Search NASA⌕ Search

SEARCH · Search NASA

Results for “model weights”

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 253 records · Page 14
All sourcesNASA NTRSDOE OSTI
NASA NTRS2014

Engine Power Turbine and Propulsion Pod Arrangement Study

A study has been conducted for NASA Glenn Research Center under contract NNC10BA05B, Task NNC11TA80T to identify beneficial arrangements of the turboshaft engine, transmissions and related systems within the propulsion pod nacelle of NASA's Large Civil Tilt-Rotor 2nd iteration (LCTR2) vehicle. Propulsion pod layouts were used to investigate potential advantages, disadvantages, as well as constraints of various arrangements assuming front or aft shafted engines. Results from previous NASA LCTR2 propulsion system studies and tasks performed by Boeing under NASA contracts are used as the basis for this study. This configuration consists of two Fixed Geometry Variable Speed Power Turbine Engines and related drive and rotor systems (per nacelle) arranged in tilting nacelles near the wing tip. Entry-into-service (EIS) 2035 technology is assumed for both the engine and drive systems. The variable speed rotor system changes from 100 percent speed for hover to 54 percent speed for cruise by the means of a two speed gearbox concept developed under previous NASA contracts. Propulsion and drive system configurations that resulted in minimum vehicle gross weight were identified in previous work and used here. Results reported in this study illustrate that a forward shafted engine has a slight weight benefit over an aft shafted engine for the LCTR2 vehicle. Although the aft shafted engines provide a more controlled and centered CG (between hover and cruise), the length of the long rotor shaft and complicated engine exhaust arrangement outweighed the potential benefits. A Multi-Disciplinary Analysis and Optimization (MDAO) approach for transmission sizing was also explored for this study. This tool offers quick analysis of gear loads, bearing lives, efficiencies, etc., through use of commercially available RomaxDESIGNER software. The goal was to create quick methods to explore various concept models. The output results from RomaxDESIGNER have been successfully linked to Boeing spreadsheets that generate gear tooth geometry in Catia 3D environment. Another initial goal was to link information from RomaxDESIGNER (such as hp, rpm, gear ratio) to populate Boeing's parametric weight spreadsheet and create an automated method to estimate drive system weight. This was only partially achieved due to the variety of weight models, number of manual inputs, and qualitative assessments required. A simplified weight spreadsheet was used with data inputs from RomaxDESIGNER along with manual inputs to perform rough weight calculations.

Robuck, Mark↗
NASA NTRS2017

Modeling and Simulation Plans in Support of Low Cost, Size, Weight, and Power Surveillance Systems for Detecting and Tracking Non-Cooperative Aircraft

RTCA Special Committee (SC) 228 has initiated a second phase for the development of minimum operational performance standards (MOPS) for UAS detect and avoid (DAA) systems. Technologies to enable UAS with less available Size, Weight, and Power (SWaP) will be considered. RTCA SC-228 has established sub-working groups and one of the sub-working groups is focused on aligning modeling and simulations activities across all participating committee members. This briefing will describe NASAs modeling and simulation plans for the development of performance standards for low cost, size, weight, and power (C-SWaP) surveillance systems that detect and track non-cooperative aircraft. The briefing will also describe the simulation platform NASA intends to use to support end-to-end verification and validation for these DAA systems. Lastly, the briefing will highlight the experiment plan for our first simulation study, and provide a high-level description of our future flight test plans. This briefing does not contain any results or data.

surveillance↗
DOE OSTI2025

Dataset for "Machine Learning Ensembles Can Enhance Hydrologic Predictions and Uncertainty Quantification" Willard et al. (2025).

This data release provides all data and code used in the paper " "Machine Learning Ensembles Can Enhance Hydrologic Predictions and Uncertainty Quantifications" Willard et al. (2025)" to model stream temperature, evaluate, and assess results. The associated manuscript explores the effect of different ensemble construction techniques across different common machine learning (ML) architectures for predictions in unmonitored basins. Modeling was done using long short-term memory (LSTM), gated recurrent unit (GRU), temporal convolution network (TCN), and extreme gradient boosting (XGBoost) models, and stream site coverage spans 1362 locations across the conterminous United States. The ensemble construction techniques investigated include ensemble by random weight initialization, differing hyperparameters, different random subsets of training data, different subselections of input features, different architectures, and Monte Carlo Dropout. The data is organized into these items items:Code repository and data for the paper " "Machine Learning Ensembles Can Enhance Hydrologic Predictions and Uncertainty Quantifications" Willard et al. (2025).Code: stream_temp_ml_regionalization.zip contains the code repositoryData to run the code:- data_dir.zip -- contains all files that should be moved to the "DATA_DIR" variable defined in the "set_env_vars.sh" script in the code repository- metadata_dir.zip -- contains all files that should be moved to the "METADATA_DIR" variable defined in the "set_env_vars.sh" script in the code repositoryData produced by the code and used in the paper:- outputs_dir.zip - contains model output and results (outputs_dir/results), model weights (outputs_dir/models), and all other outputs used for the paper including feature importances.To cite this code, please use the following BibTeX or MLA entries:bibtex:@misc{willard2025streamensembles,author = {Jared Willard and Charuleka Varadharajan},title = {Dataset for "Machine Learning Ensembles Can Enhance Hydrologic Predictions and Uncertainty Quantification"},year = {2024},doi = {10.15485/2527393},publisher = {ESS-DIVE Repository},url = {https://data.ess-dive.lbl.gov/datasets/doi:10.15485/2527393}}MLA: Willard, Jared, et al. Dataset for "Machine Learning Ensembles Can Enhance Hydrologic Predictions and Uncertainty Quantification". 2025. ESS-DIVE Repository, doi:10.15485/2448016.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1992

