Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “NEURAL 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 235 records · Page 13
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Hybrid Quantum-Classical Neural Network for Calculating Ground State Energies of Molecules

We present a hybrid quantum-classical neural network that can be trained to perform electronic structure calculation and generate potential energy curves of simple molecules. The method is based on the combination of parameterized quantum circuits and measurements. With unsupervised training, the neural network can generate electronic potential energy curves based on training at certain bond lengths. To demonstrate the power of the proposed new method, we present the results of using the quantum-classical hybrid neural network to calculate ground state potential energy curves of simple molecules such as H 2 , LiH, and BeH 2 . The results are very accurate and the approach could potentially be used to generate complex molecular potential energy surfaces.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2024

Defect diffusion graph neural networks (d2gnn)

SAND2025-01004O Defect Diffusion Graph Neural Networks (d2gnn) is a software tool that assists in the discovery of new materials for high-temperature, clean-energy applications. It uses advanced graph neural networks to model the relationship between material structures and their defect properties. The application helps predict how materials will behave under different conditions and accelerates the development of innovative materials. 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.

Witman, Matthew [Sandia National Lab. (SNL-CA), Li↗
DOE OSTI2023

Reducing the cost of n modular redundancy for neural networks

An N modular redundancy method, system, and computer program product include a computer-implemented N modular redundancy method for neural networks, the method including selectively replicating the neural network by employing one of checker neural networks and selective N modular redundancy (N-MR) applied only to critical computations.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Using the Metropolis algorithm to explore the loss surface of a recurrent neural network

In the limit of small trial moves the Metropolis Monte Carlo algorithm is equivalent to gradient descent on the energy function in the presence of Gaussian white noise. This observation was originally used to demonstrate a correspondence between Metropolis Monte Carlo moves of model molecules and overdamped Langevin dynamics, but it also applies in the context of training a neural network: making small random changes to the weights of a neural network, accepted with the Metropolis probability, with the loss function playing the role of energy, has the same effect as training by explicit gradient descent in the presence of Gaussian white noise. We explore this correspondence in the context of a simple recurrent neural network. We also explore regimes in which this correspondence breaks down, where the gradient of the loss function becomes very large or small. In these regimes the Metropolis algorithm can still effect training, and so can be used as a probe of the loss function of a neural network in regimes in which gradient descent struggles. We also show that training can be accelerated by making purposely-designed Monte Carlo trial moves of neural-network weights.

Casert, Corneel↗
DOE OSTI2023

Physics-guided neural networks with engineering domain knowledge for hybrid process modeling

As neural networks are more frequently used to solve problems in science and engineering, the methods used to incorporate scientific knowledge into these networks are becoming increasingly complex. Here, this work breaks down these complicated techniques into a set of basic strategies which can easily be applied to diverse situations. Several novel neural networks are built using the categories laid out in this work. These networks are tested on simulated data from a continuous stirred tank reactor (CSTR) model to evaluate the advantages provided by each network. The three points demonstrated in this work are: (1) architectural hybrid models can speed up convergence and reduce the amount of data necessary to train a model; (2) adding a physics-guided loss function can improve model generalization and make models more physically consistent; (3) using physics-guided initialization and transfer learning improves accuracy and speeds up convergence, but can harm generalizability if used incorrectly.

42 ENGINEERING↗
DOE OSTI2022

Analysis of Inlier and Outlier Compounds with Respect to Artificial Neural Network Cetane Number Prediction Accuracy

Artificial neural networks (ANNs) are exceptional at forming non-linear correlations between multivariate input and target variables; however, they are often seen as a “black box” approach, since how ANNs form these correlations is somewhat ambiguous. Furthermore, the process underlying how ANNs learn from inlier and outlier samples within the input dataset is not fully understood. Intuitively, it is expected that training ANNs with inlier samples will increase prediction accuracy and training with outlier samples will reduce prediction accuracy; though, in practice, this is not always true. The present work identifies and analyzes inliers and outliers of existing experimental cetane number (CN) data encompassing a variety of compounds and compound groups. It also investigates how ANNs trained to predict CN perform with and without outliers included in the training data, and whether a relationship exists between inliers/outliers and ANN prediction accuracy across the whole dataset and for individual samples. Additionally, individual outlier compounds are analyzed, highlighting how they structurally differ from inlier compounds.

