Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Deep Operator Networks”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

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

At least 1,207 records · Page 67
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Redox transistors based on TiO 2 for analogue neuromorphic computing

The ability to train deep neural networks on large data sets have made significant impacts onto artificial intelligence, but consume significant amounts of energy due to the need to move information from memory to logic units. In-memory "neuromorphic" computing presents an alternative framework that processes information directly on memory elements. In-memory computing has been limited by the poor performance of the analogue information storage element, often phase-change memory or memristors. To solve this problem, we developed two types of "redox transistors" using TiO 2 (anatase) which stores analogue information states through the electrochemical concentration of dopants in the crystal. The first type of redox transistor uses lithium as the electrochemical dopant ion, and its key advantage is low operating voltage. The second uses oxygen vacancies as the dopant, which is CMOS compatible and can retain state even when scaled to nanosized dimensions. Both devices offer significant advantages in terms of predictable analogue switching over conventional filamentary-based devices, and provide a significant advance in developing materials and devices for neuromorphic computing.

36 MATERIALS SCIENCE↗
NASA NTRS1987

Ka-band (32-GHz) downlink capability for deep space communications

The first quarter century of U.S. solar system exploration using unmanned spacecraft has involved progressively higher operating frequencies for deep space telemetry: L-band (960 MHz) in 1962 to S-band (2.3 GHz) in 1964 to X-band (8.4 GHZ) in 1977. The next logical frequency to develop for deep space is the Ka-band (32 GHz) for which a primary deep space allocation of 500 MHz between 31.8 to 32.3 GHz was established in 1979. The telecommunications capability was improved by a factor of 77 (18.9 dB) through the frequency changes from L-band to X-band. Another improvement factor of 14.5 (11.6 dB) can be achieved by going to Ka-band. Plans to develop and demonstrate Ka-band capability include the continued measurement of weather effects at Deep Space Network (DSN) sites, development of a prototype DSN ground antenna and supporting subsystems, augmentation of planned spacecraft with Ka-band beacons, and development of spacecraft prototype modules for future Ka-band transmitters. Plans for augmenting the DSN with Ka-band capability by 1995 were also developed. A companion set of articles describes the Ka-band performance and technology in greater detail.

Smith, J. G.↗
DOE OSTI2025

Neutrino interaction vertex reconstruction in DUNE with Pandora deep learning

The Pandora Software Development Kit and algorithm libraries perform reconstruction of neutrino interactions in liquid argon time projection chamber detectors. Pandora is the primary event reconstruction software used at the Deep Underground Neutrino Experiment, which will operate four large-scale liquid argon time projection chambers at the far detector site in South Dakota, producing high-resolution images of charged particles emerging from neutrino interactions. While these high-resolution images provide excellent opportunities for physics, the complex topologies require sophisticated pattern recognition capabilities to interpret signals from the detectors as physically meaningful objects that form the inputs to physics analyses. A critical component is the identification of the neutrino interaction vertex. Subsequent reconstruction algorithms use this location to identify the individual primary particles and ensure they each result in a separate reconstructed particle. A new vertex-finding procedure described in this article integrates a U-ResNet neural network performing hit-level classification into the multi-algorithm approach used by Pandora to identify the neutrino interaction vertex. The machine learning solution is seamlessly integrated into a chain of pattern-recognition algorithms. The technique substantially outperforms the previous BDT-based solution, with a more than 20% increase in the efficiency of sub-1 cm vertex reconstruction across all neutrino flavours.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRS2013

