with thermal desalination units.

Naik, Sakshi↗
DOE OSTI2026

Latent Twins

Over the past decade, scientific machine learning has transformed the development of mathematical and computational frameworks for analyzing, modeling, and predicting complex systems. From inverse problems to numerical partial differential equations (PDEs), dynamical systems, and model reduction, these advances have pushed the boundaries of what can be simulated. Yet they have often progressed in parallel, with representation learning and algorithmic solution methods evolving largely as separate pipelines. With Latent Twins, we propose a unifying mathematical framework that creates a hidden surrogate in latent space for the underlying equations. Whereas digital twins mirror physical systems in the digital world, Latent Twins mirror mathematical systems in a learned latent space governed by operators. Through this lens, classical modeling, inversion, model reduction, and operator approximation all emerge as special cases of a single principle. We establish the fundamental approximation properties of Latent Twins for both ordinary differential equations (ODEs) and PDEs and demonstrate the framework across three representative settings: (i) canonical ODEs, capturing diverse dynamical regimes; (ii) a PDE benchmark using the shallow-water equations, contrasting Latent Twin simulations with deep operator network and forecasts with a four-dimensional variational method baseline; and (iii) a challenging real-data geopotential reanalysis dataset, reconstructing and forecasting from sparse, noisy observations. Latent Twins provide a compact, interpretable surrogate for solution operators that evaluate across arbitrary time gaps in a single-shot, while remaining compatible with scientific pipelines such as assimilation, control, and uncertainty quantification. Looking forward, this framework offers scalable, theory-grounded surrogates that bridge data-driven representation learning and classical scientific modeling across disciplines.

Latent Twins↗
DOE OSTI2025

Synergistic learning with multi-task DeepONet for efficient PDE problem solving

Multi-task learning (MTL) is an inductive transfer mechanism designed to leverage useful information from multiple tasks to improve generalization performance compared to single-task learning. It has been extensively explored in traditional machine learning to address issues such as data sparsity and overfitting in neural networks. In this work, we apply MTL to problems in science and engineering governed by partial differential equations (PDEs). However, implementing MTL in this context is complex, as it requires task-specific modifications to accommodate various scenarios representing different physical processes. To this end, we present a multi-task deep operator network (MT-DeepONet) to learn solutions across various functional forms of source terms in a PDE and multiple geometries in a single concurrent training session. We introduce modifications in the branch network of the vanilla DeepONet to account for various functional forms of a parameterized coefficient in a PDE. Additionally, we handle parameterized geometries by introducing a binary mask in the branch network and incorporating it into the loss term to improve convergence and generalization to new geometry tasks. Our approach is demonstrated on three benchmark problems: (1) learning different functional forms of the source term in the Fisher equation; (2) learning multiple geometries in a 2D Darcy Flow problem and showcasing better transfer learning capabilities to new geometries; and (3) learning 3D parameterized geometries for a heat transfer problem and demonstrate the ability to predict on new but similar geometries. Finally, our MT-DeepONet framework offers a novel approach to solving PDE problems in engineering and science under a unified umbrella based on synergistic learning that reduces the overall training cost for neural operators.

42 ENGINEERING↗
DOE OSTI2024

RandONets: Shallow networks with random projections for learning linear and nonlinear operators

Deep neural networks have been extensively used for the solution of both the forward and the inverse problem for dynamical systems. However, their implementation necessitates optimizing a high-dimensional space of parameters and hyperparameters. This fact, along with the requirement of substantial computational resources, pose a barrier to achieving high numerical accuracy, but also interpretability. Here, to address the above challenges, we present Random Projection-based Operator Networks (RandONets): shallow networks with random projections and tailor-made numerical analysis methods that learn accurately and fast linear and nonlinear operators. Building on previous works, we prove that RandOnets are universal approximators of linear and nonlinear operators. Due to their simplicity, RandONets provide a one-step transformation of the input space, facilitating interpretability. For the evaluation of their performance, we focus on operators of PDEs. We show, that RandONets outperform by several orders of magnitude, both in terms of numerical approximation accuracy and computational cost, the “vanilla” DeepONets. Hence, we believe that our method will trigger further developments in the field of scientific machine learning, for the development of new ‘’light”schemes that will provide high accuracy while reducing dramatically the computational cost. A MATLAB toolbox for RandONets, including demos, is available on GitHub at https://github.com/GianlucaFabiani/RandONets.

