Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Deep operator network”

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 1,225 records · Page 68
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Machine learning-based real-time kinetic profile reconstruction in DIII-D

Abstract Kinetic equilibrium reconstruction plays a vital role in the physical analysis of plasma stability and control in fusion tokamaks. However, the traditional approach is subjective and prone to human biases. To address this, the consistent automatic kinetic equilibrium reconstruction (CAKE) method was introduced, providing objective results. Nonetheless, its offline nature limits its application in real-time plasma control systems (PCSs). To address this limitation, we present RTCAKENN, a machine learning model that approximates 7 CAKE-level output profiles, namely pressure, inverse q , toroidal current density, electron temperature and density, carbon ion impurity temperature and rotation profiles, using real-time available inputs. The deep neural network consists of an encoder layer, where the scalars and interdependent inputs such as plasma boundary coordinates and motional Stark effect data are encoded using multi-layer perceptrons (MLPs), while profile inputs are encoded by 1D convolutional layers. The encoded data is passed through a MLP for latent feature extraction, before being decoded in the decoding layers, which consist of upsampling and convolutional layers. RTCAKENN has been implemented in the DIII-D PCS and our model achieves accuracy comparable to CAKE and surpasses existing real-time alternatives. Through clever dropout training, RTCAKENN exhibits robustness and can operate even in the absence of Thomson scattering data or charge exchange recombination data. It executes in under 8 ms in the real-time environment, enabling future application in real-time control and analysis.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
NASA NTRS2005

A high performance frequency standard and distribution system for Cassini Ka-band experiment

This paper provides an overview and update of a specialized frequency reference system for the NASA Deep Space Network (DSN) to support Ka-band radio science experiments with the Cassini spacecraft, currently orbiting Saturn. Three major components, a Hydrogen Maser, Stabilized Fiber-optic Distribution Assembly (SFODA), and 10 Kelvin Cryocooled Sapphire Oscillator (10K CSO) and frequency-lock-loop, are integrated to achieve the very high performance, ground based frequency reference at a remote antenna site located 16 km from the hydrogen maser. Typical measured Allan Deviation is 1.6 -14 1 0a't 1 second and 1.7 x 10 -15 at 1000 seconds averaging intervals. Recently two 10K CSOs have been compared in situ while operating at the remote DSN site DSS-25. The CSO references were used operationally to downconvert the Ka band downlink received from the Cassini spacecraft in a series of occultation measurements performed over a 78 day period from March to June 2005.

sapphire↗
DOE OSTI2020

Deep Learning for Automated Identification of Eels in Sonar Data

Freshwater eels, such as the American eel (Anguilla rostrata) present numerous challenges related to safe downstream fish passage at hydroelectric facilities. One of those challenges is effective monitoring of their abundance, movements, and behavior to facilitate design and operation of eel protection and passage facilities. A previous EPRI study documented the ability of human analysts to reliably identify American eels in data obtained with a 1100/1800 kHz, multibeam sonar. This report describes a project to develop deep learning (a subset of artificial intelligence) tools to automate the time-consuming, subjective process of eel identification in multibeam sonar data. The project exploited new data collected in the laboratory and the existing data from the prior EPRI field study to develop and test deep learning and other data analytic tools, including wavelet filtering, differencing for static object removal, and convolutional neural network analysis. The analysis of the laboratory data demonstrated feasibility of the approach, revealed object characteristics observed with the sonar that distinguish eels from similarly sized and shaped acoustic targets, and provided additional data for algorithm selection and training. Deep learning algorithms trained and tested on the laboratory data alone achieved accuracy rates of greater than 98% when classifying acoustic images of eels and similar-sized neutrally buoyant sticks. The algorithm trained and tested on the pre-existing field data alone, and yielded classification accuracy of 9.3% false positives and 13.3% false negatives when distinguishing between eels and sticks/PVC pipes based on video clips (i.e., multiple, consecutive images). This performance is comparable to the classification accuracy achieved by human analysts in the prior study. The deep learning algorithm trained on a combination of video clips obtained in the laboratory and the field and tested on video clips from the field, was able to distinguish eels from sticks and PVC pipes (a river debris analog) of similar size with 100% accuracy. Outreach to the hardware, software, and end-user communities early in the project helped to identify needs and specify the application space. Outreach to those communities at the end of the project communicated project results and opportunities for further development. The project achieved proof of concept for automated identification of eel in multibeam sonar data. Future work should focus on acquisition of additional data for more robust algorithm training and testing; modification of the software tools to accommodate multiple acoustic targets in the acoustic field at a given time; identification of additional object classes; incorporation of motion in the object identification and classification algorithms; operationalizing the software tools, including integration with other existing sonar data analysis tools; and partnering with hardware and software providers for distribution of the software tools with their commercial products.

