Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “successive training”

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 19 records
All sourcesNASA NTRSDOE OSTI
NASA NTRS2018

Participant Training for a Flight Test Evaluation of Interval Management

Interval Management is a concept designed to be used by air traffic controllers and flight crews to more efficiently and precisely manage inter-aircraft spacing. NASA, in cooperation with Boeing, Honeywell, and United Airlines, tested an avionics prototype onboard flight test aircraft. A critical need was identified to train the pilots participating in the flight test prior to the first flight. This paper documents the flight training regimen that successfully trained the pilots on the Interval Management concepts and flight crew procedures and suggests potential improvements to future training regimens for industry use.

Roper, Roy↗
NASA NTRS1993

Perceived self-orientation and self-motion in microgravity, after landing and during preflight adaptation training

The research described in this paper is intended to support development and evaluation of preflight adaptation training (PAT) apparatus and procedures. Successful training depends on appropriate manipulation of visual and inertial stimuli that control perception of self-motion and self-orientation. For one part of this process, astronauts are trained to report their self-motion and self-orientation experiences. Before their space mission, they are exposed to the altered sensory environments produced by the PAT trainers. During and after the mission, they report their motion and orientation experiences. Subsequently, they are again exposed to the PAT trainers and are asked to describe relationships between their experiences in microgravity and following entry and their experiences in the trainers.

Space Flight↗
NASA NTRS2017

Updates and Overview of Spaceflight Medical Support in Russia and Kazakhstan

This panel presents recent updates to and a comprehensive overview of the operational medical support provided to ISS crewmembers in Star City, Russia and Kazakhstan as part of UTMB/KBRwyle's Human Health & Performance contract. With the current Soyuz training flow, physician support is required for nominal training evolutions involving pressure changes or other potential physical risks detailed in this presentation. In addition, full-time physician presence in Star City helps to address the disparity in access to health care in these relatively remote practice areas, while also developing and maintaining relationships with host nation resources. A unique part of standard training in Russia also involves survival training in both winter and water environments; logistic details and medical impacts of each of these training scenarios will be discussed. Following support of a successful training flow, UTMB/KBRwyle's Star City Medical Support Group (SCMSG) is also responsible for configuring medical packs in support of Soyuz launches and landings; we will present the rationale for current pack contents within the context of specific operational needs. With respect to contingency events, the group will describe their preparedness to respond appropriately by activating both local and global resources as necessary, detailing a specialized subset of the group who continually work and update these assets, given changes in international infrastructure and other impacts.

Chough, Natacha↗
NASA NTRSDate not supplied

Quantum Hardware-Enabled Molecular Dynamics via Transfer Learning

The ability to perform ab initio molecular dynamics simulations using potential energy surfaces provided by quantum computers would open the door to virtually exact dynamics for a variety of chemical and biochemical systems, with impacts on catalysis and biophysics. Nonetheless, performing molecular dynamics on surfaces produced by quantum hardware has been hampered by the noisy energies typically produced by quantum computers and challenges associated with computing gradients and scaling to large systems interest. A recent set of advances in machine learning, known as transfer learning, provides a new path forward for molecular dynamics simulations on quantum hardware. Transfer learning offers a workaround, where one first trains models on larger, less accurate classical datasets and then refines them on smaller, more accurate quantum datasets. We explore this approach by training machine learning models to predict a molecule's potential energy based on its geometric structure using Behler-Parrinello neural networks. When successfully trained, the model enables energy gradient predictions necessary for dynamic simulations. To reduce the quantum resources needed, the model is initially trained with data derived from classical density functional theory and subsequently refined with a smaller dataset obtained from a variational quantum eigensolver optimization of the unitary coupled cluster ansatz. We show that this approach significantly reduces the size of the needed quantum training dataset while capturing the high accuracies needed within quantum chemistry simulations. The success of this two-step training method opens more opportunities to apply machine learning models on quantum data, a significant stride towards efficient quantum-classical hybrid computational models.

quantum computing↗
NASA NTRS1987

Vestibular-visual conflict training

Squirrel monkeys susceptible to vestibular-visual conflict (similar to that produced in space motion) were successfully trained to lesser sensitivity by repeated exposures to various randomly mixed patterns of vestibular-visual conflict in sagittal plane. As long as the training effect was retained, trained animals exhibited low sickness scores and no emesis. It is suggested that the adaptation maneuvers described in this study can be effective as a training routine for crew members immediately prior to their space flight missions.

