Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “hybrid algorithm”

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 919 records · Page 51
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Disentangling Electronic Transport and Hysteresis at Individual Grain Boundaries in Hybrid Perovskites via Automated Scanning Probe Microscopy

Underlying the rapidly increasing photovoltaic efficiency and stability of metal halide perovskites (MHPs) is the advancement in the understanding of the microstructure of polycrystalline MHP thin film. Over the past decade, intense efforts have been aimed at understanding the effect of microstructures on MHP properties, including chemical heterogeneity, strain disorder, phase impurity, etc. It has been found that grain and grain boundary (GB) are tightly related to lots of microscale and nanoscale behavior in MHP thin films. Atomic force microscopy (AFM) is widely used to observe grain and boundary structures in topography and subsequently to study the correlative surface potential and conductivity of these structures. For now, most AFM measurements have been performed in imaging mode to study the static behavior; in contrast, AFM spectroscopy mode allows us to investigate the dynamic behavior of materials, e.g., conductivity under sweeping voltage. However, a major limitation of AFM spectroscopy measurements is that they require manual operation by human operators, and as such only limited data can be obtained, hindering systematic investigations of these microstructures. Here, we designed a workflow combining the conductive AFM measurement with a machine learning (ML) algorithm to systematically investigate grain boundaries in MHPs. The trained ML model can extract GBs locations from the topography image, and the workflow drives the AFM probe to each GB location to perform a current–voltage (IV) curve automatically. Then, we are able to have IV curves at all GB locations, allowing us to systematically understand the property of GBs. Using this method, we discovered that the GB junction points are less conductive, potentially more photoactive, and can play critical roles in MHP stability, while most previous works only focused on the difference between GB and grains.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Enhancing molecular design efficiency: Uniting language models and generative networks with genetic algorithms

This study examines the effectiveness of generative models in drug discovery, material science, and polymer science, aiming to overcome constraints associated with traditional inverse design methods relying on heuristic rules. Generative models generate synthetic data resembling real data, enabling deep learning model training without extensive labeled datasets. They prove valuable in creating virtual libraries of molecules for material science and facilitating drug discovery by generating molecules with specific properties. While generative adversarial networks (GANs) are explored for these purposes, mode collapse restricts their efficacy, limiting novel structure variability. To address this, we introduce a masked language model (LM) inspired by natural language processing. Although LMs alone can have inherent limitations, we propose a hybrid architecture combining LMs and GANs to efficiently generate new molecules, demonstrating superior performance over standalone masked LMs, particularly for smaller population sizes. This hybrid LM-GAN architecture enhances efficiency in optimizing properties and generating novel samples.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Federated Deep Reinforcement Learning for Decentralized VVO of BTM DERs

The future of grid control requires a hybrid approach combining centralized and decentralized methods to fully utilize the potential of smart edge devices with artificial intelligence (AI) capabilities. This paper aims to develop and evaluate a federated deep reinforcement learning (FDRL) framework for decentralized adaptive volt-var optimization (VVO) of behind-the-meter (BTM) distributed energy resources (DERs). First, this paper models a single deep reinforcement learning (DRL) agent using the Markov Decision Process (MDP) framework for decentralized adaptive VVO of BTM DERs. Two DRL algorithms, soft actor-critic (SAC) and twin-delayed deep deterministic policy gradient (TD3), are compared for their effectiveness in optimizing VVO. Results show that TD3 outperforms SAC, achieving a 71.3% improvement in mean reward. Finally, the DRL agent is deployed within the FDRL framework, using the Flower platform, to enhance learning, provide adaptive control, and ensure data privacy for BTM DERs.

Ravi, Abhijith↗
DOE OSTI2025

Hybrid data-driven cement-stabilized soil design: An integration of machine learning, multi-objective optimization, and life cycle assessment

Soil stabilization is crucial in geotechnical engineering, yet conventional methods are often time-consuming, resource-intensive, and environmentally unsustainable. Despite growing interest in Machine Learning (ML) and optimization tools for mix design, few studies integrate these methods with decision-making techniques and environmental assessment to support practical implementation. This study proposes a hybrid data-driven framework for predicting strength, optimizing mix compositions, and evaluating environmental impacts via life cycle assessment of cement-stabilized soft soils. Six ML models were evaluated, and the top-performing eXtreme Gradient Boosting (XGB) model was further improved using the Grey Wolf Optimizer (GWO). The optimized XGB-GWO model, integrated with a polynomial cost function, served as the objective function in a multi-objective optimization problem solved via the Non-Dominated Sorting Genetic Algorithm II (NSGA-II), with final mix selection guided by the entropy-weighted TOPSIS method. Validation through a case study produced mix designs offering superior strength-cost trade-offs, with the optimal mix achieving 2243.2 kPa unconfined compressive strength and a 16.07 % reduction in carbon emissions compared to the highest-cost design. In conclusion, this study offers a sustainable, scalable approach to soil stabilization and supports informed decision-making in construction.