09 BIOMASS FUELS↗
DOE OSTI2024

A Survey on Privacy in Graph Neural Networks: Attacks, Preservation, and Applications

Graph Neural Networks (GNNs) have gained significant attention owing to their ability to handle graph-structured data and the improvement in practical applications. However, many of these models prioritize high utility performance, such as accuracy, with a lack of privacy consideration, which is a major concern in modern society where privacy attacks are rampant. To address this issue, researchers have started to develop privacy-preserving GNNs. Despite this progress, there is a lack of a comprehensive overview of the attacks and the techniques for preserving privacy in the graph domain. In this survey, we aim to address this gap by summarizing the attacks on graph data according to the targeted information, categorizing the privacy preservation techniques in GNNs, and reviewing the datasets and applications that could be used for analyzing/solving privacy issues in GNNs. We also outline potential directions for future research in order to build better privacy-preserving GNNs.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

NEURAL NETWORK FOR COHERENT DIFFRACTION IMAGE INVERSION

A deep neural network model plus automatic differentiation is developed for retrieving phase information from 3D coherent diffraction images. The model is implemented using Tensorflow and the training dataset is generated using physics-based atomistic simulations. Custom codes are written to handle the resampling of diffraction images to oversampling ratios appropriate for the neural network model.

CHAN, HENRY↗
DOE OSTI2026

Multiscale Neural Networks for Approximating Green’s Functions

Neural networks (NNs) have been widely used to solve partial differential equations (PDEs) in the applications of physics, biology, and engineering. One effective approach for solving PDEs with a fixed differential operator is learning Green’s functions. However, Green’s functions are notoriously difficult to learn due to their poor regularity, which typically requires larger NNs and longer training times. In this work, we address these challenges by leveraging multiscale NNs to learn Green’s functions. Through theoretical analysis using multiscale Barron space methods and experimental validation, we show that the multiscale approach significantly reduces the necessary NN size and accelerates training.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

A framework for strategic discovery of credible neural network surrogate models under uncertainty

The widespread integration of deep neural networks in developing data-driven surrogate models for high-fidelity simulations of complex physical systems highlights the critical necessity for robust uncertainty quantification techniques and credibility assessment methodologies, ensuring the reliable deployment of surrogate models in consequential decision-making. Here, this study presents the Occam Plausibility Algorithm for surrogate models (OPAL-surrogate), providing a systematic framework to uncover predictive neural network-based surrogate models within the large space of potential models, including various neural network classes and choices of architecture and hyperparameters. The framework is grounded in hierarchical Bayesian inferences and employs model validation tests to evaluate the credibility and prediction reliability of the surrogate models under uncertainty. Leveraging these principles, OPAL-surrogate introduces a systematic and efficient strategy for balancing the trade-off between model complexity, accuracy, and prediction uncertainty. The effectiveness of OPAL-surrogate is demonstrated through two modeling problems, including the deformation of porous materials for building insulation and turbulent combustion flow for ablation of solid fuels within hybrid rocket motors.

42 ENGINEERING↗
DOE OSTI2026

Regression Convolutional Neural Network for Energy Estimation in NOvA

Regression Convolutional Neural Network for Energy Estimation in NOvA" Abstract: "NOvA (NuMI Off-Axis $\nu_e$ Appearance) is a long baseline neutrino experiment designed to measure neutrino oscillations over a distance of 810 km. NOvA employs a near and far detector to observe $\nu_\mu$ disappearance and $\nu_e$ appearance of neutrinos produced by the NuMI beam at Fermilab. Energy reconstruction is critical for precise measurements of neutrino oscillation parameters and cross sections, which are functions of neutrino energy. Energy estimation remains difficult due to the complexity of detector response and final state particle kinematics. We present a regression-based convolutional neural network (CNN) method that reconstructs neutrino and lepton energies based on raw pixel inputs for NOvA. The trained model is able to reconstruct event energy for different interaction modes and complex final states containing leptons and hadrons. Studies of regression CNN networks show improved energy resolution and reduced sensitivity to calibration scale uncertainties relative to traditional kinematics-based energy reconstruction techniques. The results demonstrate the potential of the regression CNN method for neutrino physics analyses by improving on standard kinematics-based reconstruction.

