Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “layered 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 541 records · Page 30
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Extreme Energy Dissipation via Material Evolution in Carbon Nanotube Mats

Abstract Thin layered mats comprised of an interconnected meandering network of multiwall carbon nanotubes (MWCNT) are subjected to a hypersonic micro‐projectile impact test. The mat morphology is highly compliant and while this leads to rather modest quasi‐static mechanical properties, at the extreme strain rates and large strains resulting from ballistic impact, the MWCNT structure has the ability to reconfigure resulting in extraordinary kinetic energy (KE) absorption. The KE of the projectile is dissipated via frictional interactions, adiabatic heating, tube stretching, and ultimately fracture of taut tubes and the newly formed fibrils. The energy absorbed per unit mass of the film can range from 7–12 MJ kg −1 , much greater than any other material.

36 MATERIALS SCIENCE↗
DOE OSTI2024

HDBind: encoding of molecular structure with hyperdimensional binary representations

Traditional methods for identifying “hit” molecules from a large collection of potential drug-like candidates rely on biophysical theory to compute approximations to the Gibbs free energy of the binding interaction between the drug and its protein target. These approaches have a significant limitation in that they require exceptional computing capabilities for even relatively small collections of molecules. Increasingly large and complex state-of-the-art deep learning approaches have gained popularity with the promise to improve the productivity of drug design, notorious for its numerous failures. However, as deep learning models increase in their size and complexity, their acceleration at the hardware level becomes more challenging. Hyperdimensional Computing (HDC) has recently gained attention in the computer hardware community due to its algorithmic simplicity relative to deep learning approaches. The HDC learning paradigm, which represents data with high-dimension binary vectors, allows the use of low-precision binary vector arithmetic to create models of the data that can be learned without the need for the gradient-based optimization required in many conventional machine learning and deep learning methods. This algorithmic simplicity allows for acceleration in hardware that has been previously demonstrated in a range of application areas (computer vision, bioinformatics, mass spectrometery, remote sensing, edge devices, etc.). To the best of our knowledge, our work is the first to consider HDC for the task of fast and efficient screening of modern drug-like compound libraries. We also propose the first HDC graph-based encoding methods for molecular data, demonstrating consistent and substantial improvement over previous work. We compare our approaches to alternative approaches on the well-studied MoleculeNet dataset and the recently proposed LIT-PCBA dataset derived from high quality PubChem assays. We demonstrate our methods on multiple target hardware platforms, including Graphics Processing Units (GPUs) and Field Programmable Gate Arrays (FPGAs), showing at least an order of magnitude improvement in energy efficiency versus even our smallest neural network baseline model with a single hidden layer. Our work thus motivates further investigation into molecular representation learning to develop ultra-efficient pre-screening tools. We make our code publicly available at https://github.com/LLNL/hdbind.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

The Artificial Intelligence Ontology: LLM-Assisted Construction of AI Concept Hierarchies

The Artificial Intelligence Ontology (AIO) is a systematization of artificial intelligence (AI) concepts, methodologies, and their interrelations. Developed via manual curation, with the additional assistance of large language models (LLMs), AIO aims to address the rapidly evolving landscape of AI by providing a comprehensive framework that encompasses both technical and ethical aspects of AI technologies. The primary audience for AIO includes AI researchers, developers, and educators seeking standardized terminology and concepts within the AI domain. We use the term “branches” for classes, and their subclasses, in our ontology that are subclasses of owl:Thing. AIO contains eight branches: Bias, Layer, Machine Learning Task, Mathematical Function, Model, Network, Preprocessing, and Training Strategy, each designed to support the modular composition of AI methods and facilitate a deeper understanding of deep learning architectures and ethical considerations in AI. AIO uses the Ontology Development Kit (ODK) for its creation and maintenance, with its content being more easily updated through AI-driven curation support. This approach not only ensures the ontology's relevance amidst the fast-paced advancements in AI but also significantly enhances its utility for researchers, developers, and educators by simplifying the integration of new AI concepts and methodologies. The ontology's utility is demonstrated through the annotation of AI methods data in a catalog of AI research publications and the integration into the BioPortal ontology resource, highlighting its potential for cross-disciplinary research. The AIO ontology is open source and is available on GitHub ( https://w3id.org/aio/ ) and BioPortal ( https://bioportal.bioontology.org/ontologies/AIO ).