Life cycle assessment↗
DOE OSTI2023

High-Impedance Non-Linear Fault Detection via Eigenvalue Analysis with low PMU Sampling Rates

This work presents a hybrid data-driven and physics-based framework for high-impedance fault detection in power systems. An innovative method based on eigenvalue analysis is expanded and validated. Phasor Measurement Unit data is used to estimate eigenvalues corresponding to the powerlines being monitored. The projection and drift of these eigenvalues is then tracked and evaluated. Faults are detected as they drive eigenvalues outside of their normal zones. Eigenvectors are leveraged to support and validate the decisions made by the main algorithm. This technique holds several advantages over contemporary techniques in that it utilizes technology that is already deployed in the field, it offers a significant degree of generality, and so far it has displayed a very high-level of sensitivity without sacrificing accuracy. Validation takes place in the form of simulations in the IEEE 13 Node System considering a popular high-impedance non-linear fault model. Test results are encouraging indicating potential for real-life applications.

Paramo, Gian↗
DOE OSTI2024

Medium-Duty Delivery Truck Integrated Bidirectional Wireless Power Transfer System With Grid and Stationary Energy Storage System Connectivity

Electric vehicles (EVs) can provide power to the grid or buildings similar to distributed energy resources (DER) for energy balancing applications or optimizing the operation of the microgrids in harmony with the other DER assets. This article presents the operating modes of a bidirectional wireless power transfer (WPT) system designed for a medium-duty package delivery vehicle. The WPT system designed for this study can transfer 20 kW of power across 11 in of air gap using custom-designed double-D (DD) couplers with LCC–LCC tuning networks. The proposed system utilizes a 480-V three-phase grid connection, a plug-in hybrid delivery truck with bidirectional WPT, and a stationary energy storage system (SESS) that can be connected to the primary-side dc link. Due to the differences in primary and secondary dc bus voltages, and considering the voltage of the SESS, asymmetric voltage gains were used in the system. Sensitivity analyses of this system with respect to these voltage levels are presented. Five different operating modes of the grid, SESS, and the EV battery are investigated with experimental results. Control algorithms are described for grid-to-vehicle (G2V) and vehicle-to-grid (V2G) operating modes. A bidirectional WPT system is operated with a power factor of 0.99 on the grid side in every operating mode. The EV battery was charged with 20.3 kW with an overall efficiency of 93.02% in the G2V operating mode. Finally, in the V2G operating mode, the WPT system provided 12.82 kW of power back to the grid with an overall efficiency of 89.08%.

25 ENERGY STORAGE↗
NASA NTRS2000

Hybrid Active/Passive Control of Sound Radiation from Panels with Constrained Layer Damping and Model Predictive Feedback Control