Rover Low Gain Antenna Qualification for Deep Space Thermal Environments

A method to qualify the Rover Low Gain Antenna (RLGA) for use during the Mars Science Laboratory (MSL) mission has been devised. The RLGA antenna must survive all ground operations, plus the nominal 670 Martian sol mission that includes the summer and winter seasons of the Mars thermal environment. This qualification effort was performed to verify that the RLGA design, its bonding, and packaging processes are adequate. The qualification test was designed to demonstrate a survival life of three times more than all expected ground testing, plus a nominal 670 Martian sol missions. Baseline RF tests and a visual inspection were performed on the RLGA hardware before the start of the qualification test. Functional intermittent RF tests were performed during thermal chamber breaks over the course of the complete qualification test. For the return loss measurements, the RLGA antenna was moved to a test area. A vector network analyzer was calibrated over the operational frequency range of the antenna. For the RLGA, a simple return loss measurement was performed. A total of 2,010 (3 670 or 3 times mission thermal cycles) thermal cycles was performed. Visual inspection of the RLGA hardware did not show any anomalies due to the thermal cycling. The return loss measurement results of the RLGA antenna after the PQV (Package Qualification and Verification) test did not show any anomalies. The antenna pattern data taken before and after the PQV test at the uplink and downlink frequencies were unchanged. Therefore, the developed design of RLGA is qualified for a long-duration MSL mission.

Ramesham, Rajeshuni↗
NASA NTRSDate not supplied

BioSentinel Deep Space CubeSat Mission

The BioSentinel mission was recently launched aboard the SLS launch vehicle (LV) as part of the Artemis-1 campaign. This 6U CubeSat carries yeast cells to analyze the effects of radiation at large distances from Earth, becoming the first biological payload in Deep Space. Prelaunch activities included mission design updates, orbit determination rehearsals and the development of a tracking schedule in coordination with the Artemis-1 payload office and the Deep Space Network (DSN). An important influence on the trajectories of Artemis I secondaries was the uncertainty associated with deployment from the Interim Cryogenic Propulsion System (ICPS), the upper stage of the SLS LV. The ICPS was rotating at a rate of 1 rpm; there was also uncertainty in the spin axis attitude, which translated into an unknown clock angle of deployment. The variability in this angle and magnitude of deployment implied the existence of a non-negligible risk of a lunar impact, which was evaluated for various potential launch dates. On November 16 th 2022 BioSentinel successfully deployed from ICPS and the navigation team started to receive tracking data from the DSN and ESA antennas. Soon after deployment, the spacecraft was tumbling and entered safe mode. The mission team recovered the spacecraft and after four tracking passes, we solved for a first ephemeris that was sent to the DSN for better tracking of the spacecraft. After propagating this first ephemeris solution, we determined that we avoided impact with a margin of a few hundred km from the lunar surface. More tracking data over the next few days allowed for a more refined orbit solution predicting a periselene altitude of 406 km and a lunar eclipse lasting 36.5 minutes. Therefore, BioSentinel operators avoided any correction maneuvers on the trajectory and successfully tracked and guide the spacecraft. The spacecraft performed a nominal lunar flyby which provided the pertinent energy to achieve a final Earth-trailing heliocentric orbit. Over the course of two weeks, the mission operators corroborated that the subsystems were functioning as expected after the lunar eclipse and the large ΔV incurred. Science operations started once the mission achieved the nominal orbit in Deep Space. This paper discusses in detail the BioSentinel flight performance, as well as the challenges and lessons learned prior to and during this CubeSat mission.

Andres Dono Perez↗
DOE OSTI2020

High precision control and deep learning-based corn stand counting algorithms for agricultural robot

This paper presents high precision control and deep learning-based corn stand counting algorithms for a low-cost, ultra-compact 3D printed and autonomous field robot for agricultural operations. Currently, plant traits, such as emergence rate, biomass, vigor, and stand counting, are measured manually. This is highly labor-intensive and prone to errors. The robot, termed TerraSentia, is designed to automate the measurement of plant traits for efficient phenotyping as an alternative to manual measurements. In this paper, we formulate a Nonlinear Moving Horizon Estimator that identifies key terrain parameters using onboard robot sensors and a learning-based Nonlinear Model Predictive Control that ensures high precision path tracking in the presence of unknown wheel-terrain interaction. Moreover, we develop a machine vision algorithm designed to enable an ultra-compact ground robot to count corn stands by driving through the fields autonomously. The algorithm leverages a deep network to detect corn plants in images, and a visual tracking model to re-identify detected objects at different time steps. We collected data from 53 corn plots in various fields for corn plants around 14 days after emergence (stage V3 - V4). The robot predictions have agreed well with the ground truth with C robot =1.02×C human -0.86 and a correlation coefficient R=0.96. The mean relative error given by the algorithm is -3.78%, and the standard deviation is 6.76%. These results indicate a first and significant step towards autonomous robot-based real-time phenotyping using low-cost, ultra-compact ground robots for corn and potentially other crops.

