SEARCH · Search NASA

Results for “Contrastive learning”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 649 records · Page 36

All sources NASA NTRS DOE OSTI

DOE OSTI2020

Efficient bifurcation and tabulation of multi-dimensional combustion manifolds using deep mixture of experts: An a priori study

This work describes and validates an approach for autonomously bifurcating turbulent combustion man-ifolds to divide regression tasks amongst specialized artificial neural networks (ANNs). This approach relies on the mixture of experts (MoE) framework, where each neural network is trained to be specialized in a given portion of the input space. The assignment of different input regions to the experts is determined by a gating network, which is a neural network classifier. In some previous studies, it has been demonstrated that bifurcation of a complex combustion manifold and fitting different ANNs for each part leads to better fits or faster inference speeds. However, the manner of bifurcation in these studies was based on heuristic approaches or clustering techniques. In contrast, the proposed technique enables automatic bifurcation using non-linear planes in high-dimensional turbulent combustion manifolds that are often associated with complex behavior due to different dominating physics in various zones. Furthermore, the proposed concept is validated using 4-dimensional (4D) and 5D flamelet tables, showing that the errors obtained with a given network size, or conversely the network size required to achieve a given accuracy, is considerably reduced. The effect of the number of experts on inference speed is also investigated, showing that by increasing the number of experts from 1 to 8, the inference time can be approximately reduced by a factor of two. Moreover, it is shown that the MoE approach divides the input manifold in a physically intuitive manner, suggesting that the MoE framework can elucidate high-dimensional datasets in a physically meaningful way.

42 ENGINEERING↗

DOE OSTI2024

Source shape estimation for neutron imaging systems using convolutional neural networks

Neutron imaging systems are important diagnostic tools for characterizing the physics of inertial confinement fusion reactions at the National Ignition Facility (NIF). In particular, neutron images give diagnostic information on the size, symmetry, and shape of the fusion hot spot and surrounding cold fuel. Images are formed via collection of neutron flux from the source using a system of aperture arrays and scintillator-based detectors. Currently, reconstruction of fusion source geometry from the collected neutron images is accomplished by solving a computationally intensive maximum likelihood estimation problem via expectation maximization. In contrast, it is often useful to have simple representations of the overall source geometry that can be computed quickly. In this work, we develop convolutional neural networks (CNNs) to reconstruct the outer contours of simple source geometries. We compare the performance of the CNN for penumbral and pinhole data and provide experimental demonstrations of our methods on both non-noisy and noisy data.

Machine learning, neutron imaging, source reconstr↗

DOE OSTI2024

Regional climate change: consensus, discrepancies, and ways forward

Climate change has emerged across many regions. Some observed regional climate changes, such as amplified Arctic warming and land-sea warming contrasts have been predicted by climate models. However, many other observed regional changes, such as changes in tropical sea surface temperature and monsoon rainfall are not well simulated by climate model ensembles even when taking into account natural internal variability and structural uncertainties in the response of models to anthropogenic radiative forcing. This suggests climate model predictions may not fully reflect what our future will look like. The discrepancies between models and observations are not well understood due to several real and apparent puzzles and limitations such as the “signal-to-noise paradox” and real-world record-shattering extremes falling outside of the possible range predicted by models. Addressing these discrepancies, puzzles and limitations is essential, because understanding and reliably predicting regional climate change is necessary in order to communicate effectively about the underlying drivers of change, provide reliable information to stakeholders, enable societies to adapt, and increase resilience and reduce vulnerability. The challenges of achieving this are greater in the Global South, especially because of the lack of observational data over long time periods and a lack of scientific focus on Global South climate change. To address discrepancies between observations and models, it is important to prioritize resources for understanding regional climate predictions and analyzing where and why models and observations disagree via testing hypotheses of drivers of biases using observations and models. Gaps in understanding can be discovered and filled by exploiting new tools, such as artificial intelligence/machine learning, high-resolution models, new modeling experiments in the model hierarchy, better quantification of forcing, and new observations. Conscious efforts are needed toward creating opportunities that allow regional experts, particularly those from the Global South, to take the lead in regional climate research. This includes co-learning in technical aspects of analyzing simulations and in the physics and dynamics of regional climate change. Finally, improved methods of regional climate communication are needed, which account for the underlying uncertainties, in order to provide reliable and actionable information to stakeholders and the media.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2021

Tracking the phase changes in micelle-based NiGa nanocatalysts for methanol synthesis under activation and working conditions