There has been considerable interest over the past several years in applying feedback control methods to problems of structural acoustics. One problem of particular interest is the control of sound radiation from aircraft panels excited on one side by a turbulent boundary layer (TBL). TBL excitation appears as many uncorrelated sources acting on the panel, which makes it difficult to find a single reference signal that is coherent with the excitation. Feedback methods have no need for a reference signal, and are thus suited to this problem. Some important considerations for the structural acoustics problem include the fact that the required controller bandwidth can easily extend to several hundred Hertz, so a digital controller would have to operate at a few kilohertz. In addition, aircraft panel structures have a reasonably high modal density over this frequency range. A model based controller must therefore handle the modally dense system, or have some way to reduce the bandwidth of the problem. Further complicating the problem is the fact that the stiffness and dynamic properties of an aircraft panel can vary considerably during flight due to altitude changes resulting in significant resonant frequency shifts. These considerations make the tradeoff between robustness to changes in the system being controlled and controller performance especially important. Recent papers concerning the design and implementation of robust controllers for structural acoustic problems highlight the need to consider both performance and robustness when designing the controller. While robust control methods such as H1 can be used to balance performance and robustness, their implementation is not easy and requires assumptions about the types of uncertainties in the plant being controlled. Achieving a useful controller design may require many tradeoff studies of different types of parametric uncertainties in the system. Another approach to achieving robustness to plant changes is to make the controller adaptive. For example, a mathematical model of the plant could be periodically updated as the plant changes, and the feedback gains recomputed from the updated model. To be practical, this approach requires a simple plant model that can be updated quickly with reasonable computational requirements. A recent paper by the authors discussed one way to simplify a feedback controller, by reducing the number of actuators and sensors needed for good performance. The work was done on a tensioned aircraft-style panel excited on one side by TBL flow in a low speed wind tunnel. Actuation was provided by a piezoelectric (PZT) actuator mounted on the center of the panel. For sensing, the responses of four accelerometers, positioned to approximate the response of the first radiation mode of the panel, were summed and fed back through the controller. This single input-single output topology was found to have nearly the same noise reduction performance as a controller with fifteen accelerometers and three PZT patches. This paper extends the previous results by looking at how constrained layer damping (CLD) on a panel can be used to enhance the performance of the feedback controller thus providing a more robust and efficient hybrid active/passive system. The eventual goal is to use the CLD to reduce sound radiation at high frequencies, then implement a very simple, reduced order, low sample rate adaptive controller to attenuate sound radiation at low frequencies. Additionally this added damping smoothes phase transitions over the bandwidth which promotes robustness to natural frequency shifts. Experiments were conducted in a transmission loss facility on a clamped-clamped aluminum panel driven on one side by a loudspeaker. A generalized predictive control (GPC) algorithm, which is suited to online adaptation of its parameters, was used in single input-single output and multiple input-single output configurations. Because this was a preliminary look at the potential constrained layer damping for adaptive control, static feedback control with no online adaptation was used. Two configurations of CLD in addition to a bare panel configuration were studied. For each CLD configuration, two sensor arrangements for the feedback controller were compared. The first arrangement used fifteen accelerometers on the panel to estimate the responses of the first six radiation modes of the panel. The second sensor arrangement was simpler, using the summed responses of only four accelerometers to approximate the response of the first radiation mode of the panel. In all cases a PZT patch was mounted at the center of the panel for control input. The performance of the controller was quantified using the responses of the fifteen accelerometers on the panel to estimate radiated sound power. The paper begins with a brief discussion of the GPC algorithm and the experimental setup. The experimental results are discussed next, comparing the CLD and sensor configurations, followed by discussion and conclusions.

Cabell, Randolph H.↗
DOE OSTI2025

Direct NeTS sampling of nuclear graphite $S(α, β, T)$ in Serpent

For advanced reactor applications, Neural Thermal Scattering (NeTS) modules were developed to predict the thermal scattering law (TSL or $S(α, β, T)$) of a nuclear graphite neutron moderator. NeTS are multi-layer, feedforward artificial neural networks, which act as universal function approximators designed for TSL datasets. In this case, a 4-layer neural network with 164 neurons per layer is trained using FLASSH evaluated data in PyTorch and serialized as a torchscript dictionary to predict $S(α, β, T)$ on-the-fly. Relative, absolute and maximum percent deviations of NeTS from File 7 data generated using the FLASSH code are on the order of 0.01%, 0.1% and 1%, respectively, with low inference latencies of 0.000172 s per $S(α, β, T)$ at a given temperature. Capturing the full dimensionality of possible inelastic neutron-lattice interactions, NeTS functionality is embedded in the Serpent Monte Carlo code, where $S(α, β, T)_{NeTS}$ sampling is conducted on-the-fly and compared to ACE look-up-tables for predicting TREAT criticality. k-eff differences between sampling algorithms of 6 pcm are observed and are within the order of Monte Carlo uncertainty. Compared to discrete and continuous-energy ACE files (30 MB and 131 MB per temperature), the NeTS format is on the order of 200–300 kB for a continuous-temperature, interpolation-free representation of $S(α, β, T)$ and cross sections. NeTS-in-Serpent runtimes comparable with ACE look-up tables are achieved by scaling NeTS for high performance computing architectures with hybrid OpenMP + MPI parallelization. This work validates a novel, self-contained reactor physics framework for predictive cross sections, and demonstrates a general methodology for embedding modern machine learning libraries within existing neutronic analysis frameworks.

