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 271 records · Page 15

Parallel-in-Time Solution of Allen-Cahn Equations by Integrating Operator Learning into the Parareal Method

While recent advances in deep learning have shown promising efficiency gains in solving time-dependent partial differential equations (PDEs), matching the accuracy of conventional numerical solvers still remains a challenge. One strategy to improve the accuracy of deep learning-based solutions for time-dependent PDEs is to use the learned model as the coarse propagator in the Parareal method and a traditional numerical method as the fine solver. However, successful integration of deep learning into the Parareal method requires consistency between the coarse and fine solvers, particularly for PDEs exhibiting rapid changes such as sharp transitions. Here, to ensure this consistency, we propose using convolutional neural networks (CNNs) to learn the fully discrete time-stepping operator defined by the same numerical scheme employed as the fine solver. We demonstrate the effectiveness of the proposed method in solving the classical and mass-conservative Allen–Cahn (AC) equations. Through iterative updates in the Parareal algorithm, our approach achieves a significant computational speedup compared to traditional fine solvers while converging to high-accuracy solutions. Our results highlight that the proposed hybrid Parareal algorithm effectively accelerates simulations, particularly when implemented on multiple GPUs, and converges to the desired accuracy in only a few iterations. Another advantage of our method is that the CNN model is trained on trajectory-based data generated from random initial conditions, such that the trained model can be used to solve the AC equations with various initial conditions without retraining. This work demonstrates the potential of integrating neural network methods into parallel-in-time frameworks for efficient and accurate simulations of time-dependent PDEs.

97 MATHEMATICS AND COMPUTING↗

Ensemble learning-iterative training machine learning for uncertainty quantification and automated experiment in atom-resolved microscopy

Deep learning has emerged as a technique of choice for rapid feature extraction across imaging disciplines, allowing rapid conversion of the data streams to spatial or spatiotemporal arrays of features of interest. However, applications of deep learning in experimental domains are often limited by the out-of-distribution drift between the experiments, where the network trained for one set of imaging conditions becomes sub-optimal for different ones. This limitation is particularly stringent in the quest to have an automated experiment setting, where retraining or transfer learning becomes impractical due to the need for human intervention and associated latencies. Here we explore the reproducibility of deep learning for feature extraction in atom-resolved electron microscopy and introduce workflows based on ensemble learning and iterative training to greatly improve feature detection. This approach allows incorporating uncertainty quantification into the deep learning analysis and also enables rapid automated experimental workflows where retraining of the network to compensate for out-of-distribution drift due to subtle change in imaging conditions is substituted for human operator or programmatic selection of networks from the ensemble. This methodology can be further applied to machine learning workflows in other imaging areas including optical and chemical imaging.

36 MATERIALS SCIENCE↗

Rapid 3D nanoscale coherent imaging via physics-aware deep learning

Phase retrieval, the problem of recovering lost phase information from measured intensity alone, is an inverse problem that is widely faced in various imaging modalities ranging from astronomy to nanoscale imaging. The current process of phase recovery is iterative in nature. As a result, the image formation is time consuming and computationally expensive, precluding real-time imaging. Here, we use 3D nanoscale X-ray imaging as a representative example to develop a deep learning model to address this phase retrieval problem. We introduce 3D-CDI-NN, a deep convolutional neural network and differential programing framework trained to predict 3D structure and strain, solely from input 3D X-ray coherent scattering data. Our networks are designed to be “physics-aware” in multiple aspects; in that the physics of the X-ray scattering process is explicitly enforced in the training of the network, and the training data are drawn from atomistic simulations that are representative of the physics of the material. We further refine the neural network prediction through a physics-based optimization procedure to enable maximum accuracy at lowest computational cost. 3D-CDI-NN can invert a 3D coherent diffraction pattern to real-space structure and strain hundreds of times faster than traditional iterative phase retrieval methods. Our integrated machine learning and differential programing solution to the phase retrieval problem is broadly applicable across inverse problems in other application areas.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Parallel Dynamics Simulation Using a Krylov-Schwarz Linear Solution Scheme