97 MATHEMATICS AND COMPUTING↗
NASA NTRSDate not supplied

An Envisioned Future for Space Optical Communications

Since the beginning of the Space Age, NASA has been a leader in developing space communications and navigation technologies— especially during the Apollo missions to the Moon and NASA’s initial foray into deep space. To support future exploration and science needs, NASA is gradually introducing optical communications technologies to augment its radio frequency (RF) systems. Optical communications will enable new science and exploration missions by providing high data rates and better navigation over long distances. NASA has already flown several optical communications demonstrations, including the Lunar Laser Communications Demonstration (LLCD), the Laser Communications Relay Demonstration (LCRD), and the Terabyte Infrared Delivery (TBIRD) system. Historically, NASA has partnered with the Jet Propulsion Laboratory (JPL) and the Massachusetts Institute of Technology Lincoln Laboratory (MIT/LL) to develop optical communications technology. In addition to pursuing optical communications, NASA’s Space Communications and Navigation (SCaN) Program is undergoing a paradigm shift and moving away from government owned and operated networks to using commercial services whenever possible. In partnership with SCaN, NASA’s Space Technology Mission Directorate (STMD) has identified key technologies that need to be developed to support future space communications and navigation, including enhanced RF, optical, and 3rd Generation Partnership (3GPP) cellular capabilities, as well as high-speed networking. This paper briefly describes some current and upcoming optical demonstrations and provides an overview of STMD’s envisioned future for optical communications and navigation in the 2030+ timeframe.

Bernard L Edwards↗
DOE OSTI2022

Exact constraints and appropriate norms in machine-learned exchange-correlation functionals

Machine learning techniques have received growing attention as an alternative strategy for developing general-purpose density functional approximations, augmenting the historically successful approach of human-designed functionals derived to obey mathematical constraints known for the exact exchange-correlation functional. More recently, efforts have been made to reconcile the two techniques, integrating machine learning and exact-constraint satisfaction. We continue this integrated approach, designing a deep neural network that exploits the exact constraint and appropriate norm philosophy to de-orbitalize the strongly constrained and appropriately normed (SCAN) functional. The deep neural network is trained to replicate the SCAN functional from only electron density and local derivative information, avoiding the use of the orbital-dependent kinetic energy density. The performance and transferability of the machine-learned functional are demonstrated for molecular and periodic systems.

Artificial neural networks↗
DOE OSTI2021

Deep reinforcement learning for dynamic control of fuel injection timing in multi-pulse compression ignition engines

Conventional compression-ignition (CI) engines have long offered high thermal efficiencies and torque across a wide range of loads, but often require extensive exhaust gas treatment that decreases efficiency to meet ever-increasing emissions regulations. One strategy to decrease emissions is to split the fuel injection into a series of smaller injections. In this paper, we explore a new way of discovering optimal control strategies for the next generation of CI engines using deep reinforcement learning (DRL). We outline a DRL procedure to maximize the weighted reward of engine work while minimizing end-of-cycle NO x emissions. Through the procedure outlined in this paper, we show that the DRL agent is able to reduce NO x emissions threefold while only decreasing network by 2%. We demonstrate the use of transfer learning (TL) across hierarchies of physical models to accelerate the learning process, making this approach feasible for a range of control problems within this space. This paper presents a framework and demonstration for using DRL to design control systems in technology areas such as multi-pulse engine control where a hierarchy of models combined with multi-objective rewards are used for optimal operation.