Nuclear Criticality Safety Program (NCSP)↗
NASA NTRS1993

Optimization by decomposition in structural and multidisciplinary applications

An algorithm for a general, multilevel structural optimization by substructuring is derived, based on the linear decomposition concept that is rooted in the Bellman's Optimality Criterion enhanced with the optimum sensitivity derivatives used as a means to account for coupling among the subproblems, each of which is limited to optimization of a substructure. The algorithm applies also to those multidisciplinary problems whose subproblems form a hierarchy similar to that of substructures. In systems where the subproblems communicate with each other at the same level, the decomposition becomes non-hierarchic and the system may be optimized as a whole based on the derivatives of the system behavior with respect to the design variables computed by a method that bypasses finite differencing on the system analysis. When a multidisciplinary system includes a structure as its part, a hybrid, hierarchic/non-hierarchic decomposition applies. Numerical examples and references to computational experience accumulated to date illustrate the discussion.

Sobieszczanski-Sobieski, Jaroslaw↗
NASA NTRS1998

Simulation of Conformal Spiral Slot Antennas on Composite Platforms

During the course of the grant, we wrote and distributed about 12 reports and an equal number of journal papers supported fully or in part by this grant. The list of reports (title & abstract) and papers are given in Appendices A and B. This grant has indeed been instrumental in developing a robust hybrid finite element method for the analysis of complex broadband antennas on doubly curved platforms. Previous to the grant, our capability was limited to simple printed patch antennas on mostly planar platforms. More specifically: (1) mixed element formulations were developed and new edge-based prisms were introduced; (2) these elements were important in permitting flexibility in geometry gridding for most antennas of interest; (3) new perfectly matched absorbers were introduced for mesh truncations associated with highly curved surfaces; (4) fast integral algorithms were introduced for boundary integral truncations reducing CPU time from O(N-2) down to O(N-1.5) or less; (5) frequency extrapolation schemes were developed for efficient broadband performance evaluations. This activity has been successfully continued by NASA researchers; (6) computer codes were developed and extensively tested for several broadband configurations. These include FEMA-CYL, FEMA-PRISM and FEMA-TETRA written by L. Kempel, T. Ozdemir and J. Gong, respectively; (7) a new infinite balun feed was designed nearly constant impedance over the 800-3000 MHz operational band; (8) a complete slot spiral antenna was developed, fabricated and tested at NASA Langley. This new design is a culmination of the projects goals and integrates the computational and experimental efforts. this antenna design resulted in a U.S. patent and was revised three times to achieve the desired bandwidth and gain requirements from 800-3000 MHz.

Volakis, J. L.↗
DOE OSTI2025

Generative AI models for learning flow maps of stochastic dynamical systems in bounded domains

Simulating stochastic differential equations (SDEs) in bounded domains, presents significant computational challenges due to particle exit phenomena, which requires accurate modeling of interior stochastic dynamics and boundary interactions. Despite the success of machine learning-based methods in learning SDEs, existing learning methods are not applicable to SDEs in bounded domains because they cannot accurately capture the particle exit dynamics. We present a unified hybrid data-driven approach that combines a conditional diffusion model with an exit prediction neural network to capture both interior stochastic dynamics and boundary exit phenomena. Our ML model consists of two major components: a neural network that learns exit probabilities using binary cross-entropy loss with rigorous convergence guarantees, and a training-free diffusion model that generates state transitions for non-exiting particles using closed-form score functions. The two components are integrated through a probabilistic sampling algorithm that determines particle exit at each time step and generates appropriate state transitions. Here, the performance of the proposed approach is demonstrated via three test cases: a one-dimensional simplified problem for theoretical verification, a two-dimensional advection-diffusion problem in a bounded domain, and a three-dimensional problem of interest to magnetically confined fusion plasmas.

Bounded domains↗
DOE OSTI2024

Force-Free Identification of Minimum-Energy Pathways and Transition States for Stochastic Electronic Structure Theories

Here, the accurate mapping of potential energy surfaces (PESs) is crucial to our understanding of the numerous physical and chemical processes mediated by atomic rearrangements, such as conformational changes and chemical reactions, and the thermodynamic and kinetic feasibility of these processes. Stochastic electronic structure theories, e.g., Quantum Monte Carlo (QMC) methods, enable highly accurate total energy calculations that in principle can be used to construct the PES. However, their stochastic nature poses a challenge to the computation and use of forces and Hessians, which are typically required in algorithms for minimum-energy pathway (MEP) and transition state (TS) identification, such as the nudged elastic band (NEB) algorithm and its climbing image formulation. Here, we present strategies that utilize the surrogate Hessian line-search method, previously developed for QMC structural optimization, to efficiently identify MEP and TS structures without requiring force calculations at the level of the stochastic electronic structure theory. By modifying the surrogate Hessian algorithm to operate in path-orthogonal subspaces and at saddle points, we show that it is possible to identify MEPs and TSs by using a force-free QMC approach. We demonstrate these strategies via two examples, the inversion of the ammonia (NH 3 ) molecule and the nucleophilic substitution (S N 2) reaction F – + CH 3 F → FCH 3 + F – . We validate our results using Density Functional Theory (DFT)- and Coupled Cluster (CCSD, CCSD(T))-based NEB calculations. We then introduce a hybrid DFT-QMC approach to compute thermodynamic and kinetic quantities, free energy differences, rate constants, and equilibrium constants that incorporates stochastically optimized structures and their energies, and show that this scheme improves upon DFT accuracy. Our methods generalize straightforwardly to other systems and other high-accuracy theories that similarly face challenges computing energy gradients, paving the way for highly accurate PES mapping, transition state determination, and thermodynamic and kinetic calculations at significantly reduced computational expense.

