Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “neural algorithms”

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 577 records · Page 32
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Semantic segmentation with a sparse convolutional neural network for event reconstruction in MicroBooNE

We present the performance of a semantic segmentation network, SparseSSNet, that provides pixel-level classification of MicroBooNE data. The MicroBooNE experiment employs a liquid argon time projection chamber for the study of neutrino properties and interactions. SparseSSNet is a submanifold sparse convolutional neural network, which provides the initial machine learning based algorithm utilized in one of MicroBooNE's ν e -appearance oscillation analyses. The network is trained to categorize pixels into five classes, which are re-classified into two classes more relevant to the current analysis. The output of SparseSSNet is a key input in further analysis steps. This technique, used for the first time in liquid argon time projection chambers data and is an improvement compared to a previously used convolutional neural network, both in accuracy and computing resource utilization. Here, the accuracy achieved on the test sample is ≥ 99%. For full neutrino interaction simulations, the time for processing one image is ≈ 0.5 sec, the memory usage is at 1 GB level, which allows utilization of most typical CPU worker machine.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2020

Fast widefield imaging of neuronal structure and function with optical sectioning in vivo

Optical microscopy, owing to its noninvasiveness and subcellular resolution, enables in vivo visualization of neuronal structure and function in the physiological context. Optical-sectioning structured illumination microscopy (OS-SIM) is a widefield fluorescence imaging technique that uses structured illumination patterns to encode in-focus structures and optically sections 3D samples. However, its application to in vivo imaging has been limited. In this study, we optimized OS-SIM for in vivo neural imaging. We modified OS-SIM reconstruction algorithms to improve signal-to-noise ratio and correct motion-induced artifacts in live samples. Incorporating an adaptive optics (AO) module to OS-SIM, we found that correcting sample-induced optical aberrations was essential for achieving accurate structural and functional characterizations in vivo. With AO OS-SIM, we demonstrated fast, high-resolution in vivo imaging with optical sectioning for structural imaging of mouse cortical neurons and zebrafish larval motor neurons, and functional imaging of quantal synaptic transmission at Drosophila larval neuromuscular junctions.

42 ENGINEERING↗
DOE OSTI2025

PruningAMR

PruningAMR is an algorithm that reads in data stored as a neural network and outputs an adaptive mesh with values of the neural network stored at its vertices. Examples of input data include implicit neural representations (INRs) and physics-informed neural networks (PINNs). The output mesh is a grid-based adaptive mesh with larger elements for regions in which the neural network has coarse-scale variation and smaller elements for regions with fine scale variation. The goal of the software is to discretize the neural network to a mesh that faithfully captures the details encoded in the data without resorting a fine scale mesh.

Zvonek, JenniferE [Lawrence Livermore National Lab↗
DOE OSTI2020

Progress Toward the First Search for Bound Neutron Oscillation into Antineutron in a Liquid Argon TPC

This note presents current progress for a neutron-antineutron oscillation ($n–\overline{n}$) search in MicroBooNE paving the way for the first search analysis of such process in a Liquid Argon Time Projection Chamber (LArTPC). Convolutional Neural Network (CNN) and Boosted Decision Tree (BDT) algorithms were used to select signal $n–\overline{n}$ events over cosmogenic backgrounds. The CNN-only, BDT-only, and the combined (CNN+BDT) methods were demonstrated on the Monte-Carlo signal and background events. Validation of the CNNonly and the BDT-only methods was carried out on a small dataset of MicroBooNE Run1 off-beam data, setting the starting point toward further improvement of the analysis.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2026

Status of 1eNp0π Charged-Current Electron Neutrino Cross Section on Argon in the NuMI Beam at ICARUS

The Short-Baseline Neutrino (SBN) Program is designed to probe short-baseline neutrino anomalies, including the LSND electron neutrino excess and the MiniBooNE low-energy excess. Essential to interpreting these anomalies and to the success of future experiments like DUNE, is the precise measurement of neutrino-argon interaction cross sections. The program utilizes two liquid argon time projection chamber (LArTPC) detectors: the Short-Baseline Near Detector (SBND) located 110 meters downstream from the Booster Neutrino Beam (BNB) target, and the ICARUS detector positioned 600 meters downstream. Additionally, the ICARUS detector lies off-axis to the NuMI beamline, providing a unique, high-statistics flux of electron neutrinos and sensitivity to energies that overlap with the DUNE spectrum. To analyze the data from these detectors, we have begun employing a machine-learning-based reconstruction algorithm referred to as “Scalable Particle Imaging with Neural Embeddings” (SPINE). SPINE has shown improvement in the ability to reconstruct the properties of final state particles in the detector, like the particle ID and momentum, with the potential to enhance the quality of measurements achievable within the SBN Program e.g., the resolution on kinematics used in differential cross section extraction. This poster presents progress toward measuring the electron neutrino argon interaction cross section in the 1eNp0π topology using the NuMI beam, highlighting the impact of SPINE through the ability to select signal events across a wide kinematic range without sacrificing background rejection power.