Igarashi, Makoto↗
NASA NTRSDate not supplied

Predicting the Seawater Chemistry of an Ocean World Using Machine Learning on Isotopic Measurements of Volatile CO2

Introduction: Given the long time intervals required for data transmission to and from ocean worlds targets, low bandwidth for data transmission, time required for data processing and analysis, and potentially extreme radiation environments (e.g., Europa), it is clear that ocean worlds missions will need more autonomous flight instruments and software in order to achieve established science goals. Protracted time intervals for data analysis (e.g., Europa Lander) strongly motivates the development of rapid, consistent and streamlined methods for interpreting data from flight mass spectrometers to e.g., determine how mass spectra from a plume or surface liquid/ice relates to the surface/subsurface. Since mass spectrometry also has the potential to correctly identify biosignatures[1], it is imperative that such methods for interpreting data are consistent and accurate. We used 848 isotope ratio mass spectra from laboratory analyses of CO2 that interacted with ocean worlds-relevant seawaters as a ‘training’ dataset for ‘unsupervised’ machine learning. In unsupervised learning, characteristics of the data are not labeled or linked, and any similarities found only result from the neural network. CO2 isotopologues analyzed for this dataset mimic the remote measurements of CO2 by a flight mass spectrometer, and are detailed in Theiling [2]. From this dataset, we used measured features of the spectra, such as retention time, intensity, and (isotopologue) mass ratios as inputs for our autoencoder neural network. Our neural network was trained to find similarities in these and other spectral features for seawaters of a particular composition and amount of initial CO2. Successful training then created an output of these similarities for various seawaters, which included MgSO4, Na2SO4, NaCl, MgCl2, KCl, and NaHCO3, and combinations of these salts. We then applied dimensionality reduction techniques such as Principal Component Analysis (PCA), T-Distributed Stochastic Neighbor Embedding (TSNE), and Uniform Manifold Approximation and Projection (UMAP) to demonstrate latent data features as a two-dimensional projection in a unitless, high-dimensional space. In this projection, a data point represents the combined effect of spectral features such as intensity, retention time, and isotope ratio. Our initial UMAP demonstrates data clustering (organization of the data by the neural network) based on the amount of CO2 that had initially interacted with each seawater. Further training using more ‘supervised’ learning techniques demonstrate strong clustering of preliminary data based on initial CO2 concentration, seawater chemical composition, and ionic strength (salinity). Our preliminary work therefore suggests that machine learning has the potential to identify compositional variants of an ocean world seawater based on mass spectra from volatile CO2 measurements. Acknowledgments: This work was funded through a Strategic Task Group at NASA Goddard Space Flight Center. The training dataset was collected through funding from the Oklahoma Space Grant Consortium. References: [1] Pappalardo, R. et al. (2013) Astrobiology, 13, 740–773. [2] Theiling (2020) Icarus, 114216.

Europa↗
NASA NTRS1999

Neural Networks Used to Compare Designed and Measured Time-Average Patterns

Electronic time-average holograms are convenient for comparing the measured vibration modes of fan blades with those calculated by finite-element models. At the NASA Lewis Research Center, neural networks recently were trained to perform what had been a simple visual comparison of the predictions of the design models with the measurements. Finite-element models were used to train neural networks to recognize damage and strain information encoded in subtle changes in the time-average patterns of cantilevers. But the design-grade finite element models were unable to train the neural networks to detect damage in complex blade shapes. The design-model-generated patterns simply did not agree well enough with the measured patterns. Instead, hybrid-training records, with measured time-average patterns as the input and model-generated strain information as the output, were used to effect successful training.

