Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “ITERATIVE NETWORK”

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 487 records · Page 27
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

Iterative ML and Experiments for Emerging VOCs

SAND2026-17074O Iterative ML and Experiments for Emerging VOCs is a tool that analyzes and predicts the behaviors of SARS-CoV-2 variants. It processes experimental data on ACE2 (the receptor for the SARS-CoV-2 virus that allows it to infect the cell) and antibody binding using machine learning models, including neural networks, to forecast ACE2 interactions and variant expression. The tool employs transfer learning and global epistasis modeling, integrating public datasets with proprietary data to enhance prediction accuracy. Additionally, it fits concentration-response curves to determine dissociation constants and generates visualizations to support research findings, thereby aiding in the identification of new antibodies for emerging variants of concern. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

Sheffield, Thomas [Sandia National Lab. (SNL-NM), ↗
NASA NTRSDate not supplied

Physics Informed Neural Nets for Systems Health Management

To facilitate and solve the prediction problem, awareness of the current health state of the system is key, since it is necessary to perform condition-based predictions. To accurately predict the future state of any system, it is required to possess knowledge of its current health state and future operational conditions. Development in data-driven algorithms in regression of complex nonlinear functions and classification tasks have generated a growing interest in artificial intelligence for industrial applications. Complex multi-physics models as well as digital twins, once purely built on physics and corresponding simplified lumped parameter iterations, can now benefit from machine learning algorithms to mitigate the lack of understanding of some complex behavior. The research work presents application of physics-informed neural nets application to a representative electric powertrain for unmanned aerial vehicles. The model is composed of physics-derived and empirical equations, integrated with connected networks that are strategically placed within the model to substitute equations that are subject to large uncertainty. Polynomial fit driven by heuristics or empirical observations can be substituted by more flexible networks that can minimize the error between model predictions and observations without being restricted to a predefined functional form. This modeling strategy allows training of networks deep inside the model and unknown parameters in a single learning stage.

Physics Informed↗
DOE OSTI2021

Revisiting Dropout: Escaping Pressure for Training Neural Networks with Multiple Costs

A common approach to jointly learn multiple tasks with a shared structure is to optimize the model with a combined landscape of multiple sub-costs. However, gradients derived from each sub-cost often conflicts in cost plateaus, resulting in a subpar optimum. In this work, we shed light on such gradient conflict challenges and suggest a solution named Cost-Out, which randomly drops the sub-costs for each iteration. We provide the theoretical and empirical evidence of the existence of escaping pressure induced by the Cost-Out mechanism. While simple, the empirical results indicate that the proposed method can enhance the performance of multi-task learning problems, including two-digit image classification sampled from MNIST dataset and machine translation tasks for English from and to French, Spanish, and German WMT14 datasets.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Benders Cut Classification via Support Vector Machines for Solving Two-Stage Stochastic Programs

In this work, we consider Benders decomposition for solving two-stage stochastic programs with complete recourse based on finite samples of the uncertain parameters. We define the Benders cuts binding at the final optimal solution or the ones significantly improving bounds over iterations as valuable cuts. We propose a learning-enhanced Benders decomposition (LearnBD) algorithm, which adds a cut classification step in each iteration to selectively generate cuts that are more likely to be valuable cuts. The LearnBD algorithm includes two phases: (i) sampling cuts and collecting information from training problems and (ii) solving testing problems with a support vector machine (SVM) cut classifier. We run the LearnBD algorithm on instances of capacitated facility location and multicommodity network design under uncertain demand. Our results show that SVM cut classifier works effectively for identifying valuable cuts, and the LearnBD algorithm reduces the total solving time of all instances for different problems with various sizes and complexities.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1994

Improved Multi-Partition Method for Line-Based Iteration Schemes

Regular 3-dimensional multi-partitioning has been shown to be an efficient domain decomposition method for the parallelization of ADI-type algorithms on MIMD architectures. This paper discusses further improvements that can be made to the scheme that increase the granularity and reduce the communication density. These improvements, which are illustrated by simulation and parallel benchmark results, make multi-partitioning the method of choice on systems with relatively poor communication capabilities, such as networks of workstations, or on massively parallel machines with very fast processors, such as the IBM SP2.