Zhao, Larry [UC, Irvine (main)]↗
DOE OSTI2023

Lowering the computational barrier: Partially Bayesian neural networks for transparency in medical imaging AI

Deep Neural Networks (DNNs) can provide clinicians with fast and accurate predictions that are highly valuable for high-stakes medical decision-making, such as in brain tumor segmentation and treatment planning. However, these models largely lack transparency about the uncertainty in their predictions, potentially giving clinicians a false sense of reliability that may lead to grave consequences in patient care. Growing calls for Transparent and Responsible AI have promoted Uncertainty Quantification (UQ) to capture and communicate uncertainty in a systematic and principled manner. However, traditional Bayesian UQ methods remain prohibitively costly for large, million-dimensional tumor segmentation DNNs such as the U-Net. In this work, we discuss a computationally-efficient UQ approach via the partially Bayesian neural networks (pBNN). In pBNN, only a single layer, strategically selected based on gradient-based sensitivity analysis, is targeted for Bayesian inference. We illustrate the effectiveness of pBNN in capturing the full uncertainty for a 7.8-million parameter U-Net. We also demonstrate how practitioners and model developers can use the pBNN's predictions to better understand the model's capabilities and behavior.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Decentralized Filtering Adaptive Neural Network Control for Uncertain Switched Interconnected Nonlinear Systems

This article presents a novel decentralized filtering adaptive neural network control framework for uncertain switched interconnected nonlinear systems. Each subsystem has its own decentralized controller based on the established decentralized state predictor. For each subsystem, the nonlinear uncertainties are approximated by a Gaussian radial basis function (GRBF) neural network incorporated with a piecewise constant adaptive law, where the adaptive law will update adaptive parameters from the error dynamics between the host system and the decentralized state predictor by discarding the unknowns, whereas a decentralized filtering control law is derived to cancel both local and mismatched uncertainties from other subsystems, as well as achieve the local objective tracking of the host system. The achievement of global objective depends on the achievement of local objective for each subsystem. The matched uncertainties are canceled directly by adopting their opposite in the control signal, whereas a dynamic inversion of the system is required to eliminate the effect of the mismatched uncertainties on the output. By exploiting the average dwell time principle, the error bounds between the real system and the virtual reference system, which defines the best performance that can be achieved by the closed-loop system, are derived. A numerical example is given to illustrate the effectiveness of the decentralized filtering adaptive neural network control architecture by comparing against the model reference adaptive control (MRAC).

Average dwell time, decentralized, filtering adapt↗
DOE OSTI2021

Neural network model of the multi-mode anomalous transport module for accelerated transport simulations

Here, a neural network version of the Multi-Mode Anomalous Transport Module, known as MMMnet, has been developed to calculate plasma turbulent diffusivities in DIII-D with a calculation time suitable for control applications. MMMnet uses a simple artificial neural network structure to predict the ion thermal, electron thermal, and toroidal momentum diffusivities while reproducing Multi-Mode Model (MMM) data with good accuracy and keeping the calculation time as a fraction of that associated with MMM. Model-based control techniques require models with fast calculation times, making many existing physics-oriented predictive codes unsuitable. The control-oriented predictive code COTSIM (Control Oriented Transport SIMulator) calculates the most significant plasma dynamics in response to the different actuators while running at a speed useful for control design. In order to achieve this calculation speed, COTSIM often relies on scaling laws and control-level models. Replacing some of these scaling laws and control-level models with neural network versions of more complex physics-level models has the potential of increasing the range of validity and the level of accuracy of COTSIM without compromising its computational speed. In this work, MMMnet is integrated into COTSIM to improve the turbulent diffusivity predictions, which will in turn improve the prediction accuracy associated with the dynamics of many plasma properties.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

Binary Complex Neural Network Acceleration on FPGA

Being able to learn from complex data with phase information is imperative for many signal processing applications. Today’s real-valued deep neural networks (DNNs) have shown efficiency in latent information analysis but fall short when applied to the complex domain. Deep complex networks (DCN) , in contrast, can learn from complex data, but have high computational costs; therefore, they cannot satisfy the instant decision making requirements of many deployable systems dealing with short observations or short signal bursts. Recent, Binarized Complex Neural Network (BCNN), which integrates DCNs with binarized neural networks (BNN), shows great potential in classifying complex data in real-time. In this paper, we propose a structural pruning based accelerator of BCNN, which is able to provide more than 5000 frames/s inference throughput on edge devices. The high performance comes from both the algorithm and hardware sides. On the algorithm side, we conduct structural pruning to the original BCNN models and obtain 20 × pruning rates with negligible accuracy loss; on the hardware side, we propose a novel 2D convolution operation accelerator for the binary complex neural network. Experimental results show that the proposed design works with over 90% utilization and is able to achieve the inference throughput of 5882 frames/s and 4938 frames/s for complex NIN-Net and ResNet-18 using CIFAR-10 dataset and Alveo U280 Board.