Relating multifrequency radar backscattering to forest biomass: Modeling and AIRSAR measurement

During the last several years, significant efforts in microwave remote sensing were devoted to relating forest parameters to radar backscattering coefficients. These and other studies showed that in most cases, the longer wavelength (i.e. P band) and cross-polarization (HV) backscattering had higher sensitivity and better correlation to forest biomass. This research examines this relationship in a northern forest area through both backscatter modeling and synthetic aperture radar (SAR) data analysis. The field measurements were used to estimate stand biomass from forest weight tables. The backscatter model described by Sun et al. was modified to simulate the backscattering coefficients with respect to stand biomass. The average number of trees per square meter or radar resolution cell, and the average tree height or diameter breast height (dbh) in the forest stand are the driving parameters of the model. The rest of the soil surface, orientation, and size distributions of leaves and branches, remain unchanged in the simulations.

Sun, Guo-Qing↗
NASA NTRS1991

COSP - A computer model of cyclic oxidation

A computer model useful in predicting the cyclic oxidation behavior of alloys is presented. The model considers the oxygen uptake due to scale formation during the heating cycle and the loss of oxide due to spalling during the cooling cycle. The balance between scale formation and scale loss is modeled and used to predict weight change and metal loss kinetics. A simple uniform spalling model is compared to a more complex random spall site model. In nearly all cases, the simpler uniform spall model gave predictions as accurate as the more complex model. The model has been applied to several nickel-base alloys which, depending upon composition, form Al2O3 or Cr2O3 during oxidation. The model has been validated by several experimental approaches. Versions of the model that run on a personal computer are available.

Lowell, Carl E.↗
DOE OSTI2025

Foundational Dataset for Developing Large-Sample Stream Temperature Models in the Conterminous United States

This dataset provides inputs, evaluation results, and trained weights from a large-sample Long Short-Term Memory (LSTM) model designed to predict daily stream temperatures across unregulated river reaches in the conterminous United States (CONUS). It includes dynamic meteorological and hydrologic forcings, static physiographic attributes, and model outputs from cross-validation experiments spanning 300 basins. It supports reproducible modeling, direct application for new basins, and provides data suitable for integration with reservoir and river simulations under current and future climates. It contains two .zip files described below · RQ-AI_runs.zip: Model outputs from 10-fold cross-validation experiments, including observed and predicted daily stream temperatures, along with test performance metrics for water years 2017–2019. Two versions are included: 1. Model trained and validated using subbasin-area weighted dynamic features. 2. Model trained and validated using whole-basin area weighted dynamic features. · RQ-AI_inputs.zip: Collection of all formatted dynamic and static predictor datasets (meteorological, hydrologic, and physiographic features) used in model training and analysis. Detailed instructions and data structure is held at the following GitLab repository: https://code.ornl.gov/tempwise/training.

Gomez-Velez, Jesus [Oak Ridge National Laboratory ↗
NASA NTRS1973

A search for global and seasonal variation of methane from Nimbus 4 IRIS measurements

The Nimbus 4 infrared interferometer spectrometer (IRIS) measurements in the region around wave number 1304 show absorption due to methane in the earth's atmosphere. From the laboratory measurements of the absorption coefficient and a selected vertical distribution corresponding to 1.13 atm cm of methane, a theoretical model for the transmittance at wave number 1304 is developed. The weighting function deduced from this model shows a maximum around 300 mb. Some weak absorption due to nitrous oxide in the atmosphere has been taken into account. The vertical temperature profile, derived from the 15 micron CO2 band in the IRIS spectrum, together with the methane weighting function have been used in a consistent way to compute the upwelling intensity at wave number 1304. The brightness temperature corresponding to the IRIS observed radiance at wave number 1304 has been compared with the brightness temperature deduced from the calculated upwelling intensity from 80 deg North to 80 deg South and for different periods of the year. This comparison shows that the two brightness temperatures agree with one another to within the accuracy of measurements about 2 K. From this result it was found that global or seasonal variability of methane is less than + or - 0.25 atm cm.