13 HYDRO ENERGY↗
DOE OSTI2017

Deconstruction of the Ras switching cycle through saturation mutagenesis

Ras proteins are highly conserved signaling molecules that exhibit regulated, nucleotide-dependent switching between active and inactive states. The high conservation of Ras requires mechanistic explanation, especially given the general mutational tolerance of proteins. Here, we use deep mutational scanning, biochemical analysis and molecular simulations to understand constraints on Ras sequence. Ras exhibits global sensitivity to mutation when regulated by a GTPase activating protein and a nucleotide exchange factor. Removing the regulators shifts the distribution of mutational effects to be largely neutral, and reveals hotspots of activating mutations in residues that restrain Ras dynamics and promote the inactive state. Evolutionary analysis, combined with structural and mutational data, argue that Ras has co-evolved with its regulators in the vertebrate lineage. Overall, our results show that sequence conservation in Ras depends strongly on the biochemical network in which it operates, providing a framework for understanding the origin of global selection pressures on proteins.

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRS2011

Hayabusa Re-Entry: Trajectory Analysis and Observation Mission Design

On June 13th, 2010, the Hayabusa sample return capsule successfully re-entered Earth s atmosphere over the Woomera Prohibited Area in southern Australia in its quest to return fragments from the asteroid 1998 SF36 Itokawa . The sample return capsule entered at a super-orbital velocity of 12.04 km/sec (inertial), making it the second fastest human-made object to traverse the atmosphere. The NASA DC-8 airborne observatory was utilized as an instrument platform to record the luminous portion of the sample return capsule re-entry (~60 sec) with a variety of on-board spectroscopic imaging instruments. The predicted sample return capsule s entry state information at ~200 km altitude was propagated through the atmosphere to generate aerothermodynamic and trajectory data used for initial observation flight path design and planning. The DC- 8 flight path was designed by considering safety, optimal sample return capsule viewing geometry and aircraft capabilities in concert with key aerothermodynamic events along the predicted trajectory. Subsequent entry state vector updates provided by the Deep Space Network team at NASA s Jet Propulsion Laboratory were analyzed after the planned trajectory correction maneuvers to further refine the DC-8 observation flight path. Primary and alternate observation flight paths were generated during the mission planning phase which required coordination with Australian authorities for pre-mission approval. The final observation flight path was chosen based upon trade-offs between optimal viewing requirements, ground based observer locations (to facilitate post-flight trajectory reconstruction), predicted weather in the Woomera Prohibited Area and constraints imposed by flight path filing deadlines. To facilitate sample return capsule tracking by the instrument operators, a series of two racetrack flight path patterns were performed prior to the observation leg so the instruments could be pointed towards the region in the star background where the sample return capsule was expected to become visible. An overview of the design methodologies and trade-offs used in the Hayabusa re-entry observation campaign are presented.