Smith, Merritt H.↗
NASA NTRS2003

Masked Proportional Routing

Distributed approach for determining a path connecting adjacent network nodes, for probabilistically or deterministically transporting an entity, with entity characteristic mu from a source node to a destination node. Each node i is directly connected to an arbitrary number J(mu) of nodes, labeled or numbered j=jl, j2, .... jJ(mu). In a deterministic version, a J(mu)-component baseline proportion vector p(i;mu) is associated with node i. A J(mu)-component applied proportion vector p*(i;mu) is determined from p(i;mu) to preclude an entity visiting a node more than once. Third and fourth J(mu)-component vectors, with components iteratively determined by Target(i;n(mu);mu),=alpha(mu).Target(i;n(mu)-1;mu)j+beta(mu).p* (i;mu)j and Actual(i;n(mu);+a(mu)j. Actual(i;n(mu)-l;mu)j+beta(mu).Sent(i;j'(mu);n(mu)-1;mu)j, are computed, where n(mu) is an entity sequence index and alpha(mu) and beta(mu) are selected numbers. In one embodiment, at each node i, the node j=j'(mu) with the largest vector component difference, Target(i;n(mu);mu)j'- Actual (i;n(mu);mu)j'. is chosen for the next link for entity transport, except in special gap circumstances, where the same link is optionally used for transporting consecutively arriving entities. The network nodes may be computer-controlled routers that switch collections of packets, frames, cells or other information units. Alternatively, the nodes may be waypoints for movement of physical items in a network or for transformation of a physical item. The nodes may be states of an entity undergoing state transitions, where allowed transitions are specified by the network and/or the destination node.

David H Wolpert↗
NASA NTRS1996

Development of a Computer Architecture to Support the Optical Plume Anomaly Detection (OPAD) System

The NASA OPAD spectrometer system relies heavily on extensive software which repetitively extracts spectral information from the engine plume and reports the amounts of metals which are present in the plume. The development of this software is at a sufficiently advanced stage where it can be used in actual engine tests to provide valuable data on engine operation and health. This activity will continue and, in addition, the OPAD system is planned to be used in flight aboard space vehicles. The two implementations, test-stand and in-flight, may have some differing requirements. For example, the data stored during a test-stand experiment are much more extensive than in the in-flight case. In both cases though, the majority of the requirements are similar. New data from the spectrograph is generated at a rate of once every 0.5 sec or faster. All processing must be completed within this period of time to maintain real-time performance. Every 0.5 sec, the OPAD system must report the amounts of specific metals within the engine plume, given the spectral data. At present, the software in the OPAD system performs this function by solving the inverse problem. It uses powerful physics-based computational models (the SPECTRA code), which receive amounts of metals as inputs to produce the spectral data that would have been observed, had the same metal amounts been present in the engine plume. During the experiment, for every spectrum that is observed, an initial approximation is performed using neural networks to establish an initial metal composition which approximates as accurately as possible the real one. Then, using optimization techniques, the SPECTRA code is repetitively used to produce a fit to the data, by adjusting the metal input amounts until the produced spectrum matches the observed one to within a given level of tolerance. This iterative solution to the original problem of determining the metal composition in the plume requires a relatively long period of time to execute the software in a modern single-processor workstation, and therefore real-time operation is currently not possible. A different number of iterations may be required to perform spectral data fitting per spectral sample. Yet, the OPAD system must be designed to maintain real-time performance in all cases. Although faster single-processor workstations are available for execution of the fitting and SPECTRA software, this option is unattractive due to the excessive cost associated with very fast workstations and also due to the fact that such hardware is not easily expandable to accommodate future versions of the software which may require more processing power. Initial research has already demonstrated that the OPAD software can take advantage of a parallel computer architecture to achieve the necessary speedup. Current work has improved the software by converting it into a form which is easily parallelizable. Timing experiments have been performed to establish the computational complexity and execution speed of major components of the software. This work provides the foundation of future work which will create a fully parallel version of the software executing in a shared-memory multiprocessor system.

Katsinis, Constantine↗
DOE OSTI2017

Development Of A Parallel Performance Model For The THOR Neutral Particle Transport Code

The THOR neutral particle transport code enables simulation of complex geometries for various problems from reactor simulations to nuclear non-proliferation. It is undergoing a thorough V&V requiring computational efficiency. This has motivated various improvements including angular parallelization, outer iteration acceleration, and development of peripheral tools. For guiding future improvements to the code’s efficiency, better characterization of its parallel performance is useful. A parallel performance model (PPM) can be used to evaluate the benefits of modifications and to identify performance bottlenecks. Using INL’s Falcon HPC, the PPM development incorporates an evaluation of network communication behavior over heterogeneous links and a functional characterization of the per-cell/angle/group runtime of each major code component. After evaluating several possible sources of variability, this resulted in a communication model and a parallel portion model. The former’s accuracy is bounded by the variability of communication on Falcon while the latter has an error on the order of 1%.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