The hydrogenation of CO 2 into high energy density fuels such as methanol, where the required H 2 is obtained from renewable sources, is of utmost importance for a sustainable society. In recent years, NiGa alloys have attracted attention as promising catalyst material systems for the hydrogenation of CO 2 into methanol at ambient pressures. They thus represent an energy-saving alternative to the Cu-based catalysts employed in today's catalytic industry that require high pressures for the CO 2 hydrogenation. However, the underlying reaction mechanisms for the NiGa system are still under debate. One of the challenges here is to unravel the evolution and coexistence of the different species in the heterogeneous NiGa catalyst system under activation and reaction conditions. To shed light on their evolution under activation in H 2 and their catalytic roles under CO 2 hydrogenation working conditions on well-defined Ni 3 Ga 1 and Ni 5 Ga 3 nanoparticle (NP) catalysts, we employed a multi-probe approach in this study. It included advanced machine learning-based analysis of operando X-ray absorption spectroscopy data combined with operando powder X-ray diffraction and near ambient pressure X-ray photoelectron spectroscopy measurements, as well as reactivity studies using bed-packed mass flow reactors. In addition, we employed atomic force microscopy and scanning transmission electron microscopy for structural characterization. Under H 2 activation at 1 bar total pressure, we concluded the formation of metallic Ni, starting for Ni 3 Ga 1 at 300 °C, and for Ni 5 Ga 3 at 400 °C. At higher temperatures, the formation of NiGa alloys follows. The α'-Ni 3 Ga 1 alloy phase is predominantly formed for the Ni 3 Ga 1 NPs, while the coexistence of α'-Ni 3 Ga 1 , δ-Ni 5 Ga 3 and Ga 2 O 3 phases is observed for the Ni 5 Ga 3 NPs after the H 2 activation. The formation of the Ga 2 O 3 phase also results in the presence of excess metallic Ni. Under CO 2 hydrogenation reaction conditions, Ga partially oxidizes again to form a Ga 2 O 3 -rich particle shell for both NP compositions, yet, to a larger extent for the Ni 3 Ga 1 NPs, which, in turn, feature a higher amount of excess Ni. We reveal that metallic Ni is responsible for the high selectivity of the Ni 3 Ga 1 NPs towards the production of methane in our catalytic tests. In contrast, the Ni 5 Ga 3 NPs display a strong selectivity toward methanol production (>92%), more than one order of magnitude higher than that for the Ni 3 Ga 1 NPs, which we ascribe to the presence of the δ-Ni 5 Ga 3 phase.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

NASA NTRS2013

Sleuthing the MSL EDL Performance from an X Band Carrier Perspective

During the Entry, Descent, and Landing (EDL) of NASA's Mars Science Laboratory (MSL), or Curiosity, rover to Gale Crater on Mars on August 6, 2012 UTC, the rover transmitted an X-band signal composed of carrier and tone frequencies and a UHF signal modulated with an 8kbps data stream. During EDL, the spacecraft's orientation is determined by its guidance and mechanical subsystems to ensure that the vehicle land safely at its destination. Although orientation to maximize telecom performance is not possible, antennas are especially designed and mounted to provide the best possible line of sight to Earth and to the Mars orbiters supporting MSL's landing. The tones and data transmitted over these links are selected carefully to reflect the most essential parameters of the vehicle's state and the performance of the EDL subsystems for post-EDL reconstruction should no further data transmission from the vehicle be possible. This paper addresses the configuration of the X band receive system used at NASA / JPL's Deep Space Network (DSN) to capture the signal spectrum of MSL's X band carrier and tone signal, examines the MSL vehicle state information obtained from the X band carrier signal only and contrasts the Doppler-derived information against the post-EDL known vehicle state. The paper begins with a description of the MSL EDL sequence of events and discusses the impact of the EDL maneuvers such as guided entry, parachute deploy, and powered descent on the frequency observables expected at the DSN. The range of Doppler dynamics possible is derived from extensive 6 Degrees-Of-Freedom (6 DOF) vehicle state calculations performed by MSL's EDL simulation team. The configuration of the DSN's receive system, using the Radio Science Receivers (RSR) to perform open-loop recording for both for nominal and off-nominal EDL scenarios, is detailed. Expected signal carrier power-to-noise levels during EDL are shown and their impact on signal detection is considered. Particular attention is given to the selection of the appropriate RSR processing bandwidths and to its configuration for real-time signal detection. The X-band carrier frequency obtained through post-processing of the open-loop recorded spectrum is given. Detection of spacecraft status and completion of key vehicle events through their Doppler signature is discussed and illustrated. This Doppler-derived information is compared against the very accurate vehicle data obtained post-EDL via MSL's UHF radio subsystem. The paper concludes with a discussion on the advantages and disadvantages of transmitting the X-band carrier and tone signal in the general context of EDL communications and lessons learned for future missions with EDL sequences are given