Cassell, Alan M.↗
DOE OSTI2020

Artificial Intelligence Applications in Renewable Energy

Addressing new methodologies in deep learning (DL), machine learning (ML) and artificial intelligence (AI), the webinar speakers will provide an overview of the literature spanning these three overlapping fields as applied to energy systems research. The audience will learn how developments in these areas have added new capabilities for pattern recognition and predictive modeling that are complementary to more traditional modeling approaches used in energy systems research. The speakers will illustrate several use cases of these new approaches in the energy space, such as physics-guided neural networks to improve ML regressions for solar data and how AI agents can be used to explore power plant operations.

40 EE - Geothermal Technologies Office (EE-4G)↗
DOE OSTI2024

Efficacy of using a dynamic length representation vs. a fixed-length for neuroarchitecture search

Deep learning neuroarchitecture and hyperparameter search are important in finding the best configuration that maximizes learned model accuracy. However, the number of types of layers, their associated hyperparameters, and the myriad of ways to connect layers poses a significant computational challenge in discovering ideal model configurations. Here, we assess two different approaches for neuroarchitecture search for a LeNet style neural network, one that uses a fixed-length approach where there is a preset number of possible layers that can be toggled on or off via mutation, and a variable-length approach where layers can be freely added or removed via special mutation operators. We found that the variable-length implementation trained better models while discovering unusual layer configurations worth further exploration.

Coletti, Mark↗
DOE OSTI2022

DeepZipper: A Novel Deep-learning Architecture for Lensed Supernovae Identification

Large-scale astronomical surveys have the potential to capture data on large numbers of strongly gravitationally lensed supernovae (LSNe). To facilitate timely analysis and spectroscopic follow-up before the supernova fades, an LSN needs to be identified soon after it begins. To quickly identify LSNe in optical survey data sets, we designed ZipperNet, a multibranch deep neural network that combines convolutional layers (traditionally used for images) with long short-term memory layers (traditionally used for time series). We tested ZipperNet on the task of classifying objects from four categories—no lens, galaxy-galaxy lens, lensed Type-Ia supernova, lensed core-collapse supernova—within high-fidelity simulations of three cosmic survey data sets: the Dark Energy Survey, Rubin Observatory’s Legacy Survey of Space and Time (LSST), and a Dark Energy Spectroscopic Instrument (DESI) imaging survey. Among our results, we find that for the LSST-like data set, ZipperNet classifies LSNe with a receiver operating characteristic area under the curve of 0.97, predicts the spectroscopic type of the lensed supernovae with 79% accuracy, and demonstrates similarly high performance for LSNe 1–2 epochs after first detection. We anticipate that a model like ZipperNet, which simultaneously incorporates spatial and temporal information, can play a significant role in the rapid identification of lensed transient systems in cosmic survey experiments.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
NASA NTRS1993

PC4CAST: A tool for DSN load forecasting and capacity planning

Effectively planning the use and evolution of the Deep Space Network (DSN) is a complex problem involving many parameters. The tool that models many of these complexities, yet requires simple structured inputs and provides concise easy-to-understand metrics to aid in the planning process is discussed. The tool, PC4CAST, is used for both load forecasting (predicting how well planned that DSN resources meet expected demand) and as a decision support tool in the capacity-planning process (determining the relative benefits of capacity expansion options). It is now in use in the TDA Planning Office, has been used in numerous studies, and is also being used by the JPL Multimission Operations System Office (MOSO) as an integral part of Resource Allocation Team activities. Experience using the tool has helped to identify additional requirements that will further improve the planning process, which can be met by future PC4CAST versions.