Decker, Arthur J.↗
DOE OSTI2024

ReLIC: Full-Scale Realization of Reinforcement Learning for Infrastructure Control

Prior efforts have shown that deep reinforcement learning (DRL) may provide a new method for controlling networked power systems. Though successful, prior approaches have not yet demonstrated their behavior on systems of realistic scale. This effort examined multiple theoretical and technical approaches to allow a DRL model to operate over a system of 2,000 buses or more. We find that allowing the DRL models to run training episodes in parallel provides near limitless efficiency gains, allowing us to train successful agents to behave on our Kuramoto transmission model of up to 4,000 buses. We further show that we can expand our PowerWorld DRL implementation to systems of up to 25 buses but struggle to go beyond this limit due to PowerWorld’s inability to run multiple instances at once. Finally, we examine a multi-agent approach and find that it performs as well if not better than our existing centralized approach.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2009

A Methodology for Training International Space Station Crews to Respond to On-Orbit Emergencies

Most spaceflight crewmembers agree that emergency training is among the most important training they receive. If an emergency event occurs on-orbit crewmembers want to be able to rely on a thorough and proficient knowledge of emergency operations and procedures. The inherent complexity of ISS and the international nature of the onboard operations have resulted in emergency procedures that are complex by any measure; as a result, a very robust apparatus has been developed to give crewmembers initial training on emergency procedures and ensure proficiency up to (and even after) launch. One of the most important aspects of complex onboard operations in general, and emergency operations specifically, is learning how to coordinate roles and responsibilities with fellow crewmembers. A primary goal of NASA s emergency training program is to allow the crewmembers who will actually be together on-orbit to practice executing the emergency responses together before they fly. As with any operation that includes the use of software and hardware, the fidelity of the simulation environment is a critical element to successful training. The NASA training division has spent considerable time and effort to develop a simulator that addresses the most important aspects of emergency response, working within very difficult space and budgetary constraints.

Balmain, Clinton↗
NASA NTRS1992

Fast temporal neural learning using teacher forcing

A neural network is trained to output a time dependent target vector defined over a predetermined time interval in response to a time dependent input vector defined over the same time interval by applying corresponding elements of the error vector, or difference between the target vector and the actual neuron output vector, to the inputs of corresponding output neurons of the network as corrective feedback. This feedback decreases the error and quickens the learning process, so that a much smaller number of training cycles are required to complete the learning process. A conventional gradient descent algorithm is employed to update the neural network parameters at the end of the predetermined time interval. The foregoing process is repeated in repetitive cycles until the actual output vector corresponds to the target vector. In the preferred embodiment, as the overall error of the neural network output decreasing during successive training cycles, the portion of the error fed back to the output neurons is decreased accordingly, allowing the network to learn with greater freedom from teacher forcing as the network parameters converge to their optimum values. The invention may also be used to train a neural network with stationary training and target vectors.

Toomarian, Nikzad↗
NASA NTRS1995

Fast temporal neural learning using teacher forcing

A neural network is trained to output a time dependent target vector defined over a predetermined time interval in response to a time dependent input vector defined over the same time interval by applying corresponding elements of the error vector, or difference between the target vector and the actual neuron output vector, to the inputs of corresponding output neurons of the network as corrective feedback. This feedback decreases the error and quickens the learning process, so that a much smaller number of training cycles are required to complete the learning process. A conventional gradient descent algorithm is employed to update the neural network parameters at the end of the predetermined time interval. The foregoing process is repeated in repetitive cycles until the actual output vector corresponds to the target vector. In the preferred embodiment, as the overall error of the neural network output decreasing during successive training cycles, the portion of the error fed back to the output neurons is decreased accordingly, allowing the network to learn with greater freedom from teacher forcing as the network parameters converge to their optimum values. The invention may also be used to train a neural network with stationary training and target vectors.

Toomarian, Nikzad↗
NASA NTRS1997

The Design of Model-Based Training Programs

This paper proposes a model-based training program for the skills necessary to operate advance avionics systems that incorporate advanced autopilots and fight management systems. The training model is based on a formalism, the operational procedure model, that represents the mission model, the rules, and the functions of a modem avionics system. This formalism has been defined such that it can be understood and shared by pilots, the avionics software, and design engineers. Each element of the software is defined in terms of its intent (What?), the rationale (Why?), and the resulting behavior (How?). The Advanced Computer Tutoring project at Carnegie Mellon University has developed a type of model-based, computer aided instructional technology called cognitive tutors. They summarize numerous studies showing that training times to a specified level of competence can be achieved in one third the time of conventional class room instruction. We are developing a similar model-based training program for the skills necessary to operation the avionics. The model underlying the instructional program and that simulates the effects of pilots entries and the behavior of the avionics is based on the operational procedure model. Pilots are given a series of vertical flightpath management problems. Entries that result in violations, such as failure to make a crossing restriction or violating the speed limits, result in error messages with instruction. At any time, the flightcrew can request suggestions on the appropriate set of actions. A similar and successful training program for basic skills for the FMS on the Boeing 737-300 was developed and evaluated. The results strongly support the claim that the training methodology can be adapted to the cockpit.