Mars Science Laboratory (MSL)↗

DOE OSTI2022

Classification of Photovoltaic Failures with Hidden Markov Modeling, an Unsupervised Statistical Approach

Failure detection methods are of significant interest for photovoltaic (PV) site operators to help reduce gaps between expected and observed energy generation. Current approaches for field-based fault detection, however, rely on multiple data inputs and can suffer from interpretability issues. In contrast, this work offers an unsupervised statistical approach that leverages hidden Markov models (HMM) to identify failures occurring at PV sites. Using performance index data from 104 sites across the United States, individual PV-HMM models are trained and evaluated for failure detection and transition probabilities. This analysis indicates that the trained PV-HMM models have the highest probability of remaining in their current state (87.1% to 93.5%), whereas the transition probability from normal to failure (6.5%) is lower than the transition from failure to normal (12.9%) states. A comparison of these patterns using both threshold levels and operations and maintenance (O&M) tickets indicate high precision rates of PV-HMMs (median = 82.4%) across all of the sites. Although additional work is needed to assess sensitivities, the PV-HMM methodology demonstrates significant potential for real-time failure detection as well as extensions into predictive maintenance capabilities for PV.

classification↗

NASA NTRS2005

Physics-Based Spectra of Accretion Disks around Black Holes

The purpose of this grant was to begin the process of deriving the light output of accretion disks around black holes directly from the actual processes that inject heat into the accreting matter, rather than from guessed dependences of heating rate on physical parameters. At JHU, the effort has focussed so far on models of accretion onto "intermediate mass black holes", a possible class of black holes, examples of which may have recently been discovered in nearby galaxies. There, Krolik and his student (Yawei Hui) have computed stellar atmospheres for uniformly-heated disks around this class of black holes. Their models serve two purposes: they are the very first serious attempts to compute the spectrum from accreting black holes in this mass range; and a library of such models can be used later in this program as contrasts for those computed on the basis of real disk dynamics. The output from these local disk calculations has also been successfully coupled to a program that applies the appropriate relativistic transformations and computes photon trajectories in order to predict the spectrum received by observers located at different polar angles. The principal new result of these calculations is the discovery of potentially observable ionization edges of H-like C and O at frequencies near the peak in flux from these objects. Most of the grant money at UCSB was spent on supporting graduate student Shane Davis. In addition. some money was spent on supporting two other students: Ari Socrates (now a Hubble Fellow at Princeton), and Laura Melling. Davis spent the year constructing stellar atmosphere models of accretion disks appropriate for the high/soft (thermal) state of black hole X-ray binaries. As with AGN models published previously by our collaboration with NASA support. our models include a complete general relativistic treatment of both the disk structure and the propagation of photons from the disk to a distant observer. They also include all important continuum opacity sources, including Compton scattering and bound-free opacity from abundant metal species. The principal new result is that bound-free opacity is very significant in altering the continuum spectral shape, resulting for example in quite different "color correction factors" compared to those predicted previously. In addition, the models predict a relationship between luminosity and inner disk temperature that is, for the first time, in accord with that observed. The primary purpose of the grant was to incorporate more realistic accretion disk physics, learned largely from simulations, into such spectral models. The Davis et al. paper includes consideration of a vertical dissipation profile computed from radiation magneto-hydrodynamic simulations of MRI turbulence by N. J. Turner (2004). So long as the disk is effectively thick, such dissipation profiles do not affect the predicted spectrum significantly. (More work needs to be done on these simulations, however.) A potentially more serious issue is that MRI turbulence produces substantial inhomogeneities, as do photon bubble instabilities. These inhomogeneities can affect the spectra by enhancing the effects of absorption opacity over scattering opacity. We have done some preliminary Monte Carlo calculations to explore these effects.