Loyola, S. J.↗
NASA NTRS2011

X-Band Acquisition Aid Software

The X-band Acquisition Aid (AAP) software is a low-cost acquisition aid for the Deep Space Network (DSN) antennas, and is used while acquiring a spacecraft shortly after it has launched. When enabled, the acquisition aid provides corrections to the antenna-predicted trajectory of the spacecraft to compensate for the variations that occur during the actual launch. The AAP software also provides the corrections to the antenna-predicted trajectory to the navigation team that uses the corrections to refine their model of the spacecraft in order to produce improved antenna-predicted trajectories for each spacecraft that passes over each complex. The software provides an automated Acquisition Aid receiver calibration, and provides graphical displays to the operator and remote viewers via an Ethernet connection. It has a Web server, and the remote workstations use the Firefox browser to view the displays. At any given time, only one operator can control any particular display in order to avoid conflicting commands from more than one control point. The configuration and control is accomplished solely via the graphical displays. The operator does not have to remember any commands. Only a few configuration parameters need to be changed, and can be saved to the appropriate spacecraft-dependent configuration file on the AAP s hard disk. AAP automates the calibration sequence by first commanding the antenna to the correct position, starting the receiver calibration sequence, and then providing the operator with the option of accepting or rejecting the new calibration parameters. If accepted, the new parameters are stored in the appropriate spacecraft-dependent configuration file. The calibration can be performed on the Sun, greatly expanding the window of opportunity for calibration. The spacecraft traditionally used for calibration is in view typically twice per day, and only for about ten minutes each pass.

Britcliffe, Michael J.↗
NASA NTRS2012

Telerobotics Workstation (TRWS) for Deep Space Habitats

On medium- to long-duration human spaceflight missions, latency in communications from Earth could reduce efficiency or hinder local operations, control, and monitoring of the various mission vehicles and other elements. Regardless of the degree of autonomy of any one particular element, a means of monitoring and controlling the elements in real time based on mission needs would increase efficiency and response times for their operation. Since human crews would be present locally, a local means for monitoring and controlling all the various mission elements is needed, particularly for robotic elements where response to interesting scientific features in the environment might need near- instantaneous manipulation and control. One of the elements proposed for medium- and long-duration human spaceflight missions, the Deep Space Habitat (DSH), is intended to be used as a remote residence and working volume for human crews. The proposed solution for local monitoring and control would be to provide a workstation within the DSH where local crews can operate local vehicles and robotic elements with little to no latency. The Telerobotics Workstation (TRWS) is a multi-display computer workstation mounted in a dedicated location within the DSH that can be adjusted for a variety of configurations as required. From an Intra-Vehicular Activity (IVA) location, the TRWS uses the Robot Application Programming Interface Delegate (RAPID) control environment through the local network to remotely monitor and control vehicles and robotic assets located outside the pressurized volume in the immediate vicinity or at low-latency distances from the habitat. The multiple display area of the TRWS allows the crew to have numerous windows open with live video feeds, control windows, and data browsers, as well as local monitoring and control of the DSH and associated systems.

Mittman, David S.↗
DOE OSTI2015

Wireless Advanced Lighting Controls Retrofit Demonstration

Awareness of the economic costs and environmental consequences of electric energy use in buildings is steadily growing. Implementing energy efficiency measures for reducing energy consumption in buildings can be an effective strategy for managing these impacts. Within commercial buildings in the United States, electricity use for lighting accounts for 26% (around 346 TWh in 2010) and represents a large potential energy savings source. Significant lighting energy savings have already been achieved through the widespread adoption of efficient fluorescent lamps and ballasts in the past several decades. Looking towards efficient lighting operations’ innovations, wireless advanced lighting controls technology represents an increasingly viable option for capturing the next major level of lighting energy savings in new construction and building retrofits. Fundamentally, lighting is the most amenable building end use load for producing deep energy savings because of its dynamic controllability. Advanced lighting controls encompass control strategies from occupancy sensing to continuous dimming, institutional tuning and daylight harvesting. These strategies offer greater flexibility and higher granularity of control than traditional basic control methods. Historically, advanced control systems have required extensive control wiring, which has driven system costs. However, recently developed wireless lighting controls systems can be used to network lighting components, while potentially minimizing installation time and labor costs during retrofit in comparison to wired control systems.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2026

Machine Learning-Enabled Wearable Piezoelectric Acoustic Sensor for Real-Time Breast Abnormality Detection