Interpretable machine learning↗
DOE OSTI2024

First-generation college student forges ahead, now key to Lab's mission

Fatima Woody was just 17 and a student at Pojoaque Valley High School when she first started her career as a Los Alamos Neutron Science Center receptionist. Now she's in a crucial role that keeps plutonium pit production and other mission processes operating with as little interruption as possible. Over nearly four decades, Fatima has gradually advanced from her initial positions as a receptionist and administrative secretary to become a computer technician, then a computer system professional who specializes in project management. Today, she coordinates the workflow for nearly two dozen deployed information technology technicians who keep the computer systems and networks operating at the high-tech complex that houses the Lab's Plutonium Facility.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2026

SPARTAN and IMPROVE Comparison Experiment (SPICE) Interim Campaign Report

SPICE (the SPARTAN and IMPROVE Comparison Experiment) aims to obtain and quantify comparisons between aerosol PM 2.5 mass concentration measurements from the University of Oklahoma (OU) Surface Particulate Matter Network (SPARTAN) station and the U.S. Environmental Protection Agency (EPA) Interagency Monitoring of Protected Visual Environments (IMPROVE) station hosted by the U.S. Department of Energy Atmospheric Radiation Measurement (ARM) User Facility at ARM’s Southern Great Plains (SGP) observatory in Oklahoma. Aerosols—both natural and anthropogenic—affect human populations in multiple ways. Much of ARM’s research focuses on how aerosols influence weather and climate through their optical and radiative effects, as well as their impacts on clouds and precipitation. However, aerosols also pose direct risks to humans and other organisms through inhalation, with the severity of health impacts depending on particle size, chemical composition, and duration of exposure. The IMPROVE network was established by the U.S. EPA to monitor air quality, including visible clarity as well as total and chemically speciated aerosol mass concentrations. The ARM SGP site hosts the IMPROVE SOGP station. Separately, the SPARTAN network operates a globally distributed set of stations similar to IMPROVE but with an emphasis on remote deployment and semi-autonomous operation for use beyond the borders of the United States (IMPROVE only operates within the U.S.). The University of Oklahoma operates a SPARTAN station. To establish confidence in the OU SPARTAN instrumentation and measurement protocol relative to the EPA-certified IMPROVE station, the OU SPARTAN station is currently deployed at SGP in close proximity to the IMPROVE SOGP station. The SPICE campaign was envisioned as a contiguous calendar-year effort for 2025 to capture seasonal variation in mass loading as well as composition. However, independent of the SPICE campaign, the SPARTAN network adopted a new filter construction part-way through the year, interrupting our contiguous data set. Thus, to obtain a contiguous data set with a uniform consistent configuration, SPICE desires an extension through 2026.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

The Urban Deployment Model: A Toolset for the Simulation and Performance Characterization of Radiation Detector Deployments in Urban Environments

Static and mobile radiation detectors can be deployed in urban environments for a range of nuclear security applications, including radiological source search-and-tracking scenarios. Modeling detector performance for such applications is challenging, as it does not depend solely on the detector capabilities themselves. Many factors must be taken into consideration, including specific source and background signatures, the topology and constraints of the deployment environment, the presence of nuisance sources, and whether detectors are mobile or static. When considering the simultaneous deployment of multiple, heterogeneous detectors, assessment of the system-wide performance requires the simulation of the individual detectors, and a system-level analysis of the detection performance. In radiological source search-and-tracking scenarios, performance is mostly dominated by the probability of encounter, which depends on the specifics of a given deployment, e.g., static vs. mobile detectors or a combination of both modalities, the number of detectors deployed, the dynamic vs. static setting of false alarm rates, and individual vs. networked operation. The Urban Deployment Model (UDM) toolset was specifically developed to cover the gap in the available generic frameworks for the simulation of radiation detector deployments at city scales. UDM provides a unified and modular framework to support the simulation and performance characterization of heterogeneous detector deployments in urban environments. This paper presents the key components along the UDM workflow.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗
DOE OSTI2025

From clutter to clarity: Emergent neural operators via questionnaire metrics

Real-world datasets in chemical engineering and bioengineering processes—such as those from catalytic reactors, multiphase flows, polymerization reactors, bioreactors, and clinical trials—can often be unlabeled or disorganized, rendering the training of existing supervised learning models ineffective at learning the underlying dynamics. To salvage these datasets for decision-making, we first seek to obtain clarity from the cluttered data. Here, we present a framework for developing “structural” generative models, discovering emergent equations, and constructing efficient emulators from scrambled datasets by integrating unsupervised organizational learning techniques (Questionnaires) with advanced deep learning architectures (Deep Hidden Physics Models and Deep Operator Networks). Our approach is demonstrated on two illustrative model systems: (a) a 1D advection–diffusion partial differential equation representing a winding underground pipe and (b) an ensemble of Stuart–Landau oscillators, an agent-based system of coupled ordinary differential equations. In both cases, we successfully reconstruct meaningful spatial, temporal, and parameter embeddings from scrambled data, enabling good predictions of system dynamics. As a result, we highlight the framework’s potential for broader applications, enabling data-driven system identification in fields with inherently disorganized or hidden parameter spaces.