Carber, Dan [Colorado State U.] (ORCID:00090006451↗
DOE OSTI2024

Improving the Concrete Crack Detection Process via a Hybrid Visual Transformer Algorithm

Inspections of concrete bridges across the United States represent a significant commitment of resources, given their biannual mandate for many structures. With a notable number of aging bridges, there is an imperative need to enhance the efficiency of these inspections. This study harnessed the power of computer vision to streamline the inspection process. Our experiment examined the efficacy of a state-of-the-art Visual Transformer (ViT) model combined with distinct image enhancement detector algorithms. We benchmarked against a deep learning Convolutional Neural Network (CNN) model. These models were applied to over 20,000 high-quality images from the Concrete Images for Classification dataset. Traditional crack detection methods often fall short due to their heavy reliance on time and resources. This research pioneers bridge inspection by integrating ViT with diverse image enhancement detectors, significantly improving concrete crack detection accuracy. Notably, a custom-built CNN achieves over 99% accuracy with substantially lower training time than ViT, making it an efficient solution for enhancing safety and resource conservation in infrastructure management. These advancements enhance safety by enabling reliable detection and timely maintenance, but they also align with Industry 4.0 objectives, automating manual inspections, reducing costs, and advancing technological integration in public infrastructure management.

42 ENGINEERING↗
NASA NTRS1994

Neural net controller for inlet pressure control of rocket engine testing

Many dynamic systems operate in select operating regions, each exhibiting characteristic modes of behavior. It is traditional to employ standard adjustable gain proportional-integral-derivative (PID) loops in such systems where no apriori model information is available. However, for controlling inlet pressure for rocket engine testing, problems in fine tuning, disturbance accommodation, and control gains for new profile operating regions (for research and development) are typically encountered. Because of the capability of capturing I/O peculiarities, using NETS, a back propagation trained neural network is specified. For select operating regions, the neural network controller is simulated to be as robust as the PID controller. For a comparative analysis, the higher order moment neural array (HOMNA) method is used to specify a second neural controller by extracting critical exemplars from the I/O data set. Furthermore, using the critical exemplars from the HOMNA method, a third neural controller is developed using NETS back propagation algorithm. All controllers are benchmarked against each other.

Trevino, Luis C.↗
NASA NTRS1995

The effect of lossy image compression on image classification

We have classified four different images, under various levels of JPEG compression, using the following classification algorithms: minimum-distance, maximum-likelihood, and neural network. The training site accuracy and percent difference from the original classification were tabulated for each image compression level, with maximum-likelihood showing the poorest results. In general, as compression ratio increased, the classification retained its overall appearance, but much of the pixel-to-pixel detail was eliminated. We also examined the effect of compression on spatial pattern detection using a neural network.

Paola, Justin D.↗
NASA NTRS2002

Multiple Component Event-Related Potential (mcERP) Estimation

We show how model-based estimation of the neural sources responsible for transient neuroelectric signals can be improved by the analysis of single trial data. Previously, we showed that a multiple component event-related potential (mcERP) algorithm can extract the responses of individual sources from recordings of a mixture of multiple, possibly interacting, neural ensembles. McERP also estimated single-trial amplitudes and onset latencies, thus allowing more accurate estimation of ongoing neural activity during an experimental trial. The mcERP algorithm is related to informax independent component analysis (ICA); however, the underlying signal model is more physiologically realistic in that a component is modeled as a stereotypic waveshape varying both in amplitude and onset latency from trial to trial. The result is a model that reflects quantities of interest to the neuroscientist. Here we demonstrate that the mcERP algorithm provides more accurate results than more traditional methods such as factor analysis and the more recent ICA. Whereas factor analysis assumes the sources are orthogonal and ICA assumes the sources are statistically independent, the mcERP algorithm makes no such assumptions thus allowing investigators to examine interactions among components by estimating the properties of single-trial responses.