In contemporary society, breast health has become a significant public health concern, particularly among women. According to statistics from the World Health Organization, both the incidence and mortality rates of breast tumors have steadily increased in recent years. Therefore, effective early-stage screening and postoperative monitoring are essential for maintaining breast health. However, conventional clinical diagnostic modalities are typically bulky, operationally complex, and unsuitable for continuous real-time monitoring, which limits their use in portable and everyday health management applications. To address these limitations, this study proposes a machine learning-integrated wearable piezoelectric sensing platform as an auxiliary tool for breast health assessment. The device consists of a PDMS matching layer embedded with flexible silver nanowires, a P(VDF-TrFE) piezoelectric layer, and a multi-channel low-noise signal acquisition circuit. It is capable of acquiring acoustic echo signals from tissue-mimicking environments and automatically evaluating signal validity using a convolutional neural network (CNN). By integrating piezoelectric sensing with deep learning-based signal analysis, the proposed system achieves a signal-to-noise ratio exceeding 70 dB and a real-time classification accuracy above 96% under controlled conditions. These results demonstrate that the platform provides a compact, portable, and intelligent approach for wearable sensing of mechanical heterogeneity and highlight its potential for future development in continuous biomedical monitoring technologies.

He, Shuaitong↗
DOE OSTI2023

Informed Feature Selection for Data Clustering of CSP Plant Production

To make concentrating solar power (CSP) more cost competitive, rigourous optimizations must be run to improve plant design and operations. However, these optimizaitons rely on time consuming annual simulations that solve an electricity dispatch scheduling problem to maximize plant revenue. To reduce the runtime of annual dispatch simulations of CSP plants, a data clustering approach is utilized. This approach assumes that like days of revenue and electricity generation can be identified using weather and price data. Although weather and price are important factors for electricity production, this work investigates how thermal energy storage (TES) inventory at the beginning of a day, denoted as Si, can be used as a supplemental feature to group like days. A framework for creating and training a deep neural network to predict Si is proposed. This model is validated and assessed using eleven sets of testing data that were not used during training. Then, the data clustering approach is performed three seperate times with features of weather and price along with either Si from the neural network, Si from the full annual simulation, or no Si. Ultimately, the results suggest that using Si as an additional clustering feature improves the data clustering simulation accuracy by 1.4%.

Tuman, Matthew J. (ORCID:000900038772051X)↗
DOE OSTI2022

Metaplastic and energy-efficient biocompatible graphene artificial synaptic transistors for enhanced accuracy neuromorphic computing

CMOS-based computing systems that employ the von Neumann architecture are relatively limited when it comes to parallel data storage and processing. In contrast, the human brain is a living computational signal processing unit that operates with extreme parallelism and energy efficiency. Although numerous neuromorphic electronic devices have emerged in the last decade, most of them are rigid or contain materials that are toxic to biological systems. In this work, we report on biocompatible bilayer graphene-based artificial synaptic transistors (BLAST) capable of mimicking synaptic behavior. The BLAST devices leverage a dry ion-selective membrane, enabling long-term potentiation, with ~50 aJ/µm 2 switching energy efficiency, at least an order of magnitude lower than previous reports on two-dimensional material-based artificial synapses. The devices show unique metaplasticity, a useful feature for generalizable deep neural networks, and we demonstrate that metaplastic BLASTs outperform ideal linear synapses in classic image classification tasks. With switching energy well below the 1 fJ energy estimated per biological synapse, the proposed devices are powerful candidates for bio-interfaced online learning, bridging the gap between artificial and biological neural networks.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Can Third-Body Stabilization of Bimolecular Collision Complexes in Cold Molecular Clouds Happen?