Joachimiak, Marcin P. [Biosystems Data Science Dep↗
DOE OSTI2024

Z-Target Radiography Postprocessing With A Deep Convolution Neural Network

Analyzing X-ray radiographs is crucial for understanding target behavior in Inertial Confinement Fusion (ICF) and High Energy Density (HED) platforms. However, the density of Magneto Raleigh Taylor (MRT) bands and limitations of target materials often obscure relevant spike growth and density information. To address this issue, machine learning postprocessing techniques can be applied to remove darkened regions in radiography images. In this study, a novel method is presented for removing MRT darkened regions from z-target radiographs using a convolutional neural network (CNN). The CNN, consisting of six layers, treats the darkened regions as noise and employs a mixed loss function and end-to-end frameworks to suppress them while preserving sharpness. The six-layer architecture is designed to effectively learn features when provided with a larger volume of learning space. Each layer is optimized using a mixed loss function that combines a standard loss pixel approach with a multi-scaled structural similarity index loss, which considers luminance, contrast, and structure in local neighborhoods. This approach is particularly beneficial for capturing the stochastic structure of MRT limbs. Due to the limited availability of experimental data, training is conducted using synthetic target radiography from 3D Alegra simulations.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
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↗
NASA NTRS1988

Local area networking: Ames centerwide network

A computer network can benefit the user by making his/her work quicker and easier. A computer network is made up of seven different layers with the lowest being the hardware, the top being the user, and the middle being the software. These layers are discussed.

Price, Edwin↗
NASA NTRS1989

Error detecting capabilities of the shortened Hamming codes adopted for error detection in IEEE Standard 802.3

The error-detecting capabilities of the shortened Hamming codes adopted for error detection in IEEE Standard 802.3 are investigated. These codes are also used for error detection in the data link layer of the Ethernet, a local area network. The weight distributions for various code lengths are calculated to obtain the probability of undetectable error and that of detectable error for a binary symmetric channel with bit-error rate between 0.00001 and 1/2.

Fujiwara, Toru↗
NASA NTRS1994

Comparisons of neural networks to standard techniques for image classification and correlation

Neural network techniques for multispectral image classification and spatial pattern detection are compared to the standard techniques of maximum-likelihood classification and spatial correlation. The neural network produced a more accurate classification than maximum-likelihood of a Landsat scene of Tucson, Arizona. Some of the errors in the maximum-likelihood classification are illustrated using decision region and class probability density plots. As expected, the main drawback to the neural network method is the long time required for the training stage. The network was trained using several different hidden layer sizes to optimize both the classification accuracy and training speed, and it was found that one node per class was optimal. The performance improved when 3x3 local windows of image data were entered into the net. This modification introduces texture into the classification without explicit calculation of a texture measure. Larger windows were successfully used for the detection of spatial features in Landsat and Magellan synthetic aperture radar imagery.

Paola, Justin D.↗
NASA NTRS1993

AVIRIS and TIMS data processing and distribution at the land processes distributed active archive center