Krolik, Julian H.↗

DOE OSTI2023

First Ever Field Pilot on Alaska's North Slope to Validate the Use of Polymer Floods for Heavy Oil EOR a.k.a Alaska North Slope Field Laboratory (ANSFL)

Alaska’s high viscosity oil resources that range between 20–30+ billion barrels represent about a third of known North Slope original oil in place (OOIP). These resources are primarily concentrated in the Schrader Bluff formation (also called West Sak on the Western North Slope) and Ugnu reservoirs and are categorized as “viscous oils” and “heavy oils” owing to their in-situ viscosities between 5–10,000 cP and up to a million+ cP respectively. The viscous oil deposits are relatively deeper (2,000 – 5,000 ft), whereas the heavy oils are somewhat shallower (2,000 – 4,000 ft). The typically shallow depths and the proximity to the continuous permafrost results in relatively lower formation temperatures and pressures, and consequently higher viscosities. The vertical depth vs. viscosity delineated in Paskvan et al. (2016) differentiates the viscous and heavy oils. As depicted in Paskvan et al. (2016), currently the main focus (referred to as “developing”) is on the viscous oils in the Schrader Bluff formation in the Milne Point Unit (MPU). Notwithstanding this Alaska North Slope (ANS) specific categorization, we use the industry adopted, all-inclusive term “heavy oil” for all high viscosity oils. Resource characterization and additional details can be found in topical publications of Paskvan et al. (2016) and Targac et al. (2005). Despite the vast resource base, the development pace, vis-à-vis the production of heavy oils has been very slow and limited due to multiple factors such as cost, logistics, challenging arctic environment, poor waterflood sweep efficiency due to mobility contrasts, and significantly high minimum miscibility pressures (MMP). Most importantly, typical or standard thermal methods that are commonplace elsewhere (Canada, California) are inapplicable due to the continuous permafrost. As a consequence, cumulative production of heavy and viscous oils is a little over 1% of OOIP slope wide and currently, there is hardly any production from Ugnu. However, on a broader level, these unfavorable factors are outweighed by the fact that (1) these resources, within the established infrastructure, are too large to ignore because of their strategic importance to the Nation and the State of Alaska and (2) Prudhoe Bay type diluent crude oil is still available for heavy oil transport through the Trans Alaska Pipeline System (TAPS). Similarly, from a reservoir standpoint, the following factors also are important offsets: (1) favorable rock characteristics of Schrader Bluff; (2) the promise demonstrated by the initial scoping studies (Seright 2010, 2011) suggesting significant increase of heavy oil recovery using polymer flooding; (3) successful field implementation in Canada, China and elsewhere in the world, and (4) availability of the existing pairs of horizontal injector-producer in Schrader Bluff The foregoing was recognized as the best readily available opportunity for significant investment by the US Department of Energy and the field operator Hilcorp Alaska LLC to conduct the first ever field scale experiment to test the polymer flooding technology to unlock the vast heavy oil resources on ANS. With this primary goal in mind, the research team embarked on a ~4.5 years long project that focused on the field polymer pilot complemented by supporting laboratory and simulation studies. As documented in this final report, over the course of the project, many lessons have been learned and valuable field and supporting laboratory data has been collected, which also is complemented by numerical reservoir simulations. We have been able to establish the injectivity of polymer solution, evidence of significant reduction in the water cut of previously waterflooded pattern, effective propagation of a hydrolyzed polyacrylamide (HPAM), benefits of low salinity water, provide practical guidance on handling of produced fluids containing breakthrough polymer, fit-for-purpose forecast-worthy history matched simulation model, polymer EOR benefit of 700-1000 bopd over waterflood, and most importantly a low polymer utilization factor of ~1.7 lb/stb. In summary this project is deemed as a scientific, technical and economic success, having met all objectives, fulfilled deliverables and within budget, providing impetus to apply polymer EOR throughout the Milne Point Field paving the way for even heavier viscosity oils in the Ugnu area, eventually extending the economic life of TAPS.

02 PETROLEUM↗

NASA NTRS2014

High Speed, Low Cost Telemetry Access from Space Development Update on Programmable Ultra Lightweight System Adaptable Radio (PULSAR)