Polson, Peter↗
DOE OSTI2024

Combining High-Throughput Experiments and Active Learning to Characterize Deep Eutectic Solvents

The high tunability of deep eutectic solvents (DESs) stems from the ease of changing their precursors and relative compositions. However, measuring the physicochemical properties across large composition and temperature ranges, necessary to properly design target-specific DESs, is tedious and error-prone and represents a bottleneck in the advancement and scalability of DES-based applications. As such, active learning (AL) methodologies based on Gaussian processes (GPs) were developed in this work to minimize the experimental effort necessary to characterize DESs. Owing to its importance for large-scale applications, the reduction of DES viscosity through the addition of a low-molecular-weight solvent was explored as a case study. A high-throughput experimental screening was initially performed on nine different ternary DESs. Then, GPs were successfully trained to predict DES viscosity from its composition and temperature, showcasing the ability of these stochastic, nonparametric models to accurately describe the physicochemical properties of complex mixtures. Finally, the ability of GPs to provide estimates of their own uncertainty was leveraged through an AL framework to minimize the number of data points necessary to obtain accurate viscosity modes. This led to a significant reduction in data requirements, with many systems requiring only five independent viscosity data points to be properly described.

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

Identification of carbohydrate gene clusters obtained from in vitro fermentations as predictive biomarkers of prebiotic responses

Prebiotic fibers are non-digestible substrates that modulate the gut microbiome by promoting expansion of microbes having the genetic and physiological potential to utilize those molecules. Although several prebiotic substrates have been consistently shown to provide health benefits in human clinical trials, responder and non-responder phenotypes are often reported. These observations had led to interest in identifying, a priori, prebiotic responders and non-responders as a basis for personalized nutrition. In this study, we conducted in vitro fecal enrichments and applied shotgun metagenomics and machine learning tools to identify microbial gene signatures from adult subjects that could be used to predict prebiotic responders and non-responders. Using short chain fatty acids as a targeted response, we identified genetic features, consisting of carbohydrate active enzymes, transcription factors and sugar transporters, from metagenomic sequencing of in vitro fermentations for three prebiotic substrates: xylooligosacharides, fructooligosacharides, and inulin. A machine learning approach was then used to select substrate-specific gene signatures as predictive features. These features were found to be predictive for XOS responders with respect to SCFA production in an in vivo trial. Our results confirm the bifidogenic effect of commonly used prebiotic substrates along with inter-individual microbial responses towards these substrates. We successfully trained classifiers for the prediction of prebiotic responders towards XOS and inulin with robust accuracy (≥ AUC 0.9) and demonstrated its utility in a human feeding trial. Overall, the findings from this study highlight the practical implementation of pre-intervention targeted profiling of individual microbiomes to stratify responders and non-responders.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Building collaboration to advance our understanding of regional climate impacts of dust in California's San Joaquin Valley

This project successfully achieved its central objective of building collaborative research capabilities at UC Merced, a Hispanic-Serving Institution, to advance understanding of the regional climate impacts of dust in California's San Joaquin Valley. Through strategic partnerships with three DOE national laboratories (PNNL, LLNL, and LBNL), we developed critical expertise in the Energy Exascale Earth System Model (E3SM) and Atmospheric Radiation Measurement (ARM) facilities. Among the project's scientific contributions, one key publication includes demonstrating that fallowed agricultural lands are the primary source of anthropogenic dust in California's Central Valley, with dust activities increasing substantially between 2008 and 2022, in correlation with drought severity and expanded fallowed land coverage. This finding suggests that current climate models, including E3SM, likely underestimate the dust burden due to inadequate representation of agricultural land-use changes. Beyond the scientific contributions, the project successfully trained a PhD student, established ongoing collaborations resulting in multiple manuscripts in preparation, and positioned UC Merced to participate in the DUSTIEAIM campaign for 2026-2027, thereby building sustainable research capacity while addressing climate science questions directly relevant to the California Central Valley.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