The U.S. Government has initiated the Global Change Research program, a systematic study of the Earth as a complete system. NASA's contribution of the Global Change Research Program is the Earth Observing System (EOS), a series of orbital sensor platforms and an associated data processing and distribution system. The EOS Data and Information System (EOSDIS) is the archiving, production, and distribution system for data collected by the EOS space segment and uses a multilayer architecture for processing, archiving, and distributing EOS data. The first layer consists of the spacecraft ground stations and processing facilities that receive the raw data from the orbiting platforms and then separate the data by individual sensors. The second layer consists of Distributed Active Archive Centers (DAAC) that process, distribute, and archive the sensor data. The third layer consists of a user science processing network. The EOSDIS is being developed in a phased implementation. The initial phase, Version 0, is a prototype of the operational system. Version 0 activities are based upon existing systems and are designed to provide an EOSDIS-like capability for information management and distribution. An important science support task is the creation of simulated data sets for EOS instruments from precursor aircraft or satellite data. The Land Processes DAAC, at the EROS Data Center (EDC), is responsible for archiving and processing EOS precursor data from airborne instruments such as the Thermal Infrared Multispectral Scanner (TIMS), the Thematic Mapper Simulator (TMS), and Airborne Visible and Infrared Imaging Spectrometer (AVIRIS). AVIRIS, TIMS, and TMS are flown by the NASA-Ames Research Center ARC) on an ER-2. The ER-2 flies at 65000 feet and can carry up to three sensors simultaneously. Most jointly collected data sets are somewhat boresighted and roughly registered. The instrument data are being used to construct data sets that simulate the spectral and spatial characteristics of the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) instrument scheduled to be flown on the first EOS-AM spacecraft. The ASTER is designed to acquire 14 channels of land science data in the visible and near-IR (VNIR), shortwave-IR (SWIR), and thermal-IR (TIR) regions from 0.52 micron to 11.65 micron at high spatial resolutions of 15 m to 90 m. Stereo data will also be acquired in the VNIR region in a single band. The AVIRIS and TMS cover the ASTER VNIR and SWIR bands, and the TIMS covers the TIR bands. Simulated ASTER data sets have been generated over Death Valley, California, Cuprite, Nevada, and the Drum Mountains, Utah using a combination of AVIRIS, TIMS, amd TMS data, and existing digital elevation models (DEM) for the topographic information.

Mah, G. R.↗
NASA NTRS2005

System for Processing Coded OFDM Under Doppler and Fading

An advanced communication system has been proposed for transmitting and receiving coded digital data conveyed as a form of quadrature amplitude modulation (QAM) on orthogonal frequency-division multiplexing (OFDM) signals in the presence of such adverse propagation-channel effects as large dynamic Doppler shifts and frequency-selective multipath fading. Such adverse channel effects are typical of data communications between mobile units or between mobile and stationary units (e.g., telemetric transmissions from aircraft to ground stations). The proposed system incorporates novel signal processing techniques intended to reduce the losses associated with adverse channel effects while maintaining compatibility with the high-speed physical layer specifications defined for wireless local area networks (LANs) as the standard 802.11a of the Institute of Electrical and Electronics Engineers (IEEE 802.11a). OFDM is a multi-carrier modulation technique that is widely used for wireless transmission of data in LANs and in metropolitan area networks (MANs). OFDM has been adopted in IEEE 802.11a and some other industry standards because it affords robust performance under frequency-selective fading. However, its intrinsic frequency-diversity feature is highly sensitive to synchronization errors; this sensitivity poses a challenge to preserve coherence between the component subcarriers of an OFDM system in order to avoid intercarrier interference in the presence of large dynamic Doppler shifts as well as frequency-selective fading. As a result, heretofore, the use of OFDM has been limited primarily to applications involving small or zero Doppler shifts. The proposed system includes a digital coherent OFDM communication system that would utilize enhanced 802.1la-compatible signal-processing algorithms to overcome effects of frequency-selective fading and large dynamic Doppler shifts. The overall transceiver design would implement a two-frequency-channel architecture (see figure) that would afford frequency diversity for reducing the adverse effects of multipath fading. By using parallel concatenated convolutional codes (also known as Turbo codes) across the dual-channel and advanced OFDM signal processing within each channel, the proposed system is intended to achieve at least an order of magnitude improvement in received signal-to-noise ratio under adverse channel effects while preserving spectral efficiency.