Software Defined Radio (SDR) technology has been proven in the commercial sector since the early 1990's. Today's rapid advancement in mobile telephone reliability and power management capabilities exemplifies the effectiveness of the SDR technology for the modern communications market. In contrast, the foundations of transponder technology presently qualified for satellite applications were developed during the early space program of the 1960's. Conventional transponders are built to a specific platform and must be redesigned for every new bus while the SDR is adaptive in nature and can fit numerous applications with no hardware modifications. A SDR uses a minimum amount of analog / Radio Frequency (RF) components to up/down-convert the RF signal to/from a digital format. Once the signal is digitized, all processing is performed using hardware or software logic. Typical SDR digital processes include; filtering, modulation, up/down converting and demodulation. NASA Marshall Space Flight Center (MSFC) Programmable Ultra Lightweight System Adaptable Radio (PULSAR) leverages existing MSFC SDR designs and commercial sector enhanced capabilities to provide a path to a radiation tolerant SDR transponder. These innovations (1) reduce the cost of NASA Low Earth Orbit (LEO) and Deep Space standard transponders, (2) decrease power requirements, and (3) commensurately reduce volume. A second pay-off is the increased SDR flexibility by allowing the same hardware to implement multiple transponder types simply by altering hardware logic - no change of hardware is required - all of which will ultimately be accomplished in orbit. Development of SDR technology for space applications will provide a highly capable, low cost transponder to programs of all sizes. The MSFC PULSAR Project results in a Technology Readiness Level (TRL) 7 low-cost telemetry system available to Smallsat and CubeSat missions, as well as other platforms. This paper documents the continued development and verification/validation of the MSFC SDR, called PULSAR, which contributes to advancing the state-of-the-art in transponder design - directly applicable to the SmallSat and CubeSat communities. This paper focuses on lessons learned on the first sub-orbital flight (high altitude balloon) and the follow-on steps taken to validate PULSAR. A sounding rocket launch, currently planned for 03/2015, will further expose PULSAR to the high dynamics of sub-orbital flights. Future opportunities for orbiting satellite incorporation reside in the small satellite missions (FASTSat, CubeSat. etc.).

Simms, William Herbert, III↗

NASA NTRS1996

Development of Collaborative Research Initiatives to Advance the Aerospace Sciences-via the Communications, Electronics, Information Systems Focus Group

The primary goal of the Adaptive Vision Laboratory Research project was to develop advanced computer vision systems for automatic target recognition. The approach used in this effort combined several machine learning paradigms including evolutionary learning algorithms, neural networks, and adaptive clustering techniques to develop the E-MOR.PH system. This system is capable of generating pattern recognition systems to solve a wide variety of complex recognition tasks. A series of simulation experiments were conducted using E-MORPH to solve problems in OCR, military target recognition, industrial inspection, and medical image analysis. The bulk of the funds provided through this grant were used to purchase computer hardware and software to support these computationally intensive simulations. The payoff from this effort is the reduced need for human involvement in the design and implementation of recognition systems. We have shown that the techniques used in E-MORPH are generic and readily transition to other problem domains. Specifically, E-MORPH is multi-phase evolutionary leaming system that evolves cooperative sets of features detectors and combines their response using an adaptive classifier to form a complete pattern recognition system. The system can operate on binary or grayscale images. In our most recent experiments, we used multi-resolution images that are formed by applying a Gabor wavelet transform to a set of grayscale input images. To begin the leaming process, candidate chips are extracted from the multi-resolution images to form a training set and a test set. A population of detector sets is randomly initialized to start the evolutionary process. Using a combination of evolutionary programming and genetic algorithms, the feature detectors are enhanced to solve a recognition problem. The design of E-MORPH and recognition results for a complex problem in medical image analysis are described at the end of this report. The specific task involves the identification of vertebrae in x-ray images of human spinal columns. This problem is extremely challenging because the individual vertebra exhibit variation in shape, scale, orientation, and contrast. E-MORPH generated several accurate recognition systems to solve this task. This dual use of this ATR technology clearly demonstrates the flexibility and power of our approach.

Knasel, T. Michael↗

DOE OSTI2021

Utilizing computer vision and artificial intelligence algorithms to predict and design the mechanical compression response of direct ink write 3D printed foam replacement structures

Additive Manufacturing (AM) of porous polymeric materials, such as foams, recently became a topic of intensive research due their unique combination of low density, impressive mechanical properties, and stress dissipation capabilities. Conventional methods for fabricating foams rely on complex and stochastic processes, making it challenging to achieve precise architectural control of structured porosity. In contrast, AM provides access to a wide range of printable materials, where precise spatial control over structured porosity can be modulated during the fabrication process enabling the production of foam replacement structures (FRS). Current approaches for designing FRS are based on intuitive understanding of their properties or an extensive number of finite element method (FEM) simulations. These approaches, however, are computationally expensive and time consuming. As such, in this work, we present a novel methodology for determining the mechanical compression response of direct ink write (DIW) 3D printed FRS using a simple cross-sectional image. By obtaining measurement data for a relatively small number of samples, an artificial neural network (ANN) was trained, and a computer vision algorithm was used to make inferences about foam compression characteristics from a single cross-sectional image. Finally, a genetic algorithm (GA) was used to solve the inverse design problem, generating the AM printing parameters that an engineer should use to achieve a desired compression response from a DIW printed FRS. The methods developed herein present an avenue for entirely autonomous design and analysis of additively manufactured structures using artificial intelligence.