42 ENGINEERING↗
DOE OSTI2024

Separable physics-informed DeepONet: Breaking the curse of dimensionality in physics-informed machine learning

The deep operator network (DeepONet) has shown remarkable potential in solving partial differential equations (PDEs) by mapping between infinite-dimensional function spaces using labeled datasets. However, in scenarios lacking labeled data, the physics-informed DeepONet (PI-DeepONet) approach, which utilizes the residual loss of the governing PDE to optimize the network parameters, faces significant computational challenges, particularly due to the curse of dimensionality. This limitation has hindered its application to high-dimensional problems, making even standard 3D spatial with 1D temporal problems computationally prohibitive. Additionally, the computational requirement increases exponentially with the discretization density of the domain. Here, to address these challenges and enhance scalability for high-dimensional PDEs, we introduce the Separable physics-informed DeepONet (Sep-PI-DeepONet). This framework employs a factorization technique, utilizing sub-networks for individual one-dimensional coordinates, thereby reducing the number of forward passes and the size of the Jacobian matrix required for gradient computations. By incorporating forward-mode automatic differentiation (AD), we further optimize computational efficiency, achieving linear scaling of computational cost with discretization density and dimensionality, making our approach highly suitable for high-dimensional PDEs. We demonstrate the effectiveness of Sep-PI-DeepONet through three benchmark PDE models: the viscous Burgers’ equation, Biot’s consolidation theory, and a parameterized heat equation. Our framework maintains accuracy comparable to the conventional PI-DeepONet while reducing training time by two orders of magnitude. Notably, for the heat equation solved as a 4D problem, the conventional PI-DeepONet was computationally infeasible (estimated 289.35 h), while the Sep-PI-DeepONet completed training in just 2.5 h. These results underscore the potential of Sep-PI-DeepONet in efficiently solving complex, high-dimensional PDEs, marking a significant advancement in physics-informed machine learning.

Neural operator↗
DOE OSTI2025

Data-driven emulation of modal aerosol microphysics via neural operator-based modeling

The complexity and the small characteristic scales of aerosol microphysical processes pose a big challenge for accurate and efficient Earth system simulations at regional and global scales. In this work, we construct and evaluate a surrogate model: the aerosol deep operator network (ADON), a physics-inspired dual-net architecture for emulating the aerosol microphysics parameterization suite in the version 2 of the Energy Earth System Model (E3SMv2). The current version of the surrogate model is trained on a dataset comprising 9.8 million samples obtained from a global E3SMv2 simulation with the horizontal resolution of about one degree under cloud-free conditions. Incorporating domain spatial and temporal coordinates, as well as principle components extracted from training data, the dual-net surrogate model effectively captures the intricate representations of aerosol and the relationship with atmospheric state variables, achieving an R-squared score over $$95.7\%$$ for all the lognormal aerosol modes in the extrapolated regime. The validated model provides feature importance of input variables and their impact on the predictive capacity of the surrogate model in relation to the E3SM. The computational cost of online inference time deployed on CPUs and GPUs with lower precisions highlights ADON’s efficiency and potential in robust predictive modeling for large-scale Earth system computations.

Bai, Zhe↗
DOE OSTI2024

Graph neural network for neutrino physics event reconstruction

Liquid argon time projection chamber (LArTPC) detector technology offers a wealth of high-resolution information on particle interactions, and leveraging that information to its full potential requires sophisticated automated reconstruction techniques. Here, this article describes NUGRAPH 2, a graph neural network for low-level reconstruction of simulated neutrino interactions in a LArTPC detector. Simulated neutrino interactions in the MicroBooNE detector geometry are described as heterogeneous graphs, with energy depositions on each detector plane forming nodes on planar subgraphs. The network utilizes a multihead attention message-passing mechanism to perform background filtering and semantic labeling on these graph nodes, identifying those associated with the primary physics interaction with 98.0% efficiency and labeling them according to particle type with 94.9% efficiency. The network operates directly on detector observables across multiple two-dimensional representations but utilizes a three-dimensional-context-aware mechanism to encourage consistency between these representations. Model inference takes 0.12 s / event on a CPU and 0.005 s / event batched on a GPU. This architecture is designed to be a general-purpose solution for particle reconstruction in neutrino physics, with the potential for deployment across a broad range of detector technologies, and offers a core convolution engine that can be leveraged for a variety of tasks beyond the two described in this paper.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Implementation of Detailed Polyethylene Pyrolysis Kinetics into CFD Simulations using Machine Learning

Municipal solid waste (MSW) and waste plastics have received significant attention due to the issues of waste generation and storage, as well as their potential as an energy resource. High-density polyethylene (HDPE) makes up a large portion of plastic waste and has been the subject of several conversion studies. However, the mechanisms associated with converting HDPE through pyrolysis and gasification are extensive and complex making them difficult to implement into high-fidelity computational fluid dynamic (CFD) simulations. For this project, a primary pyrolysis mechanism containing 42 unique species and 737 heterogeneous reactions was used to generate kinetic data over a range of operating conditions. A machine learning (ML) model was developed to replicate the results of the detailed pyrolysis mechanism while significantly increasing the computational efficiency. A deep operator network (DeepONet) architecture was adopted to train the model using time steps relevant to CFD simulations. The ML used physics-based loss functions to ensure mass conservation. The ML model has been deployed in simple MFiX CFD simulations, single particle, and an experimental drop tube reactor, and has shown promising performance compared to the original scheme.