For more than half a century, networks of radiative association, dissociative recombination, and bimolecular reactions have been postulated to drive the low-temperature chemistry of cold molecular clouds. Third-body stabilizations of collision complexes have been assumed to be 'irrelevant' due to short lifetime of such complexes. Here, we conduct crossed molecular beam studies of ground state atomic silicon with diacetylene in combination with electronic structure calculations and microcanonical kinetics models operating under cold molecular cloud conditions. Our combined experimental, electronic structure, and microcanonical kinetics modelling investigations provide compelling evidence that three-body collisions of molecular hydrogen with long-lived reaction intermediates accessed through intersystem crossing are prevalent deep inside molecular clouds. This concept might be exportable to reactions involving polycyclic aromatic hydrocarbons thus affording a versatile machinery to complex organics via third-body stabilizations of bimolecular collision complexes deep inside cold molecular clouds.

astrochemistry↗
DOE OSTI2020

nPINNs: nonlocal Physics-Informed Neural Networks for a parametrized nonlocal universal Laplacian operator. Algorithms and Applications

Physics-informed neural networks (PINNs) are effective in solving inverse problems based on differential and integro-differential equations with sparse, noisy, unstructured, and multifidelity data. PINNs incorporate all available information, including governing equations (reflecting physical laws), initial-boundary conditions, and observations of quantities of interest, into a loss function to be minimized, thus recasting the original problem into an optimization problem. In this paper, we extend PINNs to parameter and function inference for integral equations such as nonlocal Poisson and nonlocal turbulence models, and we refer to them as nonlocal PINNs (nPINNs). The contribution of the paper is three-fold. First, we propose a unified nonlocal Laplace operator, which converges to the classical Laplacian as one of the operator parameters, the nonlocal interaction radius δ goes to zero, and to the fractional Laplacian as δ goes to infinity. This universal operator forms a super-set of classical Laplacian and fractional Laplacian operators and, thus, has the potential to fit a broad spectrum of data sets. We provide theoretical convergence rates with respect to δ and verify them via numerical experiments. Second, we use nPINNs to estimate the two parameters, δ and α, characterizing the kernel of the unified operator. The strong non-convexity of the loss function yielding multiple (good) local minima reveals the occurrence of the operator mimicking phenomenon, that is, different pairs of estimated parameters could produce multiple solutions of comparable accuracy. Third, we propose another nonlocal operator with spatially variable order α(γ), which is more suitable for modeling turbulent Couette flow. Our results show that nPINNs can jointly infer this function as well as δ. More importantly, these parameters exhibit a universal behavior with respect to the Reynolds number, a finding that contributes to our understanding of nonlocal interactions in wall-bounded turbulence.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS2017

AI and Autonomy Initiatives for NASA’s Deep Space Network (DSN)

NASA’s Deep Space Network (DSN) consists of thirteen large (34- and 70-meter) antennas that are used to communicate with approximately 40 NASA and partner spacecraft, all at great distance from the earth (generally at Lunar distances and beyond). The DSN has a long history — over 50 years — and has evolved with cutting edge, often custom, telecommunications equipment and associated software systems. In recent years, and in preparation for an increasing future demand, there has been an effort to invest in initiatives that will result in significant cost savings in the future. These efforts are building on, or augmenting, the recent deployment of “Follow-the-Sun” operations (day shift remote operational control of the entire network from each of the three antenna complexes in turn) — which is being deployed in 2017. This paper focuses on Adaptive Demand Access: in a paradigm shift from completely pre-planned operations, this concept calls for spacecraft to signal their intent (or not) for near-future contacts, in case they have science results of interest, or have experienced an anomaly. This would take advantage of a beacon tone transmission, which can be detected using smaller antennas. When a connection request is received, the DSN ground systems would adaptively accommodate the request, inserting the contact into the plan as soon as possible, subject to constraints and priorities. The demand access concept incorporates onboard data analysis and science data processing, so that beacon tones can be generated with maximum information. This area is representative of several where infusing AI technologies can lead to improved effectiveness of the DSN as the network readies for support of expanded Mars exploration efforts in the 2020’s and beyond.

Wyatt, E. Jay↗