Pre-commercial Demonstration of Direct Non-iterative State Estimator for Operational Use of Synchrophasor Management Systems

The objective of this project was to demonstrate functionality and performance of a Direct Non-iterative State Estimator (DNSE) integrated with NYPA’s Energy Management System (EMS) and with enhanced Real Time Dynamics Monitoring System (RTDMS) synchrophasor platform from Electric Power Group (EPG). DNSE is designed to overcome a major obstacle to operational use of Synchro-Phasor Management Systems (SPMS) by providing to synchrophasor management systems (SPMS) applications a consistent and a complete synchrophasor data foundation in the same way that a traditional EMS State Estimator (SE) provides to EMS applications. Specifically, DNSE is designed to use synchrophasor measurements collected by a central PDC, Supervisory Control and Data Acquisition (SCADA) measurements, and Energy Management System (EMS) network model, to obtain the complete state of the utility’s operating model at rates that are close to the synchrophasor data rates. In this way, the system is comprehensive in that it does not only cover the part of the network that is “visible” via synchrophasors, but also the part that is only “visible” through the SCADA measurements. Visualization needs associated with the use of DNSE results are fulfilled through suitably enhanced Real Time Dynamics Monitoring System (RTDMS), with the enhancements implemented by EPG. This project had the following goals in mind: To advance the deployment of commercial grade DNSE software application that relies on synchrophasor and SCADA data ; Apply DNSE at other utilities, to address a generic and fundamental need for “clean” operational data for synchrophasor applications; Provide means for “live” estimated data access by control system operators; Enhance potential for situational awareness through full system operational model coverage; Sub-second execution rate of the Direct Non-iterative State Estimator, eventually at a near-phasor data rate execution speed, i.e. < 0.1 sec. Anticipated benefits from this projects are: Enhanced reliability and improvements in the economic efficiency of bulk power system planning and operations; Providing “clean” data to other synchrophasor applications; Enhancement of situational awareness by providing the full operational model updated at near synchrophasor rate; A production-grade software tool that incorporate synchrophasor and SCADA data; Provides a basis for development of next generation monitoring and control applications, based on both SCADA and PMU data. Quanta Technology (QT) team worked in collaboration with Electric Power Group (EPG) whose team has enhanced its commercial Real Time Dynamics Monitoring System (RTDMS) to accommodate the requirements posed by DNSE application. EPG also provided its ePDC and Model-less Data Conditioning (PDVC) software for integration with DNSE+. QT developed the system requirements for DNSE; developed system architecture and defined interfaces between internal DNSE components. The core DNSE algorithm with all surrounding interfaces was named DNSE+. Since the DNSE development was done in a simulated system environment, QT used its PMU simulator that was enhanced during this project for development and factory acceptance testing (FAT). SCADA data in this stage was simulated by commercial PSS/e software. The output of DNSE are estimates of System states in C37.118-2 format, sent to RTDMS for further processing and display. As the number of these states is large, it was necessary to expand the C37.111-2 standard to accommodate large data sets. This enhancement was implemented in RTDMS. The demonstration of pre-production DNSE technology was done at NYPA using streaming field data from NYPA PMUs and from its RTUs through their SCADA system. NYPA provided ICCP interface as well as Common Information Model (CIM). The relevance of the DNSE+ application is that it provides state estimation of the power system based on hybrid set of data, consisting of both available PMU data and SCADA measurements. As this is a direct, non-iterative method of calculation of the system states, if does not suffer from convergence issues which is potential problem for conventional state estimators. Also, it can take any available PMU measurements, so it does not need to have a high percentage of PMU coverage needed in the case of Linear State Estimator. As the DNSE calculates synchrophasors of the system states (both phase and absolute value) as sub-second rate, this application can provide a basis for development of next generation of applications based both on SCADA and PMU data.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2016

Fracture Network and Fluid Flow Imaging for Enhanced Geothermal Systems Applications from Multi-Dimensional Electrical Resistivity Structure

We have developed an algorithm for the inversion of magnetotelluric (MT) data to a 3D earth resistivity model based upon the finite element method. Hexahedral edge finite elements are implemented to accommodate discontinuities in the electric field across resistivity boundaries, and to accurately simulate topographic variations. All matrices are reduced and solved using direct solution modules which avoids ill-conditioning endemic to iterative solvers such as conjugate gradients, principally PARDISO for the finite element system and PLASMA for the parameter step estimate. Large model parameterizations can be handled by transforming the Gauss-Newton estimator to data-space form. Accuracy of the forward problem and jacobians has been checked by comparison to integral equations results and by limiting asymptotes. Inverse accuracy and performance has been verified against the public Dublin Secret Test Model 2 and the well-known Mount St Helens 3D MT data set. This algorithm we believe is the most capable yet for forming 3D images of earth resistivity structure and their implications for geothermal fluids and pathways.