Knuth, K. H.↗
DOE OSTI2021

Benchmarking the Performance of Neuromorphic and Spiking Neural Network Simulators

Software simulators play a critical role in the development of new algorithms and system architectures in any field of engineering. Neuromorphic computing, which has shown potential in building brain-inspired energy-efficient hardware, suffers a slow-down in the development cycle due to a lack of flexible and easy-to-use simulators of either neuromorphic hardware itself or of spiking neural networks (SNNs), the type of neural network computation executed on most neuromorphic systems. While there are several openly available neuromorphic or SNN simulation packages developed by a variety of research groups, they have mostly targeted computational neuroscience simulations, and only a few have targeted small-scale machine learning tasks with SNNs. Evaluations or comparisons of these simulators have often targeted computational neuroscience-style workloads. In this work, we seek to evaluate the performance of several publicly available SNN simulators with respect to non-computational neuroscience workloads, in terms of speed, flexibility, and scalability. We evaluate the performance of the NEST, Brian2, Brian2GeNN, BindsNET and Nengo packages under a common front-end neuromorphic framework. Our evaluation tasks include a variety of different network architectures and workload types to mimic the computation common in different algorithms, including feed-forward network inference, genetic algorithms, and reservoir computing. We also study the scalability of each of these simulators when running on different computing hardware, from single core CPU workstations to multi-node supercomputers. Our results show that the BindsNET simulator has the best speed and scalability for most of the SNN workloads (sparse, dense, and layered SNN architectures) on a single core CPU. However, when comparing the simulators leveraging the GPU capabilities, Brian2GeNN outperforms the others for these workloads in terms of scalability. NEST performs the best for small sparse networks and is also the most flexible simulator in terms of reconfiguration capability NEST shows a speedup of at least 2x compared to the other packages when running evolutionary algorithms for SNNs. The multi-node and multi-thread capabilities of NEST show at least 2x speedup compared to the rest of the simulators (single core CPU or GPU based simulators) for large and sparse networks. We conclude our work by providing a set of recommendations on the suitability of employing these simulators for different tasks and scales of operations. We also present the characteristics for a future generic ideal SNN simulator for different neuromorphic computing workloads.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2004

Aerodynamic Optimization of Rocket Control Surface Geometry Using Cartesian Methods and CAD Geometry

Aerodynamic design is an iterative process involving geometry manipulation and complex computational analysis subject to physical constraints and aerodynamic objectives. A design cycle consists of first establishing the performance of a baseline design, which is usually created with low-fidelity engineering tools, and then progressively optimizing the design to maximize its performance. Optimization techniques have evolved from relying exclusively on designer intuition and insight in traditional trial and error methods, to sophisticated local and global search methods. Recent attempts at automating the search through a large design space with formal optimization methods include both database driven and direct evaluation schemes. Databases are being used in conjunction with surrogate and neural network models as a basis on which to run optimization algorithms. Optimization algorithms are also being driven by the direct evaluation of objectives and constraints using high-fidelity simulations. Surrogate methods use data points obtained from simulations, and possibly gradients evaluated at the data points, to create mathematical approximations of a database. Neural network models work in a similar fashion, using a number of high-fidelity database calculations as training iterations to create a database model. Optimal designs are obtained by coupling an optimization algorithm to the database model. Evaluation of the current best design then gives either a new local optima and/or increases the fidelity of the approximation model for the next iteration. Surrogate methods have also been developed that iterate on the selection of data points to decrease the uncertainty of the approximation model prior to searching for an optimal design. The database approximation models for each of these cases, however, become computationally expensive with increase in dimensionality. Thus the method of using optimization algorithms to search a database model becomes problematic as the number of design variables is increased.

Nelson, Andrea↗
DOE OSTI2021

Learning the ground state of a non-stoquastic quantum Hamiltonian in a rugged neural network landscape

Strongly interacting quantum systems described by non-stoquastic Hamiltonians exhibit rich low-temperature physics. Yet, their study poses a formidable challenge, even for state-of-the-art numerical techniques. Here, we investigate systematically the performance of a class of universal variational wave-functions based on artificial neural networks, by considering the frustrated spin- 1/2 1 / 2 J_1-J_2 J 1 − J 2 Heisenberg model on the square lattice. Focusing on neural network architectures without physics-informed input, we argue in favor of using an ansatz consisting of two decoupled real-valued networks, one for the amplitude and the other for the phase of the variational wavefunction. By introducing concrete mitigation strategies against inherent numerical instabilities in the stochastic reconfiguration algorithm we obtain a variational energy comparable to that reported recently with neural networks that incorporate knowledge about the physical system. Through a detailed analysis of the individual components of the algorithm, we conclude that the rugged nature of the energy landscape constitutes the major obstacle in finding a satisfactory approximation to the ground state wavefunction, and prevents learning the correct sign structure. In particular, we show that in the present setup the neural network expressivity and Monte Carlo sampling are not primary limiting factors.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS2023