Tsou, Haiping↗
NASA NTRS2014

Enabling a Science Support Structure for NASAs Global Hawk UASs

In this paper we describe the information technologies developed by NASA for the Winter/Spring 2013/2014, and Fall 2014, NASA Earth Venture Campaigns, Hurricane and Severe Storm Sentinel (HS3) and Airborne Tropical TRopopause EXperiment (ATTREX). These campaigns utilized Global Hawk UAS vehicles equipped at the NASA Armstrong (previously Dryden) Flight Research Facility (AFRC), Edwards Air Force Base, California, and operated from there, the NASA Wallops Flight Facility (WFF), Virginia, and Anderson Air Force Base (AAFB), Guam. Part of this enabling infrastructure utilized a layer 2 encrypted terrestrial Virtual Local Area Network (VLAN) that, at times, spanned greater than ten thousand miles (AAFB <-> AFRC <-> WFF) and was routed over geosynchronous Ku band communication Satellites directly to the aircraft sensor network. This infrastructure enabled seamless hand off between Satellites, and Satellite ground stations in Guam, California and Virginia, so allowing simultaneous Aircraft Command and Control and Science operations from remote locations. Additionally, we will describe the other elements of this infrastructure, from on-board geo-enabled databases, to real time communications directly from the instruments (in some cases, more than twelve were carried, and simultaneously operated, on one aircraft) to the researchers and other interested parties, world wide.

Virtual Local Area Network (VLAN)↗
NASA NTRS2022

A Novel Machine Learning-Based Gap-Filling of Fine-Resolution Remotely Sensed Snow Cover Fraction Data By Combining Downscaling and Regression

Satellite-based remotely sensed observations of snow cover fraction (SCF) can have data gaps in spatially distributed coverage from sensor and orbital limitations. We mitigate these limitations in the example fine-resolution Moderate Resolution Imaging Spectroradiometer (MODIS) data by gap-filling using auxiliary 1-km datasets that either aid in downscaling from coarser-resolution (5 km) MODIS SCF wherever not fully covered by clouds, or else by themselves via regression wherever fully cloud covered. This study’s prototype predicts a 1-km version of the 500-m MOD10A1 SCF target. Due to noncollocatedness of spatial gaps even across input and auxiliary datasets, we consider a recent gap-agnostic advancement of partial convolution in computer vision for both training and predictive gap-filling. Partial convolution accommodates spatially consistent gaps across the input images, effectively implementing a two-dimensional masking. To overcome reduced usable data from noncollocated spatial gaps across inputs, we innovate a fully generalized three-dimensional masking in this partial convolution. This enables a valid output value at a pixel even if only a single valid input variable and its value exist in the neighborhood covered by the convolutional filter zone centered around that pixel. Thus, our gap-agnostic technique can use significantly more examples for training (∼67%) and prediction (∼100%), instead of only less than 10% for the previous partial convolution. We train an example simple three-layer legacy super-resolution convolutional neural network (SRCNN) to obtain downscaling and regression component performances that are better than baseline values of either climatology or MOD10C1 SCF as relevant. Our generalized partial convolution can enable multiple Earth science applications like downscaling, regression, classification, and segmentation that were hindered by data gaps.

Soni Yatheendradas↗
DOE OSTI2021

Impact of Blockchain Delay on Grid-Tied Solar Inverter Performance: Preprint

This paper investigates the impact of the delay resulting from a blockchain, a promising security measure, for a hierarchical control system of inverters connected to the grid. The blockchain communication network is designed at the secondary control layer for resilience against cyberattacks. To represent the latency in the communication channel, a model is developed based on the complexity of the blockchain framework. Taking this model into account, this work evaluates the plant’s performance subject to communication delays, introduced by the blockchain, among the hierarchical control agents. In addition, this article considers an optimal model-based control strategy that performs the system’s internal control loop. The work shows that the blockchain’s delay size influences the convergence of the power supplied by the inverter to the reference at the point of common coupling. In the results section, real-time simulations on OPAL-RT are performed to test the resilience of two parallel inverters with increasing blockchain complexity.