15 GEOTHERMAL ENERGY↗
NASA NTRS1991

User's guide for NETS/PROSSS

Expensive analysis programs are often combined with optimization procedures to solve engineering problems. To obtain an optimal solution requires numerous iterations between the analysis program and the optimizer. This often becomes prohibitive due to the cost and amount of computer time needed to converge to an optimal solution. NETS/PROSSS was developed to address this problem. The purpose of this paper is to serve as a user's guide for NETS/PROSSS. The key features include the neural network, determining the training pairs for the neural network, and the approximated analysis/optimization process. A small problem is given to serve as an example of how to apply the system.

Rogers, James L.↗
DOE OSTI2020

Data Driven Online Learning of Power System Dynamics

With the advancement of sensing and communication in power networks, high-frequency real-time data from a power network can be used as a resource to develop better monitoring capabilities. In this work, a systematic approach based on data-driven operator theoretic methods involving Koopman operator is proposed for the online identification of power system dynamics. In particular, a new algorithm is provided, which unlike any previously existing algorithms, updates the Koopman operator iteratively as new data points are acquired. The proposed algorithm has three advantages: a) allows for real-time monitoring of the power system dynamics b) linear power system dynamics (this linear system is usually in a higher dimensional feature space and is not same as linearization of the underlying nonlinear dynamics) and c) computationally fast and less intensive when compared to the popular Extended Dynamic Mode Decomposition (EDMD) algorithm. The efficiency of the proposed algorithm is illustrated on an IEEE 9 bus system using synthetic data from the nonlinear model and on IEEE 39 bus system using synthetic data from the linearized model.

Sinha, Subhrajit↗
DOE OSTI2025

Optimizing transmit field inhomogeneity of parallel RF transmit design in 7T MRI using deep learning

Ultrahigh field (UHF) Magnetic Resonance Imaging (MRI) provides a higher signal-to-noise ratio and, thereby, higher spatial resolution. However, UHF MRI introduces challenges such as transmit radiofrequency (RF) field (B+1) inhomogeneities, leading to uneven flip angles and image intensity anomalies. These issues can significantly degrade imaging quality and its medical applications. This study addresses B+1 field homogeneity through a novel deep learning-based strategy. Traditional methods like Magnitude Least Squares (MLS) optimization have been effective but are time-consuming and dependent on the patient’s presence. Recent machine learning approaches, such as RF Shim Prediction by Iteratively Projected Ridge Regression and deep learning frameworks, have shown promise but face limitations like extensive training times and oversimplified architectures. We propose a two-step deep learning strategy. First, we obtain the desired reference RF shimming weights from multi-channel B+1 fields using random-initialized Adaptive Moment Estimation. Then, we employ Residual Networks (ResNets) to train a model that maps B+1 fields to target RF shimming outputs. Our approach does not rely on pre-calculated reference optimizations for the testing process and efficiently learns residual functions. Comparative studies with traditional MLS optimization demonstrate our method’s advantages in terms of speed and accuracy. The proposed strategy achieves a faster and more efficient RF shimming design, significantly improving imaging quality at UHF. This advancement holds potential for broader applications in medical imaging and diagnostics.

Lu, Zhengyi [Vanderbilt University]↗
DOE OSTI2026

wa-hls4ml: A Benchmark and Surrogate Models for hls4ml Resource and Latency Estimation

As machine learning (ML) is increasingly implemented in hardware to address real-time challenges in scientific applications, the development of advanced toolchains has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as hardware synthesis, are becoming limiting factors in the rapid iteration of designs. To mitigate these emerging constraints, multiple efforts have been undertaken to develop an ML-based surrogate model that estimates resource usage of ML accelerator architectures. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of over 680,000 fully connected and convolutional neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, and the average performance across a subset of the dataset. Additionally, we introduce GNN- and transformer-based surrogate models that predict latency and resources for ML accelerators. We present the architecture and performance of the models and find that the models generally predict latency and resources for the 75% percentile within several percent of the synthesized resources on the synthetic test dataset.