Prediction of Stiffness and Fatigue Lives of Polymer Matrix Composite Laminates Using Artificial Neural Networks

Machine learning (ML) models are increasingly being used in many engineering fields due to the advancements in ML algorithms and availability of high-speed computing power. One of the most popular ML class of models is artificial neural networks (ANN). ML is increasingly being used in the design and analysis of composite materials and structures, specifically in the constitutive modeling of composite materials with the focus on greatly accelerating multiscale analyses of composite materials and structures through development of surrogate models. Towards that end, both Python and MATLAB-based neural nets have been developed to predict initial stiffness and fatigue life of an eight-ply symmetric polymer matrix composite laminate. Two types of neural networks, a Multilayer Perceptron (MLP) and a Recurrent Neural Network (RNN), have been developed for both platforms. Results show that the both neural net types can provide an excellent estimate of initial stiffness as well as fatigue life of eight-ply symmetric polymer matrix composite laminate. RNNs are better able to capture the shape of the fatigue curve of a laminate. This tool can be very useful for system level studies to obtain an estimate of desired properties and life of PMC composite laminates. The associated surrogate models could also be used in composite multiscale analyses to replace the actual physics-based calculations at lower scales and thereby significantly increase the computational efficiency of such analyses and thus make multiscale analyses a viable industrial tool for large scale structural problems.

Composite↗
NASA NTRS1994

On-line, adaptive state estimator for active noise control

Dynamic characteristics of airframe structures are expected to vary as aircraft flight conditions change. Accurate knowledge of the changing dynamic characteristics is crucial to enhancing the performance of the active noise control system using feedback control. This research investigates the development of an adaptive, on-line state estimator using a neural network concept to conduct active noise control. In this research, an algorithm has been developed that can be used to estimate displacement and velocity responses at any locations on the structure from a limited number of acceleration measurements and input force information. The algorithm employs band-pass filters to extract from the measurement signal the frequency contents corresponding to a desired mode. The filtered signal is then used to train a neural network which consists of a linear neuron with three weights. The structure of the neural network is designed as simple as possible to increase the sampling frequency as much as possible. The weights obtained through neural network training are then used to construct the transfer function of a mode in z-domain and to identify modal properties of each mode. By using the identified transfer function and interpolating the mode shape obtained at sensor locations, the displacement and velocity responses are estimated with reasonable accuracy at any locations on the structure. The accuracy of the response estimates depends on the number of modes incorporated in the estimates and the number of sensors employed to conduct mode shape interpolation. Computer simulation demonstrates that the algorithm is capable of adapting to the varying dynamic characteristics of structural properties. Experimental implementation of the algorithm on a DSP (digital signal processing) board for a plate structure is underway. The algorithm is expected to reach the sampling frequency range of about 10 kHz to 20 kHz which needs to be maintained for a typical active noise control application.

Lim, Tae W.↗
NASA NTRS2016

A New Operational Snow Retrieval Algorithm Applied to Historical AMSR-E Brightness Temperatures

Snow is a key element of the water and energy cycles and the knowledge of spatio-temporal distribution of snow depth and snow water equivalent (SWE) is fundamental for hydrological and climatological applications. SWE and snow depth estimates can be obtained from spaceborne microwave brightness temperatures at global scale and high temporal resolution (daily). In this regard, the data recorded by the Advanced Microwave Scanning Radiometer-Earth Orbiting System (EOS) (AMSR-E) onboard the National Aeronautics and Space Administration's (NASA) AQUA spacecraft have been used to generate operational estimates of SWE and snow depth, complementing estimates generated with other microwave sensors flying on other platforms. In this study, we report the results concerning the development and assessment of a new operational algorithm applied to historical AMSR-E data. The new algorithm here proposed makes use of climatological data, electromagnetic modeling and artificial neural networks for estimating snow depth as well as a spatio-temporal dynamic density scheme to convert snow depth to SWE. The outputs of the new algorithm are compared with those of the current AMSR-E operational algorithm as well as in-situ measurements and other operational snow products, specifically the Canadian Meteorological Center (CMC) and GlobSnow datasets. Our results show that the AMSR-E algorithm here proposed generally performs better than the operational one and addresses some major issues identified in the spatial distribution of snow depth fields associated with the evolution of effective grain size.