Fast dynamics simulation of large-scale power systems is a computational challenge because of the need to solve a large set of stiff, nonlinear differential-algebraic equations at every time step. The main bottleneck in dynamic simulations is the solution of a linear system during each nonlinear iteration of Newton’s method. In this paper, we present a parallel Krylov- Schwarz linear solution scheme that uses the Krylov subspacebased iterative linear solver GMRES with an overlapping restricted additive Schwarz preconditioner. As a result, performance tests of the proposed Krylov-Schwarz scheme for several large test cases ranging from 2,000 to 20,000 buses, including a real utility network, show good scalability on different computing architectures.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Fast-RF-Shimming: Accelerate RF shimming in 7T MRI using deep learning

Ultrahigh field (UHF) Magnetic Resonance Imaging (MRI) offers an elevated signal-to-noise ratio (SNR), enabling exceptionally high spatial resolution that benefits both clinical diagnostics and advanced research. However, the jump to higher fields introduces complications, particularly transmit radiofrequency (RF) field ($B^{+}_{1}$) inhomogeneities, manifesting as uneven flip angles and image intensity irregularities. These artifacts can degrade image quality and impede broader clinical adoption. Traditional RF shimming methods, such as Magnitude Least Squares (MLS) optimization, effectively mitigate $B^{+}_{1}$ inhomogeneity, but remain time-consuming. Recent machine learning approaches, including RF Shim Prediction by Iteratively Projected Ridge Regression and other deep learning architectures, suggest alternative pathways. Although these approaches show promise, challenges such as extensive training periods, limited network complexity, and practical data requirements persist. In this paper, we introduce a holistic learning-based framework called Fast-RF-Shimming, which achieves a 5000 ​× ​speed-up compared to the traditional MLS method. In the initial phase, we employ random-initialized Adaptive Moment Estimation (Adam) to derive the desired reference shimming weights from multi-channel $B^{+}_{1}$ fields. Next, we train a Residual Network (ResNet) to map $B^{+}_{1}$ fields directly to the ultimate RF shimming outputs, incorporating the confidence parameter into its loss function. Finally, we design Non-uniformity Field Detector (NFD), an optional post-processing step, to ensure the extreme non-uniform outcomes are identified. Comparative evaluations with standard MLS optimization underscore notable gains in both processing speed and predictive accuracy, which indicates that our technique shows a promising solution for addressing persistent inhomogeneity challenges.

Deep learning↗

GradDFT. A software library for machine learning enhanced density functional theory

Density functional theory (DFT) stands as a cornerstone method in computational quantum chemistry and materials science due to its remarkable versatility and scalability. Yet, it suffers from limitations in accuracy, particularly when dealing with strongly correlated systems. To address these shortcomings, recent work has begun to explore how machine learning can expand the capabilities of DFT: an endeavor with many open questions and technical challenges. In this work, we present GradDFT a fully differentiable JAX-based DFT library, enabling quick prototyping and experimentation with machine learning-enhanced exchange–correlation energy functionals. GradDFT employs a pioneering parametrization of exchange–correlation functionals constructed using a weighted sum of energy densities, where the weights are determined using neural networks. Moreover, GradDFT encompasses a comprehensive suite of auxiliary functions, notably featuring a just-in-time compilable and fully differentiable self-consistent iterative procedure. To support training and benchmarking efforts, we additionally compile a curated dataset of experimental dissociation energies of dimers, half of which contain transition metal atoms characterized by strong electronic correlations. The software library is tested against experimental results to study the generalization capabilities of a neural functional across potential energy surfaces and atomic species, as well as the effect of training data noise on the resulting model accuracy.