Gravitational Lenses in UNIONS and Euclid (GLUE). I. A Search for Strong Gravitational Lenses in UNIONS with Subaru, CFHT, and Pan-STARRS Data

We present the results of our pipeline for discovering strong gravitational lenses in the ongoing Ultraviolet Near-Infrared Optical Northern Survey (UNIONS). We successfully train a deep residual convolutional neural network based on CMU-Deeplens architecture, which is designed to detect strong lenses in ground-based imaging surveys. We train on images of real strong lenses and deploy on a sample of 8 million galaxies in areas with full coverage in the g, r, and i filters—the first multiband search for strong gravitational lenses in UNIONS. Following human inspection and grading, we report the discovery of a total of 1346 new strong-lens candidates, of which 146 are Grade A, 199 are Grade B, and 1001 are Grade C. Of these candidates, 283 have lens galaxy spectroscopic redshifts from the Sloan Digital Sky Survey, and an additional 297 have them from the Dark Energy Spectroscopic Instrument Data Release 1. We find 15 of these systems display evidence of both lens and source galaxy redshifts in spectral superposition. We also report the spectroscopic confirmation of seven lensed sources in high-quality systems, all with z > 2.1, using the Keck Near Infrared Echellette Spectrograph and the Gemini Near-Infrared Spectrograph.

Storfer, Christopher J. [University of Hawaii, Hon↗
NASA NTRS2003

Enabling computer decisions based on EEG input

Multilayer neural networks were successfully trained to classify segments of 12-channel electroencephalogram (EEG) data into one of five classes corresponding to five cognitive tasks performed by a subject. Independent component analysis (ICA) was used to segregate obvious artifact EEG components from other sources, and a frequency-band representation was used to represent the sources computed by ICA. Examples of results include an 85% accuracy rate on differentiation between two tasks, using a segment of EEG only 0.05 s long and a 95% accuracy rate using a 0.5-s-long segment.

Validation Studies↗
NASA NTRS2008

Vertical Jump Height is more Strongly Associated with Velocity and Work Performed Prior to Take-off

Vertical jump (VJ) height is commonly used as a measure of athletic capability in strength and power sports. Although VJ has been shown to be a predictor of athletic performance, it is not clear which kinetic ground reaction force (GRF) variables, such as peak force (PF), peak power (PP), peak velocity (PV), total work (TW) or impulse (Imp) are the best correlates. To determine which kinetic variables (PF, PP, PV, TW, and Imp) best correlate with VJ height. Twenty subjects (14 males, 6 females) performed three maximal countermovement VJs on a force platform (Advanced Mechanical Technology, Inc., Watertown, MA, USA). VJ jump height was calculated as the difference between standing reach and the highest reach point measured using a Vertec. PF, PP, PV, TW, and Imp were calculated using the vertical GRF data sampled at 1000 Hz from the lowest point in the countermovement through the concentric portion until take-off. GRF data were normalized to body mass measured using a standard scale (Detecto, Webb City, MO, USA). Correlation coefficients were computed between each GRF variable and VJ height using a Pearson correlation. VJ height (43.4 plus or minus 9.1 cm) was significantly correlated (p less than 0.001) with PF (998 plus or minus 321 N; r=0.51), PP (1997 plus or minus 772 W; r=0.69), PV (2.66 plus or minus 0.40 m (raised dot) s(sup -1); r=0.85), TW (259 plus or minus 93.0 kJ; r=0.82), and Imp (204 plus or minus 51.1 N(raised dot)s; r=0.67). Although all variables were correlated to VJ height, PV and TW were more strongly correlated to VJ height than PF, PP, and Imp. Therefore, since TW is equal to force times displacement, the relative displacement of the center of mass along with the forces applied during the upward movement of the jump are critical determinants of VJ height. PV and TW are key determinants of VJ height, and therefore successful training programs to increase VJ height should focus on rapid movement (PV) and TW by increasing power over time rather than focusing on PF alone.

Bentley, J. R.↗