41 EE - Solar Energy Technologies Office (EE-4S)↗
DOE OSTI2021

Traction-separation response of bilayer graphene interfaces: The role of an intercalated single atomic layer of water molecules and hydroxyl groups

Molecular dynamics simulations of the normal and shear traction-separation response of dry and wet bilayer graphene interfaces were performed. The influence of a single, atomic layer of water molecules sandwiched between graphene layers and hydroxyl functionalization on one or both layers were investigated. Hydroxyl groups provided a shielding effect during normal separation while favoring stick-slip friction under sliding conditions. The single, atomic layer of water molecules confined between hydrophilic and asymmetric hydrophobic/hydrophilic surfaces was found to ease sliding friction. The mechanisms arose from the balance between the attractive and repulsive graphene/water interactions through bridges of single water molecules between the two surfaces and the structure of the water monolayer under confinement for which an out-of-plane water/water hydrogen bond network could not be developed. The single, atomic layer of water molecules reduced the sliding friction the most when both surfaces were hydrophilic and lowered the sliding friction of the asymmetric hydrophobic/hydrophilic interfaces to that when both surfaces were hydrophobic. An optimal hydroxyl surface coverage of less than 10% may exist for efficient water lubrication, with diminishing effect with increasing number of hydroxyl groups. Finally, the findings offer a design strategy for assembling bilayer graphene interfaces with tailored friction and sliding contact properties.

36 MATERIALS SCIENCE↗
NASA NTRS1993

Adaptive nonlinear polynomial neural networks for control of boundary layer/structural interaction

The acoustic pressures developed in a boundary layer can interact with an aircraft panel to induce significant vibration in the panel. Such vibration is undesirable due to the aerodynamic drag and structure-borne cabin noises that result. The overall objective of this work is to develop effective and practical feedback control strategies for actively reducing this flow-induced structural vibration. This report describes the results of initial evaluations using polynomial, neural network-based, feedback control to reduce flow induced vibration in aircraft panels due to turbulent boundary layer/structural interaction. Computer simulations are used to develop and analyze feedback control strategies to reduce vibration in a beam as a first step. The key differences between this work and that going on elsewhere are as follows: that turbulent and transitional boundary layers represent broadband excitation and thus present a more complex stochastic control scenario than that of narrow band (e.g., laminar boundary layer) excitation; and secondly, that the proposed controller structures are adaptive nonlinear infinite impulse response (IIR) polynomial neural network, as opposed to the traditional adaptive linear finite impulse response (FIR) filters used in most studies to date. The controllers implemented in this study achieved vibration attenuation of 27 to 60 dB depending on the type of boundary layer established by laminar, turbulent, and intermittent laminar-to-turbulent transitional flows. Application of multi-input, multi-output, adaptive, nonlinear feedback control of vibration in aircraft panels based on polynomial neural networks appears to be feasible today. Plans are outlined for Phase 2 of this study, which will include extending the theoretical investigation conducted in Phase 2 and verifying the results in a series of laboratory experiments involving both bum and plate models.

Parker, B. Eugene, Jr.↗
NASA NTRS2007

Identification and Control of Aircrafts using Multiple Models and Adaptive Critics

We compared two possible implementations of local linear models for control: one approach is based on a self-organizing map (SOM) to cluster the dynamics followed by a set of linear models operating at each cluster. Therefore the gating function is hard (a single local model will represent the regional dynamics). This simplifies the controller design since there is a one to one mapping between controllers and local models. The second approach uses a soft gate using a probabilistic framework based on a Gaussian Mixture Model (also called a dynamic mixture of experts). In this approach several models may be active at a given time, we can expect a smaller number of models, but the controller design is more involved, with potentially better noise rejection characteristics. Our experiments showed that the SOM provides overall best performance in high SNRs, but the performance degrades faster than with the GMM for the same noise conditions. The SOM approach required about an order of magnitude more models than the GMM, so in terms of implementation cost, the GMM is preferable. The design of the SOM is straight forward, while the design of the GMM controllers, although still reasonable, is more involved and needs more care in the selection of the parameters. Either one of these locally linear approaches outperform global nonlinear controllers based on neural networks, such as the time delay neural network (TDNN). Therefore, in essence the local model approach warrants practical implementations. In order to call the attention of the control community for this design methodology we extended successfully the multiple model approach to PID controllers (still today the most widely used control scheme in the industry), and wrote a paper on this subject. The echo state network (ESN) is a recurrent neural network with the special characteristics that only the output parameters are trained. The recurrent connections are preset according to the problem domain and are fixed. In a nutshell, the states of the reservoir of recurrent processing elements implement a projection space, where the desired response is optimally projected. This architecture trades training efficiency by a large increase in the dimension of the recurrent layer. However, the power of the recurrent neural networks can be brought to bear on practical difficult problems. Our goal was to implement an adaptive critic architecture implementing Bellman s approach to optimal control. However, we could only characterize the ESN performance as a critic in value function evaluation, which is just one of the pieces of the overall adaptive critic controller. The results were very convincing, and the simplicity of the implementation was unparalleled.