Chemistry↗

Reconstructing cosmic polarization rotation with ResUNet-CMB

Cosmic polarization rotation, which may result from parity-violating new physics or the presence of primordial magnetic fields, converts E-mode polarization of the cosmic microwave background (CMB) into B-mode polarization. Anisotropic cosmic polarization rotation leads to statistical anisotropy in CMB polarization and can be reconstructed with quadratic estimator techniques similar to those designed for gravitational lensing of the CMB. At the sensitivity of upcoming CMB surveys, lensing-induced B-mode polarization will act as a limiting factor in the search for anisotropic cosmic polarization rotation, meaning that an analysis which incorporates some form of delensing will be required to improve constraints on the effect with future surveys. In this study we extend the ResUNet-CMB convolutional neural network to reconstruct anisotropic cosmic polarization rotation in the presence of gravitational lensing and patchy reionization, and we show that the network simultaneously reconstructs all three effects with variance that is lower than that from the standard quadratic estimator nearly matching the performance of an iterative reconstruction method.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Cyber Physical Grid-Interactive Distributed Energy Resources Control for VPP Dispatch and Regulation

This paper presents a cyber-physical algorithm for grid-interactive Distributed Energy Resource (DER) control to enable two features of Virtual Power Plants (VPPs) dispatch and grid voltage regulation, considering the communication and security impacts. We first formulate the DER dispatch problem as a real-time, iterative, and grid-interactive DER control problem. Thereafter, we consider a probabilistic traffic model to characterize packet delays and loss in a communication network, and study how the delays enter the process of information exchange among the grid measurement units, local DER controllers and the grid control center that coordinately execute this dispatch algorithm. Finally, a strategy combining delay threshold and modified message update rules is proposed to immune the asynchrony resulting from the communications network traffic and it avoids possible numerical instabilities and sensitivities of the tracking and regulation capabilities of this DER control algorithm. By implementing the proposed cyber-physical algorithm on the modified IEEE 37-node system, our preliminary results exhibit that the uncertainties of the underlying communications infrastructure must be considered for the VPP tracking and regulation capabilities of any DER in a generic Cyber-Physical System (CPS), because the delayed voltage measurements in the uplink/bi-link cases result in the off-track in VPP dispatch and jittery in voltage regulation.

cyber-physical algorithm↗

Resilient Distributed Estimation: Sensor Attacks

Here, this paper studies multi-agent distributed estimation under sensor attacks. Individual agents make sensor measurements of an unknown parameter belonging to a compact set, and, at every time step, a fraction of the agents’ sensor measurements may fall under attack and take arbitrary values. We present the Saturated Innovation Update (SIU) algorithm for distributed estimation resilient to sensor attacks. Under the iterative SIU algorithm, if less than one half of the agent sensors fall under attack, then, all of the agents’ estimates converge at a polynomial rate (with respect to the number of iterations) to the true parameter. The resilience of SIU to sensor attacks does not depend on the topology of the inter-agent communication network, as long as it remains connected. We demonstrate the performance of SIU with numerical examples.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Cyber Physical Grid-Interactive Distributed Energy Resources Control for VPP Dispatch and Regulation: Preprint

This paper presents a cyber-physical algorithm for grid interactive DER control to enable two features of Virtual Power Plants (VPPs) dispatch and grid voltage regulation, considering the communication and security impacts. We first formulate the DER dispatch problem as a real-time, iterative, and grid-interactive DER control problem. Thereafter, we consider a probabilistic traffic model to characterize packet delays and loss in a communication network, and study how the delays enter the process of information exchange among the grid measurement units, local DER controllers and the grid control center that coordinately execute this dispatch algorithm. Finally, we propose a cyber-physical grid-interactive DER control algorithm using the previous message strategy. The tracking and regulation capabilities of this proposed algorithm can be made immune to the asynchrony resulting from the communications network traffic. We carry out simulations to show possible numerical instabilities and sensitivities of the tracking and regulation capabilities on the proposed strategy. Our results exhibit that the uncertainties of the underlying communications infrastructure must be considered in the control algorithm for the VPP tracking and regulation capabilities of any DER in a generic Cyber-Physical System (CPS).