36 MATERIALS SCIENCE↗

DOE OSTI2018

Predicting the 3D fatigue crack growth rate of small cracks using multimodal data via Bayesian networks: In-situ experiments and crystal plasticity simulations

Small crack propagation accounts for most of the fatigue life of engineering structures subject to high cycle fatigue loading conditions. Determining the fatigue crack growth rate of small cracks propagating into polycrystalline engineering alloys is critical to improving fatigue life predictions, thus lowering cost and increasing safety. In this work, cycle-by-cycle data of a small crack propagating in a beta metastable titanium alloy is available via phase and diffraction contrast tomography. Crystal plasticity simulations are used to supplement experimental data regarding the micromechanical fields ahead of the crack tip. Experimental and numerical results are combined into a multimodal dataset and sampled utilizing a non-local data mining procedure. Furthermore, to capture the propensity of body-centered cubic metals to deform according to the pencil-glide model, a non-local driving force is postulated. The proposed driving force serves as the basis to construct a data-driven probabilistic crack propagation framework using Bayesian networks as building blocks. The spatial correlation between the postulated driving force and experimental observations is obtained by analyzing the results of the proposed framework. Results show that the above correlation increases proportionally to the distance from the crack front until the edge of the plastic zone. Moreover, the predictions of the propagation framework show good agreement with experimental observations. Finally, we studied the interaction of a small crack with grain boundaries (GBs) utilizing various slip transmission criteria, revealing the tendency of a crack to cross a GB by propagating along the slip directions minimizing the residual Burgers vector within the GB.

36 MATERIALS SCIENCE↗

NASA NTRS2002

Stimulus-response compatibility and psychological refractory period effects: implications for response selection

The purpose of this paper was to provide insight into the nature of response selection by reviewing the literature on stimulus-response compatibility (SRC) effects and the psychological refractory period (PRP) effect individually and jointly. The empirical findings and theoretical explanations of SRC effects that have been studied within a single-task context suggest that there are two response-selection routes-automatic activation and intentional translation. In contrast, all major PRP models reviewed in this paper have treated response selection as a single processing stage. In particular, the response-selection bottleneck (RSB) model assumes that the processing of Task 1 and Task 2 comprises two separate streams and that the PRP effect is due to a bottleneck located at response selection. Yet, considerable evidence from studies of SRC in the PRP paradigm shows that the processing of the two tasks is more interactive than is suggested by the RSB model and by most other models of the PRP effect. The major implication drawn from the studies of SRC effects in the PRP context is that response activation is a distinct process from final response selection. Response activation is based on both long-term and short-term task-defined S-R associations and occurs automatically and in parallel for the two tasks. The final response selection is an intentional act required even for highly compatible and practiced tasks and is restricted to processing one task at a time. Investigations of SRC effects and response-selection variables in dual-task contexts should be conducted more systematically because they provide significant insight into the nature of response-selection mechanisms.

Review Literature↗

DOE OSTI2025

Improving the prediction of daily reservoir releases over the CONUS using conditioned LSTM

Reservoirs play a vital role in regulating streamflow timing and variability for hydroelectricity, flood control, water supply, irrigation, and recreation. Despite their importance, many reservoirs lack comprehensive operational guidelines, making their management complex due to conflicting operational objectives. Hence traditional policy-based reservoir models often fail to capture real-world conditions accurately and they depend on perfect streamflow predictions, which are not always available. In contrast, data-driven models like Long Short-Term Memory (LSTM) networks offer a robust alternative. This study introduces an approach that integrates reservoir characteristics—such as main use, climate, and maximum capacity—into the LSTM model to enhance reservoir release predictions. Using data from nearly 200 reservoirs in the contiguous United States (CONUS), our conditioned LSTM model (LSTM_cond) was compared with both the vanila LSTM and a traditional policy-based approach. Furthermore, our results show that while both LSTM_cond and LSTM perfoms better than the policy-based approach, LSTM_cond consistently outperforms LSTM for hydroelectric, water supply, irrigation, and recreation reservoirs. The KGE median values for LSTM_cond for out-sample reservoirs are 0.764, 0.565, 0.821, and 0.779, respectively, for the aforementioned reservoir types, which are consistently higher that the corresponding KGE values of 0.737, 0.413, 0.775, and 0.713 of LSTM, demonstrating its advantages in improving generalizability.