Hawks, Benjamin [Fermilab] (ORCID:0000000157000288↗
DOE OSTI2025

wa-hls4ml: A Benchmark and Surrogate Models for hls4ml Resource and Latency Estimation

As machine learning (ML) is increasingly implemented in hardware to address real-time challenges in scientific applications, the development of advanced toolchains has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as hardware synthesis, are becoming limiting factors in the rapid iteration of designs. To mitigate these emerging constraints, multipleefforts have been undertaken to develop an ML-based surrogate model that estimates resource usage of synthesized ML accelerator architectures. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of over 680 000 fully connected and convolutional neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, and the average performance across a subset of the dataset. Additionally, we introduce GNN- and transformer-based surrogate models that predict latency and resources for ML accelerators. We present the architecture and performance of the models and find that the models generally predict latency and resources for the 75% percentile within several percent of the synthesized resources on the synthetic test dataset.

Hawks, Benjamin G. [Fermilab]↗
DOE OSTI2025

Enhancing Unknown Waveform Detection by Learning Intra and Inter-domain Dependencies with Advanced Attention Fusion Mechanisms

Detection of unknown waveforms in mission-critical communications is a crucial area of interest for the Department of Energy (DoE). Traditional methods and recent deep learning-based approaches often assume that the training set includes all possible classes, which is impractical for detecting new waveforms. This limitation gives rise to the problem of open-set recognition (OSR), which involves correctly identifying known classes while detecting and rejecting unknown or unseen classes. To address this limitation, we propose a novel dual-domain complex-valued neural architecture that jointly processes time-domain and frequency-domain signal representations using transformer mechanisms. A transformer model is a deep learning architecture that uses self-attention mechanisms to process and learn relationships in sequential data. Our model employs a cosine similarity loss to extract domain-specific features and incorporates a transformer architecture in the latent space to weigh the importance of different features from the time and frequency domains. The transformer layer includes stacked self-attention and cross-attention modules to learn intra-domain and inter-domain dependencies, creating a more holistic signal representation. An attention-based fusion module intelligently combines the time and frequency-domain features using multi-head attention, enabling the network to learn the optimal feature for each domain in each input signal. Quantitative results demonstrate the impact of these architectural choices on overall performance, showing significant improvement after incorporating self and cross-attention modules and using complex attention fusion over simple weighted fusion. Our ongoing work will focus on addressing the limitations of threshold-based OSR methods by developing a novel generative framework that integrates a conditional diffusion probabilistic model (DPM). DPM is a generative framework that learns to synthesize complex data by reversing a gradual noising process using a neural network trained to denoise step-by-step. Our goal is to leverage the inherent strengths of DPMs for identifying unknown signals more robustly. One primary advantage of using a DPM is its ability to provide a more reliable anomaly score based on the model's reconstruction error, rather than relying solely on classifier confidence. Additionally, the iterative denoising process of DPMs makes this approach naturally resilient to low Signal-to-Noise Ratio (SNR) conditions, where traditional methods often fail. By implementing this generative framework, we aim to enhance the model's capability to accurately detect unknown waveforms and maintain performance in challenging environments.

99 - GENERAL AND MISCELLANEOUS↗
DOE OSTI2019

Machine-learning error models for approximate solutions to parameterized systems of nonlinear equations

Here, this work proposes a machine-learning framework for constructing statistical models of errors incurred by approximate solutions to parameterized systems of nonlinear equations. These approximate solutions may arise from early termination of an iterative method, a lower-fidelity model, or a projection-based reduced-order model, for example. The proposed statistical model comprises the sum of a deterministic regression-function model and a stochastic noise model. The method constructs the regression-function model by applying regression techniques from machine learning (e.g., support vector regression, artificial neural networks) to map features (i.e., error indicators such as sampled elements of the residual) to a prediction of the approximate-solution error. The method constructs the noise model as a mean-zero Gaussian random variable whose variance is computed as the sample variance of the approximate-solution error on a test set; this variance can be interpreted as the epistemic uncertainty introduced by the approximate solution. This work considers a wide range of feature-engineering methods, data-set-construction techniques, and regression techniques that aim to ensure that (1) the features are cheaply computable, (2) the noise model exhibits low variance (i.e., low epistemic uncertainty introduced), and (3) the regression model generalizes to independent test data. Finally, numerical experiments performed on several computational-mechanics problems and types of approximate solutions demonstrate the ability of the method to generate statistical models of the error that satisfy these criteria and significantly outperform more commonly adopted approaches for error modeling.

97 MATHEMATICS AND COMPUTING↗