co-simulation↗

Analysis of the phase control of the ITER ICRH antenna array. Influence on the load resilience and radiated power spectrum

The paper analyses how the phasing of the ITER ICRH 24 strap array evolves from the power sources up to the strap currents of the antenna. The study of the phasing control and coherence through the feeding circuits with prematching and automatic matching and decoupling network is made by modeling starting from the TOPICA matrix of the antenna array for a low coupling plasma profile and for current drive phasing (worst case for mutual coupling effects). The main results of the analysis are: (i) the strap current amplitude is well controlled by the antinode V{sub max} amplitude of the feeding lines, (ii) the best toroidal phasing control is done by the adjustment of the mean phase of V{sub max} of each poloidal straps column, (iii) with well adjusted system the largest strap current phasing error is ±20°, (iv) the effect on load resilience remains well below the maximum affordable VSWR of the generators, (v) the effect on the radiated power spectrum versus k{sub //} computed by means of the coupling code ANTITER II remains small for the considered cases.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Distributed energy management for community microgrids considering network operational constraints and building thermal dynamics

A distributed energy management system for community microgrids is developed here. Unlike a centralized optimization-based energy management system, it schedules distributed energy resources and energy storage systems, as well as residential appliances, indirectly through iterative interaction between the microgrid central controller and home energy management systems, based on price signals. In each iteration, the microgrid central controller adjusts the scheduling of distributed energy resources and energy storage systems at the microgrid level. Meanwhile, the home energy management system of each house adjusts the scheduling of residential appliances. Then, the energy price at each bus is updated according to the unbalanced power between generation and demand. The optimization converges when the unbalanced power of all buses is close to zero, i.e., the microgrid central controller and home energy management systems reach an agreement on the energy price and generation/consumption. In particular, a detailed thermal dynamic model of the house is integrated into the HEMS scheduling for intelligent control of the heating, ventilation, and air-conditioning system by customers. Lastly, the distribution network is also considered to make the energy price signal locational. Results of case studies validate the efficacy of the proposed distributed energy management system.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Global Solution Strategies for the Network-Constrained Unit Commitment Problem with AC Transmission Constraints

Here, we propose a novel global solution algorithm for the network-constrained unit commitment problem that incorporates a nonlinear alternating current (AC) model of the transmission network, which is a nonconvex mixed-integer nonlinear programming (MINLP) problem. Our algorithm is based on the multi-tree global optimization methodology, which iterates between a mixed-integer lower-bounding problem and a nonlinear upper-bounding problem. We exploit the mathematical structure of the unit commitment problem with AC power flow constraints (UC-AC) and leverage second-order cone relaxations, piecewise outer approximations, and optimization-based bounds tightening to guarantee a globally optimal solution at convergence. Numerical results on four benchmark problems illustrate the effectiveness of our algorithm, both in terms of convergence rate and solution quality.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Smart-DS: Synthetic Models for Advanced, Realistic Testing: Distribution Systems and Scenarios

The National Renewable Energy Laboratory (NREL) in collaboration with Massachusetts Institute of Technology (MIT), Universidad Pontificia Comillas (Comillas-IIT, Spain) and GE Grid Solutions, is working on an ARPA-E GRID DATA project, titled Smart-DS, to create: 1) High-quality, realistic, synthetic distribution network models, and 2) Advanced tools for automated scenario generation based on high-resolution weather data and generation growth projections. Through these advancements, the Smart-DS project is envisioned to accelerate the development, testing, and adoption of advanced algorithms, approaches, and technologies for sustainable and resilient electric power systems, especially in the realm of U.S. distribution systems. This talk will present the goals and overall approach of the Smart-DS project, including the process of creating the synthetic distribution datasets using reference network model (RNM) and the comprehensive validation process to ensure network realism, feasibility, and applicability to advanced use cases. The talk will provide demonstrations of early versions of synthetic models, along with the lessons learnt from expert engagements to enhance future iterations. Finally, the scenario generation framework, its development plans, and co-ordination with GRID DATA repository teams to house these datasets for public access will also be discussed.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Real-time capable modeling of neutral beam injection on NSTX-U using neural networks