Houston, Ross↗
DOE OSTI2026

Pull Force Evaluation of CCS Insulated End Caps

Insulated end caps installed on DC pins of Combined Charging System (CCS) inlets have been identified as potential debris sources within electric vehicle supply equipment (EVSE) connectors, increasing the risk of electric shock and fire hazards. Although standards such as IEC 62196-1:2022 specify mechanical pull-force requirements for these end caps, it remains unclear whether these requirements provide adequate robustness under real-world conditions. To address this concern, the National Charging Experience (ChargeX) Consortium's Hardware Task Force conducted evaluations of insulated end caps used in OEM CCS inlets. This study assesses the performance of insulated end caps used in OEM CCS inlets, specifically those installed on DC pins, using the procedures defined in IEC 62196-1:2022 (Section Sign) 26.7 and SAE J3400, supplemented by additional experimental conditioning. To ensure relevance to real-world conditions, NLR collaborated with charging network operators (CNOs) to guide the selection of charging inlet samples. Material analyses of the insulated end cap samples were conducted to confirm that the tested materials reflected the types and properties commonly observed in the field. The selected samples underwent temperature and humidity conditioning followed by pull-force testing to evaluate the end-cap integrity. Although SAE J3400 inlets were not directly tested due to limited diversity in available field data, the underlying end-cap principles between CCS and SAE J3400 are comparable, allowing the study's insights to be relevant to both technologies. Overall, this study provides a structured evaluation framework to inform potential refinements to mechanical pull-force requirements in charging standards.

33 ADVANCED PROPULSION SYSTEMS↗
DOE OSTI2026

Blueprint: Coordinated Vulnerability Disclosure (CVD) Adoption for Information Sharing and Analysis Center (ISAC)-Like Groups

The electric vehicle supply equipment (EVSE) industry is an incredibly diverse set of participants (EVSE manufacturers, charge network operators (CNOs), original equipment manufacturers (OEMs), etc.), and with the potential for an Information Sharing and Analysis Centers (ISAC) or ISAC-like group, it requires a series of guidance for doing a multiparty coordinated vulnerability disclosure (CVD) such that a group like this could be successful. This blueprint provides a template and guidance to stakeholders in the EVSE industry for conducting a multiparty CVD. It also formalizes what multiparty CVD could look like in an ISAC-like group with multiple entities as well as vulnerability coordinators by specifically calling out who in the ISAC-like group may be involved, and which industry members it may apply to. This blueprint leverages tools such as Vultron, VINCE, etc. along with open resources such as the Software Engineering Institutes guide for coordinated vulnerability disclosure, for the stakeholder in the EVSE industry to start up a CVD program of their own.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Predicting Cislunar Orbit Lifetimes from Initial Orbital Elements

The volume of space between Earth’s geosynchronous orbit out to the Moon’s sphere of influence, including the lunar Lagrange points, is crucial for the successful planning and execution of space missions, but not fully understood dynamically. This region is a part of cislunar space. Trajectories through cislunar space are influenced by the gravitational forces of the Sun, Earth, Moon, and other Solar System planets leading to typically unpredictable and chaotic trajectory behavior. It is therefore difficult to predict the stability of an trajectory through cislunar space from a set of initial conditions or orbital elements. We simulate one million cislunar orbits to train a self-organizing map (SOM) to cluster the trajectories and orbits into families based on how long they remain stable within the cislunar space. Using the trained SOM, we are able to predict the stable lifetime of a trajectory through cislunar space from a set of initial orbital elements to within an accuracy of 10% for 8% of simulated trajectories and within 50% for 43% of the simulated trajectories. Clustering in the SOM suggests that a variety of trajectory morphologies have similar lifetimes. Once trained, the SOM can predict the stable lifetime of a given cislunar trajectory within milliseconds. The methods developed in this work enable the rapid identification of stable cislunar orbits and trajectories that could be used for future space exploration. Moreover, the developed SOM method can generate orbital and trajectory lifetime estimates from minimal observational data, such as a single two line element, making it useful for early warning systems and large-scale sensor network operations.

79 ASTRONOMY AND ASTROPHYSICS↗