33 ADVANCED PROPULSION SYSTEMS↗
NASA NTRS2016

Determination of Ceres Physical Parameters Using Radiometric and Optical Data

The Dawn spacecraft was launched on September 27th, 2007. Its mission is to rendezvous with and observe the two largest bodies in the main asteroid belt, Vesta and Ceres. It has completed over a year’s worth of direct observations of Vesta from early 2011 through late 2012. In the spring of 2015, the Dawn spacecraft entered orbit around the asteroid Ceres for the start of what is expected to be more than a year of science operations. The science data collected from this encounter consist of infrared (IR) images and spectra, visible images through a number of color filters, gamma ray detections and measurements of the Ceres gravity field. These data will be collected during several science phases: an Approach phase (1500000-4860 km from Ceres), a Survey orbit (4860 km radius), a High Altitude Mapping Orbit (HAMO) (1940 km radius) and a Low Altitude Mapping Orbit (LAMO) (855 km radius). The Approach phase included three Rotational Characterization (RC) opportunities. Designing each science orbit and successfully transferring into that orbit requires a sufficiently accurate estimate of Ceres physical parameters (body fixed frame, GM and harmonics). This paper focuses on work performed to estimate Ceres physical parameters using Deep Space Network (DSN) radiometric tracking data and optical measurements derived from science camera imagery. This paper describes planning for the data acquisition, as well as processing techniques and methodology. The trajectories predicted by the gravity field estimations are also compared with the actual as-flown trajectories. Observations of the gravity at high altitudes are found to be sufficient to design precision orbits at lower altitudes. Follow-up analysis after successfully reaching LAMO is included, as is a discussion of lessons learned.

Kennedy, Brian M.↗
DOE OSTI2022

Data-Driven Prediction and Optimization of Energy Use for Transit Fleets of Electric and ICE Vehicles

Due to the high upfront cost of electric vehicles, many public transit agencies can afford only mixed fleets of internal combustion and electric vehicles. Optimizing the operation of such mixed fleets is challenging because it requires accurate trip-level predictions of electricity and fuel use as well as efficient algorithms for assigning vehicles to transit routes. We present a novel framework for the data-driven prediction of trip-level energy use for mixed-vehicle transit fleets and for the optimization of vehicle assignments, which we evaluate using data collected from the bus fleet of CARTA, the public transit agency of Chattanooga, TN. We first introduce a data collection, storage, and processing framework for system-level and high-frequency vehicle-level transit data, including domain-specific data cleansing methods. We train and evaluate machine learning models for energy prediction, demonstrating that deep neural networks attain the highest accuracy. Based on these predictions, we formulate the problem of minimizing energy use through assigning vehicles to fixed-route transit trips. We propose an optimal integer program as well as efficient heuristic and meta-heuristic algorithms, demonstrating the scalability and performance of these algorithms numerically using the transit network of CARTA.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
NASA NTRS2018

Recommendations Emerging from an Analysis of NASA's Deep Space Communications Capacity