passive microwave↗
DOE OSTI2023

Regional Heatwave Prediction Using Graph Neural Network and Weather Station Data

Abstract Heatwaves lead to catastrophic consequences on public health and the economy. Accurate and timely predictions of regional heatwaves can improve climate preparedness and foster decision‐making to alleviate the burdens due to climate change. In this paper, we propose a heatwave prediction algorithm based on a novel deep learning model, that is, Graph Neural Network (GNN). This new GNN framework can provide real time warnings of the sudden occurrence of regional heatwaves with high accuracy at lower costs of computation and data collection. In addition, its interpretable structure unravels the spatiotemporal patterns of regional heatwaves and helps to enrich our understanding of the general climate dynamics and the causal influences between locations. The proposed GNN framework can be applied for the detection and prediction of other extreme or compound climate events, which calls for future studies.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Development of neural network force fields for corrosion studies

To fully understand the chemistry and physics of corrosion, novel methods of simulation must be developed. One approach is designing machine learning (ML) algorithms integrated with density functional theory to develop adaptive force fields to gain insight into corrosion behavior namely at the surface of metal oxides. Current methods of modeling corrosion are slow due to the computational cost of resolving both reaction mechanics and mass transport processes. Machine learning methods can be implemented to obtain structure-activity relationships at both the molecular and bulk scale while still retaining the accuracy of density functional theory (DFT) and significantly decreasing the time needed for simulations of complex chemical processes in the various environments of corrosion. Multiscale models are needed for corrosion studies to fully understand its processes not only at the atomic length scale (chemical bonding, energies, and forces), but also at the nano and meso length scales (solid-state physics and material science processes). Current methods of study include DFT, molecular dynamics, and Monte Carlo. The limitation of DFT is that only a small number of atoms or molecules can be simulated at that level of theory. Density functional theory is used to study the electronic structure of atoms and molecules, and calculate the force component of each atom. However, these calculations are limited to about 1000 atoms. Custom periodic boundary conditions (PBC) can be used to describe the various environments and defects that affect the atomic forces to produce a large data set from which a training set can be derived. Machine learning can be utilized to overcome the barrier of modeling macroscopic and multi-scale processes from ab initio calculations through the development of adaptive force fields. Local environments determine the atomic forces of a given system, therefore adaptive force fields must be created to produce reliable quantum mechanical calculations. This can be achieved by developing a learning algorithm that uses the mapped atomic forces or fingerprint as an input to produce energies and magnetic moments as output. A systematic approach was used to begin to build a data set in order to accurately describe the atomic forces in various environments. In Figure 4 below, a simple PBC cell of Fe{sub 2}O{sub 3} was first optimized. A surface optimization was performed next, followed by a hydroxylated surface optimization. Once this calculation has converged, the adsorption of halide species to the hydroxylated surface will be investigated. TensorFlow is an open source platform for machine learning developed by Google. Using a high level application program interface (API) such as Keras allows for building and training ML models easily in a number of different environments and languages. For this project, a neural network was developed within Anaconda in Python. Future Work: Further development of reference data set; Refining neural network and learning algorithm; Fingerprinting atomic environment to enable mapping of atomic force components; Choosing appropriate training set from reference data; Learning from training set and enabling non-linear mapping of training set fingerprints and the atomic forces; Estimation of uncertainty to identify ranges of outside applicability; Testing and analysis of molecular dynamic simulations.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Neutrino interaction vertex reconstruction in DUNE with Pandora deep learning

The Pandora Software Development Kit and algorithm libraries perform reconstruction of neutrino interactions in liquid argon time projection chamber detectors. Pandora is the primary event reconstruction software used at the Deep Underground Neutrino Experiment, which will operate four large-scale liquid argon time projection chambers at the far detector site in South Dakota, producing high-resolution images of charged particles emerging from neutrino interactions. While these high-resolution images provide excellent opportunities for physics, the complex topologies require sophisticated pattern recognition capabilities to interpret signals from the detectors as physically meaningful objects that form the inputs to physics analyses. A critical component is the identification of the neutrino interaction vertex. Subsequent reconstruction algorithms use this location to identify the individual primary particles and ensure they each result in a separate reconstructed particle. A new vertex-finding procedure described in this article integrates a U-ResNet neural network performing hit-level classification into the multi-algorithm approach used by Pandora to identify the neutrino interaction vertex. The machine learning solution is seamlessly integrated into a chain of pattern-recognition algorithms. The technique substantially outperforms the previous BDT-based solution, with a more than 20% increase in the efficiency of sub-1 cm vertex reconstruction across all neutrino flavours.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