Iyer, Gopal R.↗
NASA NTRS2015

Scalable Lunar Surface Networks and Adaptive Orbit Access

Teranovi Technologies, Inc., has developed innovative network architecture, protocols, and algorithms for both lunar surface and orbit access networks. A key component of the overall architecture is a medium access control (MAC) protocol that includes a novel mechanism of overlaying time division multiple access (TDMA) and carrier sense multiple access with collision avoidance (CSMA/CA), ensuring scalable throughput and quality of service. The new MAC protocol is compatible with legacy Institute of Electrical and Electronics Engineers (IEEE) 802.11 networks. Advanced features include efficiency power management, adaptive channel width adjustment, and error control capability. A hybrid routing protocol combines the advantages of ad hoc on-demand distance vector (AODV) routing and disruption/delay-tolerant network (DTN) routing. Performance is significantly better than AODV or DTN and will be particularly effective for wireless networks with intermittent links, such as lunar and planetary surface networks and orbit access networks.

Wang, Xudong↗
DOE OSTI2021

ExaWind: Predictive Wind Energy Simulations

This presentation describes the ExaWind project and the team's progress in creating a suite of performance-portable codes designed for predictive simulations of wind farms on next-generation exascale-class supercomputers. Such simulations will require the resolution of scales spanning many orders of magnitude, from blade boundary layers to wind farm flow structures. In the U.S., the first exascale systems will be GPU accelerated, and different GPU manufacturers have been chosen for the different systems. At the heart of the ExaWind software is a hybrid-solver approach based on the codes Nalu-Wind and AMR-Wind, which are computational fluid dynamics solvers for the incompressible Navier-Stokes equations. Nalu-Wind is an unstructured-grid code used to resolve wind turbine geometry and blade boundary layers, whereas AMR-Wind is a structured-grid background solver for atmospheric turbulent flow and turbine wake propagation. The models are coupled with overset meshes and global linear systems are approximated through a loose-coupling algorithm. Results will include validation-quality high-fidelity simulations and strong/weak scaling results from the Summit supercomputer.

49 EE - Wind and Water Power Program - Wind (EE-4W↗
NASA NTRS1998

A Hybrid Procedural/Deductive Executive for Autonomous Spacecraft

The New Millennium Remote Agent (NMRA) will be the first AI system to control an actual spacecraft. The spacecraft domain places a strong premium on autonomy and requires dynamic recoveries and robust concurrent execution, all in the presence of tight real-time deadlines, changing goals, scarce resource constraints, and a wide variety of possible failures. To achieve this level of execution robustness, we have integrated a procedural executive based on generic procedures with a deductive model-based executive. A procedural executive provides sophisticated control constructs such as loops, parallel activity, locks, and synchronization which are used for robust schedule execution, hierarchical task decomposition, and routine configuration management. A deductive executive provides algorithms for sophisticated state inference and optimal failure recover), planning. The integrated executive enables designers to code knowledge via a combination of procedures and declarative models, yielding a rich modeling capability suitable to the challenges of real spacecraft control. The interface between the two executives ensures both that recovery sequences are smoothly merged into high-level schedule execution and that a high degree of reactivity is retained to effectively handle additional failures during recovery.