CONUS↗

DOE OSTI2023

Densely Connected G-invariant Deep Neural Networks with Signed Permutation Representations

We introduce and investigate, for finite groups G, G-invariant deep neural network (GDNN) architectures with ReLU activation that are densely connected- i.e., include all possible skip connections. In contrast to other G-invariant architectures in the literature, the preactivations of theG-DNNs presented here are able to transform by signed permutation representations (signed perm-reps) of G. Moreover, the individual layers of the G-DNNs are not required to be G-equivariant; instead, the preactivations are constrained to be G-equivariant functions of the network input in a way that couples weights across all layers. The result is a richer family of G-invariant architectures never seen previously. We derive an efficient implementation of G-DNNs after a reparameterization of weights, as well as necessary and sufficient conditions for an architecture to be "admissible"- i.e., nondegenerate and inequivalent to smaller architectures. We include code that allows a user to build a G-DNN interactively layer-by-layer, with the final architecture guaranteed to be admissible. We show that there are far more admissible G-DNN architectures than those accessible with the "concatenated ReLU" activation function from the literature. Finally, we apply G-DNNs to two example problems--(1) multiplication in --1, 1} (with theoretical guarantees) and (2) 3D object classification--finding that the inclusion of signed perm-reps significantly boosts predictive performance compared to baselines with only ordinary (i.e., unsigned) perm-reps.

97 MATHEMATICS AND COMPUTING↗

NASA NTRSDate not supplied

Shielding Considerations for CubeSat Structures During Solar Maximum

The purpose of this lessons learned is to communicate the utility of shielding in small spacecraft planning for the support of mission assurance and reliability. Numerous SmallSats have been flying in polar low earth orbit for scientific, communications, technology demonstrations, and imaging with academic, commercial, and government interests. Shielding has been part of mission assurance and reliability from the advent of long duration spacecraft missions. The Shields-1 CubeSat has been operating in polar low earth orbit since 16 December 2018 with atomic number (Z)-grade radiation shielding and demonstrates shielding effectiveness. Shields-1 has collected a representative example of solar minimum data in 2019 with 8 Teledyne dosimeters over varying shielding effectiveness. It serves as current experimental data and has been compared with NOVICE Shielding estimates using the AP8 –AE8 trapped radiation model with the Shields-1 CAD and generic CubeSat 3 unit (U) models. Using NOVICE model radiation analysis coding, the shielding effectiveness's, based on a generic CubeSat 3U structure with 4 electronic boards, were estimated for aluminum wall thicknesses ranging from 0.204 cmto4.44 cm (0.550 g/cm2–12.0 g/cm2) thick aluminum. For modeled polar orbiting spacecraft, solar maximum total ionizing dose (TID)increases by nearly a magnitude for thin-walled aluminum 0.550g/cm2-0.686 g/cm2(0.204 cm –0.254 cm) typical CubeSat structures. The shielding effectiveness by NOVICE Sigma estimates, which is a shielding sphere approximation around a detector, showed a linear relationship with wall thickness, which increased over the wall thickness by a ratio of 1.43 determined by linear regression analysis. Using NOVICE Adjoint Monte-Carlo Modeling of solar minimum and solar maximum with the inclusion of a worst-case solar particle event over a one year mission without geomagnetic shielding, the TID for minimum and maximum conditions for a generic 3U with a wall thickness of 0.254 cm is 158 RAD and 1540 RAD, respectively. The modeled total solar maximum TID is over estimated, because at low orbital latitudes a spacecraft will have shielding from Earth's magnetic field. However, TID will still be significant at high latitudes over the poles, where a spacecraft is exposed in a solar particle event. In contrast, to a thin walled generic 3U CubeSat, Shields-1 electronics enclosure has a shielding effectiveness of 21.3 g/cm2from NOVICE Sigma modeling and is expected to show reduced total ionizing dose increases during the present active Solar Cycle 25 period. Because solar particle events during solar maximum increase TID on electronic parts with thin-walled shielding in short periods of time, it is a mission assurance and reliability consideration on the spacecraft’s mission value versus adding shielding for risk reduction of premature spacecraft or instrument payload loss. Since the volumes of many instruments and system electronics have reduced with small spacecraft, shielding material costs and weight penalties have diminished. A small spacecraft project budget and schedule may limit traditional radiation-hardened part use and radiation testing requirements, where shielding can contribute to mission assurance and reliability with reduced costs.