A new model of heating, current drive, torque and other effects of neutral beam injection on NSTX-U that uses neural networks has been developed. The model has been trained and tested on the results of the Monte Carlo code NUBEAM for the database of experimental discharges taken during the first operational campaign of NSTX-U. By projecting flux surface quantities onto empirically derived basis functions, the model is able to efficiently and accurately reproduce the behavior of both scalars, like the total neutron rate and shine through, and profiles, like beam current drive and heating. The model has been tested on the NSTX-U real-time computer, demonstrating a rapid execution time orders of magnitude faster than the Monte Carlo code that is well suited for the iterative calculations needed to interpret experimental results, optimization during scenario development activities, and real-time plasma control applications. Simulation results of a proposed design for a nonlinear observer that embeds the neural network calculations to estimate the poloidal flux profile evolution, as well as effective charge and fast ion diffusivity, are presented.

Neural networks↗

CRADA: Data Acquisition and Tomographic Software Development

Dr. Raymond R. Raylman, Professor and Vice Chair for Research, Radiology Department at WVU was the Principal Investigator for the project. The Jefferson Lab partner on this project was the Radiation Detector and Imaging Group in the Physics Division, led by Dr. Drew Weisenberger who was the coordinator of the project execution and the chief contact with WVU. The responsibility of Jefferson Lab was to facilitate optimization of data acquisition software, conversion of JLab Fortran iterative positron emission mammography computer code to C++ and to act as a computer code repository. During the two year project interactions and exchange of information was directed by Dr. Raylman and Dr. Wesienberger. WVU provided the expertise that allowed JLab to continue development of the current JLab data acquisition software referred to as JDAQ. JDAQ is a highly flexible Java language based data acquisition tool that helps to minimize software induced dead-time while maintaining the highest possible data rates from a JLab built FPGA based analog-to-digital convertor (ADC). JDAQ interfaces directly with the FPGA based multi-channel ADCs. Using JDAQ it is possible process data from a number of different detector types such as the PET systems at WVU. Because of the inherent flexibility of our processing software, we are also able adjust detector readout parameters during operation to provide the best possible data presentation and calibration parameters. To accommodate the high rate nature of PET detector systems JDAQ was designed to be scalable across an Ethernet network as well as for multi-cpu computer systems, and it also has the capability to synchronize many ADCs connected to multiple computers. During the project WVU will provided iterative reconstruction algorithm expertise and Fortran programming assistance to facilitate the conversion to C++ of a JLab developed Fortran positron emission mammography (PEM) reconstruction code. Also WVU provided acceptance testing of the new C++ based PEM reconstruction code. The PEM code is a maximization (MLEM) image reconstruction program that created images from the list-mode data acquired from the PET system.

97 MATHEMATICS AND COMPUTING↗

Real-time capable modeling of neutral beam injection on NSTX-U using neural networks

A new model of heating, current drive, torque and other effects of neutral beam injection on NSTX-U that uses neural networks has been developed. The model has been trained and tested on the results of the Monte Carlo code NUBEAM for the database of experimental discharges taken during the first operational campaign of NSTX-U. By projecting flux surface quantities onto empirically derived basis functions, the model is able to efficiently and accurately reproduce the behavior of both scalars, like the total neutron rate and shine through, and profiles, like beam current drive and heating. With this being said, the model has been tested on the NSTX-U real-time computer, demonstrating a rapid execution time orders of magnitude faster than the Monte Carlo code that is well suited for the iterative calculations needed to interpret experimental results, optimization during scenario development activities, and real-time plasma control applications. Simulation results of a proposed design for a nonlinear observer that embeds the neural network calculations to estimate the poloidal flux profile evolution, as well as Z eff and fast ion diffusivity, are presented.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