Principe, Jose C.↗
DOE OSTI2021

Effects of Ink Formulation on Construction of Catalyst Layers for High-Performance Polymer Electrolyte Membrane Fuel Cells

Rational design of catalyst layers in a membrane electrode assembly (MEA) is crucial for achieving high-performance polymer electrolyte membrane fuel cells. Establishing a clear understanding of the property (catalyst ink)–structure (catalyst layer)–performance (MEA) relationship lays the foundation for this rational design. Here, a synergistic approach was taken to correlate the ink formulation, the microstructure of catalyst layers, and the resulting MEA performance to establish such a property–structure–performance relationship. The solvent composition (n-PA/H 2 O mixtures) demonstrated a strong influence on the performance of the MEA fabricated with an 830-EW (Aquivion) ionomer, especially polarization losses of cell activation and mass transport. The performance differences were studied in terms of how the solvent composition affects the catalyst/ionomer interface, ionomer network, and pore structure of the resulting catalyst layers. The ionomer aggregates mainly covered the surface of catalyst aggregates acting as oxygen reduction reaction active sites, and the aggregate sizes of the ionomer and catalyst (revealed by ultrasmall angle X-ray scattering and cryo-transmission electron microscopy) were dictated by tuning the solvent composition, which in turn determined the catalyst/ionomer interface (available active sites). In n-PA/H 2 O mixtures with 50~90 wt % H 2 O, the catalyst agglomerates could be effectively broken up into small aggregates, leading to enhanced kinetic activities. The boiling point of the mixed solvents determined the pore structure of ultimate catalyst layers, as evidenced by mercury porosimetry and scanning electron microscopy. For mixed solvents with a higher boiling point, the catalyst–ionomer aggregates in the ink tend to agglomerate during the solvent evaporation process and finally form larger catalyst–ionomer aggregates in the ultimate catalyst layer, resulting in more secondary pores and thus lower mass transport resistance. Both the enlarged catalyst/ionomer interface and appropriate pore structure were achieved with the catalyst layer fabricated from an n-PA/H 2 O mixture with 90 wt % H 2 O, leading to the best MEA performance.

25 ENERGY STORAGE↗
DOE OSTI2024

Adaptive Data-Driven Deep-Learning Surrogate Model for Frontal Polymerization in Dicyclopentadiene

Frontal polymerization (FP) is a self-sustaining curing process that enables rapid and energy-efficient manufacturing of thermoset polymers and composites. Computational methods conventionally used to simulate the FP process are time-consuming, and repeating simulations are required for sensitivity analysis, uncertainty quantification, or optimization of the manufacturing process. Here, in this work, we develop an adaptive surrogate deep-learning model for FP of dicyclopentadiene (DCPD), which predicts the evolution of temperature and degree of cure orders of magnitude faster than the finite-element method (FEM). The adaptive algorithm provides a strategy to select training samples efficiently and save computational costs by reducing the redundancy of FEM-based training samples. The adaptive algorithm calculates the residual error of the FP governing equations using automatic differentiation of the deep neural network. A probability density function expressed in terms of the residual error is used to select training samples from the Sobol sequence space. The temperature and degree of cure evolution of each training sample are obtained by a 2D FEM simulation. The adaptive method is more efficient and has a better prediction accuracy than the random sampling method. With the well-trained surrogate neural network, the FP characteristics (front speed, shape, and temperature) can be extracted quickly from the predicted temperature and degree-of-cure fields.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