During 2016-2017, NASA’s Space Communications and Navigation (SCaN) Office chartered a study of Deep Space Network (DSN) communications capacity relative to projected future-mission demand over the next 30 years. In this paper, we briefly describe the methodology used to analyze capacity vs. demand over such a broad timeframe, summarize key findings emerging from the analysis, and discuss the associated recommendations. Performing the analysis entailed: identifying key factors shaping the anticipated future mission set, identifying several alternative future mission set scenarios consistent with these factors, and then analyzing each mission set scenario in terms of required antenna capacity, downlink and uplink capabilities, and spectrum as a function of time. On the basis of these aggregate requirements, DSN loading simulations were then conducted that examined how well each of the postulated mission sets could load up onto the the DSN’s “in-plan” architecture. To the extent that capacity shortfalls emerged during these baseline simulations, architectural solutions to the shortfalls were then postulated and tested via additional simulations. In general, the trend analyses and baseline loading simulations indicated a significant progression in challenges over the next three decades. In the current decade, the DSN appeared to be operating very close to capacity. The first human exploration mission and its secondary payload launch opportunities for cubesats traveling beyond GEO contributed to this loading. As a consequence, the main challenge appeared to be managing peak assetcontention periods. In the next decade, the DSN continued to operate close to capacity but also began transitioning to more frequent human mission support. Upgrading for, and operating, a human-rated system while continuing to meet robotic mission customer requirements emerged as the key challenge. In the 2030’s and beyond, simulations suggested a need for fundamentally new capability and capacity. The high data rates and long link distances characteristic of human Mars exploration drove requirements far beyond what is currently “in plan.” The key challenge then became determining the most cost-effective combination of RF and optical assets for communicating with the postulated human Mars assets while still providing for the needs of all the other missions across the solar system. Various link budget, visibility, and loading analyses ultimately suggested that the human Mars exploration demands of the 2030’s could best be addressed with two cross-linked RF-optical areostationary relays (or an areostationary relay and deep space habitat) providing a dual “trunk link” to an array of 2-to-3 additional 34m beam waveguide antennas and an ~8.5m optical antenna at each DSN Complex. The dual “trunk link” would enable the same amount of total data return to Earth as a single trunk link at twice the data rate, but with only half the required array size on the ground, assuming use of Multiple Spacecraft Per Antenna (MSPA) techniques. MSPA techniques, including a Multiple Uplink Per Antenna (MUPA) technique currently under investigation, also showed promise for reducing asset contention in the decades prior to human Mars exploration.

Lee, Carlyn-Ann↗
DOE OSTI2022

Scalable algorithms for physics-informed neural and graph networks

Physics-informed machine learning (PIML) has emerged as a promising new approach for simulating complex physical and biological systems that are governed by complex multiscale processes for which some data are also available. In some instances, the objective is to discover part of the hidden physics from the available data, and PIML has been shown to be particularly effective for such problems for which conventional methods may fail. Unlike commercial machine learning where training of deep neural networks requires big data, in PIML big data are not available. Instead, we can train such networks from additional information obtained by employing the physical laws and evaluating them at random points in the space–time domain. Such PIML integrates multimodality and multifidelity data with mathematical models, and implements them using neural networks or graph networks. Here, we review some of the prevailing trends in embedding physics into machine learning, using physics-informed neural networks (PINNs) based primarily on feed-forward neural networks and automatic differentiation. For more complex systems or systems of systems and unstructured data, graph neural networks (GNNs) present some distinct advantages, and here we review how physics-informed learning can be accomplished with GNNs based on graph exterior calculus to construct differential operators; we refer to these architectures as physics-informed graph networks (PIGNs). We present representative examples for both forward and inverse problems and discuss what advances are needed to scale up PINNs, PIGNs and more broadly GNNs for large-scale engineering problems.

42 ENGINEERING↗
NASA NTRS2002

Converging voice and data over mission-critical networks

The U.S. National Aeronautics and Space Administration (NASA) Deep Space Network ?? DSN - is an international network of antennas that supports interplanetary spacecraft missions and radio and radar astronomy observations for the exploration of the solar system and the universe. The network also supports selected Earth-orbiting missions. This paper describes the ground communications network of the DSN and ways network infrastructure costs are being reduced by the introduction of new technology. Fundamentally, the DSN ground network architecture is a star network, and the hub is at JPL in Pasadena, California. Communications to customer sites are designed to minimize NASA costs and may be either shared IP backbone networks or dedicated circuits. One of the primary features of the network is its ability to support real-time data, voice, and video communications among antenna stations, an automated multi-mission operations systems facility at JPL (AMMOS), and mission operations centers (MOCs) at NASA and non-NASA facilities. Funding for advanced engineering to reduce ground network infrastructure costs was provided by the NASA Office of Space Science.

ground communications networks Internet voice VOIP↗
NASA NTRS2015

Kinematic and Microphysical Control of Lightning Flash Rate over Northern Alabama