Peng, Hongwu↗
DOE OSTI2025

Simplifying activations with linear approximations in neural networks

A key step in Neural Networks is activation. Among the different types of activation functions, sigmoid, tanh, and others involve the usage of exponents for calculation. From a hardware perspective, exponential implementation implies the usage of Taylor series or repeated methods involving many addition, multiplication, and division steps, and as a result are power-hungry and consume many clock cycles. We implement a piecewise linear approximation of the sigmoid function as a replacement for standard sigmoid activation libraries. This approach provides a practical alternative by leveraging piecewise segmentation, which simplifies hardware implementation and improves computational efficiency. In this paper, we detail piecewise functions that can be implemented using linear approximations and their implications for overall model accuracy and performance gain. Our results show that for the DenseNet, ResNet, and GoogLeNet architectures, the piecewise linear approximation of the sigmoid function provides faster execution times compared to the standard TensorFlow sigmoid implementation while maintaining comparable accuracy. Specifically, for MNIST with DenseNet, accuracy reaches 99.91% (Piecewise) vs. 99.97% (Base) with up to 1.31x speedup in execution time. For CIFAR-10 with DenseNet, accuracy improves to 98.97% (Piecewise) vs. 99.40% (Base) while achieving 1.24x faster execution. Similarly, for CIFAR-100 with DenseNet, the accuracy is 97.93% (Piecewise) vs. 98.39% (Base), with a 1.18x execution time reduction. These results confirm the proposed method’s capability to efficiently process large-scale datasets and computationally demanding tasks, offering a practical means to accelerate deep learning models, including LSTMs, without compromising accuracy.

Activation function↗
DOE OSTI2021

Benchmarking graph neural networks for materials chemistry

Graph neural networks (GNNs) have received intense interest as a rapidly expanding class of machine learning models remarkably well-suited for materials applications. To date, a number of successful GNNs have been proposed and demonstrated for systems ranging from crystal stability to electronic property prediction and to surface chemistry and heterogeneous catalysis. However, a consistent benchmark of these models remains lacking, hindering the development and consistent evaluation of new models in the materials field. Here, we present a workflow and testing platform, MatDeepLearn, for quickly and reproducibly assessing and comparing GNNs and other machine learning models. We use this platform to optimize and evaluate a selection of top performing GNNs on several representative datasets in computational materials chemistry. From our investigations we note the importance of hyperparameter selection and find roughly similar performances for the top models once optimized. We identify several strengths in GNNs over conventional models in cases with compositionally diverse datasets and in its overall flexibility with respect to inputs, due to learned rather than defined representations. Meanwhile several weaknesses of GNNs are also observed including high data requirements, and suggestions for further improvement for applications in materials chemistry are discussed.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Recipes for when physics fails: recovering robust learning of physics informed neural networks

Abstract Physics-informed neural networks (PINNs) have been shown to be effective in solving partial differential equations by capturing the physics induced constraints as a part of the training loss function. This paper shows that a PINN can be sensitive to errors in training data and overfit itself in dynamically propagating these errors over the domain of the solution of the PDE. It also shows how physical regularizations based on continuity criteria and conservation laws fail to address this issue and rather introduce problems of their own causing the deep network to converge to a physics-obeying local minimum instead of the global minimum. We introduce Gaussian process (GP) based smoothing that recovers the performance of a PINN and promises a robust architecture against noise/errors in measurements. Additionally, we illustrate an inexpensive method of quantifying the evolution of uncertainty based on the variance estimation of GPs on boundary data. Robust PINN performance is also shown to be achievable by choice of sparse sets of inducing points based on sparsely induced GPs. We demonstrate the performance of our proposed methods and compare the results from existing benchmark models in literature for time-dependent Schrödinger and Burgers’ equations.

97 MATHEMATICS AND COMPUTING↗