Pell, Barney↗
DOE OSTI2020

Scalable molecular dynamics on CPU and GPU architectures with NAMD

NAMD is a molecular dynamics program designed for high-performance simulations of very large biological objects on CPU- and GPU-based architectures. NAMD offers scalable performance on petascale parallel supercomputers consisting of hundreds of thousands of cores, as well as on inexpensive commodity clusters commonly found in academic environments. It is written in C++ and leans on Charm++ parallel objects for optimal performance on low-latency architectures. NAMD is a versatile, multipurpose code that gathers state-of-the-art algorithms to carry out simulations in apt thermodynamic ensembles, using the widely popular CHARMM, AMBER, OPLS, and GROMOS biomolecular force fields. Here, we review the main features of NAMD that allow both equilibrium and enhanced-sampling molecular dynamics simulations with numerical efficiency. We describe the underlying concepts utilized by NAMD and their implementation, most notably for handling long-range electrostatics; controlling the temperature, pressure, and pH; applying external potentials on tailored grids; leveraging massively parallel resources in multiple-copy simulations; and hybrid quantum-mechanical/molecular-mechanical descriptions. We detail the variety of options offered by NAMD for enhanced-sampling simulations aimed at determining free-energy differences of either alchemical or geometrical transformations and outline their applicability to specific problems. Last, we discuss the roadmap for the development of NAMD and our current efforts toward achieving optimal performance on GPU-based architectures, for pushing back the limitations that have prevented biologically realistic billion-atom objects to be fruitfully simulated, and for making large-scale simulations less expensive and easier to set up, run, and analyze. NAMD is distributed free of charge with its source code at www.ks.uiuc.edu.

high-performance computing↗
DOE OSTI2022

AutoEnRichness: A hybrid empirical and analytical approach for estimating the richness of galaxy clusters

ABSTRACT We introduce AutoEnRichness, a hybrid approach that combines empirical and analytical strategies to determine the richness of galaxy clusters (in the redshift range of 0.1 ≤ z ≤ 0.35) using photometry data from the Sloan Digital Sky Survey Data Release 16, where cluster richness can be used as a proxy for cluster mass. In order to reliably estimate cluster richness, it is vital that the background subtraction is as accurate as possible when distinguishing cluster and field galaxies to mitigate severe contamination. AutoEnRichness is comprised of a multistage machine learning algorithm that performs background subtraction of interloping field galaxies along the cluster line of sight and a conventional luminosity distribution fitting approach that estimates cluster richness based only on the number of galaxies within a magnitude range and search area. In this proof-of-concept study, we obtain a balanced accuracy of 83.20 per cent when distinguishing between cluster and field galaxies as well as a median absolute percentage error of 33.50 per cent between our estimated cluster richnesses and known cluster richnesses within r200. In the future, we aim for AutoEnRichness to be applied on upcoming large-scale optical surveys, such as the Legacy Survey of Space and Time and Euclid, to estimate the richness of a large sample of galaxy groups and clusters from across the halo mass function. This would advance our overall understanding of galaxy evolution within overdense environments as well as enable cosmological parameters to be further constrained.

79 ASTRONOMY AND ASTROPHYSICS↗
NASA NTRS1993

An overview on development of neural network technology

The study has been to obtain a bird's-eye view of the current neural network technology and the neural network research activities in NASA. The purpose was two fold. One was to provide a reference document for NASA researchers who want to apply neural network techniques to solve their problems. Another one was to report out survey results regarding NASA research activities and provide a view on what NASA is doing, what potential difficulty exists and what NASA can/should do. In a ten week study period, we interviewed ten neural network researchers in the Langley Research Center and sent out 36 survey forms to researchers at the Johnson Space Center, Lewis Research Center, Ames Research Center and Jet Propulsion Laboratory. We also sent out 60 similar forms to educators and corporation researchers to collect general opinions regarding this field. Twenty-eight survey forms, 11 from NASA researchers and 17 from outside, were returned. Survey results were reported in our final report. In the final report, we first provided an overview on the neural network technology. We reviewed ten neural network structures, discussed the applications in five major areas, and compared the analog, digital and hybrid electronic implementation of neural networks. In the second part, we summarized known NASA neural network research studies and reported the results of the questionnaire survey. Survey results show that most studies are still in the development and feasibility study stage. We compared the techniques, application areas, researchers' opinions on this technology, and many aspects between NASA and non-NASA groups. We also summarized their opinions on difficulties encountered. Applications are considered the top research priority by most researchers. Hardware development and learning algorithm improvement are the next. The lack of financial and management support is among the difficulties in research study. All researchers agree that the use of neural networks could result in cost saving. Fault tolerance has been claimed as one important feature of neural computing. However, the survey indicates that very few studies address this issue. Fault tolerance is important in space mission and aircraft control. We believe that it is worthy for NASA to devote more efforts into the utilization of this feature.

Lin, Chun-Shin↗