The Deep Convective Clouds and Chemistry (DC3) experiment seeks to examine the relationship between deep convection and the production of nitrogen oxides (NO (sub x)) via lightning (LNO (sub x)). A critical step in estimating LNO (sub x) production in a cloud-resolving model (CRM) without explicit lightning is to estimate the flash rate from available model parameters that are statistically and physically correlated. As such, the objective of this study is to develop, improve and evaluate lightning flash rate parameterizations in a variety of meteorological environments and storm types using radar and lightning mapping array (LMA) observations taken over Northern Alabama from 2005-2012, including during DC3. UAH's Advanced Radar for Meteorological and Operational Research (ARMOR) and the Weather Surveillance Radar - 1988 Doppler (WSR 88D) located at Hytop (KHTX) comprises the dual-Doppler and polarimetric radar network, which has been in operation since 2004. The northern Alabama LMA (NA LMA) in conjunction with Vaisala's National Lightning Detection Network (NLDN) allow for a detailed depiction of total lightning during this period. This study will integrate ARMOR-KHTX dual Doppler/polarimetric radar and NA LMA lightning observations from past and ongoing studies, including the more recent DC3 results, over northern Alabama to form a large data set of 15-20 case days and over 20 individual storms, including both ordinary multicell and supercell convection. Several flash rate parameterizations will be developed and tested, including those based on 1) graupel/small hail volume; 2) graupel/small hail mass, and 3) convective updraft volume. Sensitivity of the flash rate parameterizations to storm intensity, storm morphology and environmental conditions will be explored.

Carey, Lawrence D.↗
NASA NTRS1991

Geostationary Operational Environmental Satellite (GOES I-M) METSAT project

The primary objective of the METSAT project is to provide a satellite system that meets the National Environmental Satellite Data and Information Service (NESDIS) requirements as specified by NOAA. For the GOES I-M spacecraft, these requirements include an imager and sounder system, a data collection system, and a search and rescue system. The Space Transportation System and GOES I-M development philosophy and the satellite transfer orbit philosophy are explained. The spacecraft will be launched from the Kennedy Space Center using Atlas G/Centaur D-1A expendable launch vehicles. The coverage required for launch and the support of transfer and drift orbits will consist of the 26-m antenna as prime and the TBD antenna as backup for 11 days at all complexes. There will be contingency support 15 days for on-station spacecraft checkout. After the initial 30 to 45 days, the Deep Space Network (DSN) is committed for emergency support. Contingency and emergency support will be provided by Goldstone only. Information is given in tabular form for DSN support, frequency assignments, telemetry, command, and tracking support responsibility.

Blaney, K.↗
DOE OSTI2024

Sentinel

Network intrusion detection systems (NIDS) are commonplace in network security but they frequently employ algorithms that are computational demanding requiring hardware and software with significant power requirements. Two examples of such resource-intensive algorithms used for network security are regular expression matching and broader signature pattern matching which are commonly used in deep packet inspection (DPI). Network security algorithms that have large power requirements may be a challenge for low-power internet-of-things (IoT) environments, which generally lack the power resources to implement complex security measures like computationally expensive DPI at the edge. Furthermore, IoT environments incorporating 5G standalone networks have network latency constraints beyond just power that make DPI at the edge even more difficult. Programmable logic is ideally suited for machine learning inference for DPI because of its deep instruction level parallelism and single-cycle memory access. Machine learning approaches for DPI have been explored before using the programmable logic of field programmable gate arrays (FPGA) as a potential solution for NIDS approaches that would be power-suitable for IoT. However, those previous programmable logic NIDS approaches utilize either a supervised or unsupervised learning model. Sentinel utilizes the ensemble of these two machine learning approaches known as a semi-supervised approach which has shown promise in NIDS implementations. Sentinel provides a programmable logic implementation of a semi-supervised approach for DPI which operates at much lower power and latency than a GPU implementation with negligible loss of accuracy due to quantization through a logistic regressor.

Anderson, MatthewW [Idaho National Laboratory (INL↗