Shields-1↗

DOE OSTI2021

On the use of neural networks for energy reconstruction in high-granularity calorimeters

We contrasted the performance of deep neural networks — Convolutional Neural Network (CNN) and Graph Neural Network (GNN) — to current state of the art energy regression methods in a finely 3D-segmented calorimeter simulated by GEANT4. This comparative benchmark gives us some insight to assess the particular latent signals neural network methods exploit to achieve superior resolution. A CNN trained solely on a pure sample of pions achieved substantial improvement in the energy resolution for both single pions and jets over the conventional approaches. It maintained good performance for electron and photon reconstruction. We also used the Graph Neural Network (GNN) with edge convolution to assess the importance of timing information in the shower development for improved energy reconstruction. We implement a simple simulation based correction to the energy sum derived from the fraction of energy deposited in the electromagnetic shower component. This serves as an approximate dual-readout analogue for our benchmark comparison. Although this study does not include the simulation of detector effects, such as electronic noise, the margin of improvement seems robust enough to suggest these benefits will endure in real-world application. We also find reason to infer that the CNN/GNN methods leverage latent features that concur with our current understanding of the physics of calorimeter measurement.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2018

Lessons Learned From Public Outreach: Effective Incorporation of the Stakeholder Voice - 18447

The licensing process for any facility associated with nuclear power or aspects of the nuclear fuel cycle can be expected to encounter intense stakeholder interest, and often significant challenges. Over the last decade, staff of the Center for Nuclear Waste Regulatory Analyses (CNWRA{sup R}) at Southwest Research Institute{sup R} (SwRI{sup R}) provided comprehensive support to the U.S. Nuclear Regulatory Commission (NRC) for public outreach efforts across a wide variety of nuclear activities. Many of these projects included public involvement and consultation activities for compliance with the National Environmental Policy Act (NEPA), including scoping, public meetings, responding to public comments on draft documents, and holding small meetings with affected local, state, and tribal governments. In addition, we have supported NRC public outreach related to nuclear facility licensing activities (including the proposed repository at Yucca Mountain), rulemaking, and regulatory guidance development. This experience has offered us a broad view of the challenges faced by industry, regulators, and stakeholders. In the process, we have observed some best practices that lead to effective stakeholder communication. Communication about complex technical issues is challenging, particularly for societally controversial topics, such as civilian use of nuclear power and management of radioactive waste. Most commonly, we have observed a lack of mutual trust and respect among industry, regulators, and stakeholders that creates an atmosphere ripe for suspicion and misunderstanding. Stakeholders have, in public comments, criticized the lack of appropriate venues to raise concerns in which they feel they can get meaningful responses from the nuclear industry and regulators. Conversely, industry faces difficulty in reaching compromise with stakeholders, who sometimes misinterpret or misrepresent factual information. Within this contentious environment, regulators are constrained by their statutory framework, even as they are challenged to be more responsive, transparent, and fair to both stakeholders and industry. In spite of these obstacles, there are examples where feedback from stakeholders was heard and utilized. Because of successful collection of appropriate feedback, regulators became more adept at understanding the full range of perspectives from both industry and stakeholders, and subsequently did a better job at effectively incorporating feedback and communicating their decisions and findings in a clear and understandable way. Ultimately, the success of these examples was based on a measure of mutual respect, which was reached through effective communication and implementation of some best practices. These best practices include (i) outreach planning that is abundantly mindful of the various stakeholder points of view, site history, regional characteristics, and probable future and related actions, combined with advance identification of key stakeholders; (ii) development and distribution of information that facilitates public participation; and (iii) use of avenues by which stakeholders can provide meaningful and effective comments. Illustrative examples include (i) how concerns raised by stakeholders were heard by an applicant and resulted in a revised project proposal; (ii) use of a variety of interaction formats that incorporated input across a broad spectrum of stakeholders; (iii) utilizing stakeholder feedback received during scoping; and (iv) how a facility underestimated the impact of 'outsider' stakeholder voices that were in contrast to the known local community. By moving past the viewpoint that public outreach is a 'necessary evil' and engaging only to check a procedural box, each participant can benefit from mutually satisfactory and improved outcomes. (author)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