Prabhakara, C.↗
DOE OSTI2025

FPGA-accelerated SpeckleNN with SNL for real-time X-ray single-particle imaging

We present the implementation of a specialized version of our previously published unified embedding model, SpeckleNN, for real-time speckle pattern classification in X-ray Single-Particle Imaging (SPI), using the SLAC Neural Network Library (SNL) on an FPGA platform. This hardware realization transitions SpeckleNN from a prototypic model into a practical edge solution, optimized for running inference near the detector in high-throughput X-ray free-electron laser (XFEL) facilities, such as those found at the Linac Coherent Light Source (LCLS). To address the resource constraints inherent in FPGAs, we developed a more specialized version of SpeckleNN. The original model, which was designed for broader classification across multiple biological samples, comprised ~5.6 million parameters. The new implementation, while reducing the parameter count to 64.6K (a 98.8% reduction), focuses on maintaining the model's essential functionality for real-time operation, achieving an accuracy of 90%. Furthermore, we compressed the latent space from 128 to 50 dimensions. This implementation was demonstrated on the KCU1500 FPGA board, utilizing 71% of available DSPs, 75% of LUTs, and 48% of FFs, with an average power consumption of 9.4W according to the Vivado post-implementation report. The FPGA performed inference on a single image with a latency of 45.015 microseconds at a 200 MHz clock rate. In comparison, running the same inference on an NVIDIA A100 GPU resulted in an average power consumption of ~73W and an image processing latency of around 400 microseconds. Our FPGA-accelerated version of SpeckleNN demonstrated significant improvements, achieving an 8.9 × speedup and a 7.8 × reduction in power consumption compared to the GPU implementation. Key advancements include model specialization and dynamic weight loading through SNL, which eliminates the need for time-consuming FPGA design re-synthesis, allowing fast and continuous deployment of models (re)trained online. These innovations enable real-time adaptive classification and efficient vetoing of speckle patterns, making SpeckleNN more suited for deployment in XFEL facilities. This implementation has the potential to significantly accelerate SPI experiments and enhance adaptability to evolving experimental conditions.

47 OTHER INSTRUMENTATION↗
NASA NTRS1985

Application of lagged averaged forecasting to medium range forecasting

The lagged average forecast (LAF) method for predicting and compensating for forecast error growth is applied to two 100 day samples of 10 day forecasts at the 500 mb altitude for winter-summer 1980-81. The LAF parameterizes the forecast error growth in order to weight the forecast with regression calculations. A 5 day LAF forecast was generated for a 10 day forecast made with a 1681 d.o.f. model which produced predictions using a truncated spherical harmonic expansion. A 100 day data set was employed, with the first 90 days serving for LAF forecasts for the model forecast. Significant improvements were obtained in the model forecast when the LAF weightings were introduced into the model variables.

Dalcher, A.↗
NASA NTRS1984

Tissue fluid shift, forelimb loading, and tail tension in tail-suspended rats

The tail suspension model (head-down tilt) simulates hypogravity in terms of musculoskeletal loss in the rat. However, little is known of tissue fluid shifts and body weight distribution in this model. Tissue fluid pressures were measured by wick catheters in 12 Munich-Wistar rats before, during, and after 48 hrs of tail suspension (about 30 deg head-down tilt). Subcutaneous tissue fluid pressure in the neck increased from -2.2 + or - 0.4 (normal horizontal position) to +4.0 + or - 1.5 cm H2O during tail suspension, indicating a cephalic fluid shift and significant edema during head-down tilt. In a separate study, six rats were suspended at 30-70 deg, and forelimb load and tail tension were measured by a balance and force transducer, respectively. Approximately 50 percent of body weight (BW) was loaded on forelimbs at a head-down tilt angle of 30 deg and forelimb load declined linearly to 10 percent BW at 70 deg. Furthermore, tail tension increased from 50 percent BW at 30 deg to 85 percent BW at 70 deg. These results indicate that less than normal loads are applied to forelimbs of rats suspended at angles of less than 30 deg and that the tail bears an increasing proportion of the rat's body weight at head-down tilt angles of less than 30 deg.

Hargens, A. R.↗
DOE OSTI2024

Integrating multi-modal remote sensing, deep learning, and attention mechanisms for yield prediction in plant breeding experiments

In both plant breeding and crop management, interpretability plays a crucial role in instilling trust in AI-driven approaches and enabling the provision of actionable insights. The primary objective of this research is to explore and evaluate the potential contributions of deep learning network architectures that employ stacked LSTM for end-of-season maize grain yield prediction. A secondary aim is to expand the capabilities of these networks by adapting them to better accommodate and leverage the multi-modality properties of remote sensing data. In this study, a multi-modal deep learning architecture that assimilates inputs from heterogeneous data streams, including high-resolution hyperspectral imagery, LiDAR point clouds, and environmental data, is proposed to forecast maize crop yields. The architecture includes attention mechanisms that assign varying levels of importance to different modalities and temporal features that, reflect the dynamics of plant growth and environmental interactions. The interpretability of the attention weights is investigated in multi-modal networks that seek to both improve predictions and attribute crop yield outcomes to genetic and environmental variables. This approach also contributes to increased interpretability of the model's predictions. The temporal attention weight distributions highlighted relevant factors and critical growth stages that contribute to the predictions. The results of this study affirm that the attention weights are consistent with recognized biological growth stages, thereby substantiating the network's capability to learn biologically interpretable features. Accuracies of the model's predictions of yield ranged from 0.82-0.93 R 2 ref in this genetics-focused study, further highlighting the potential of attention-based models. Further, this research facilitates understanding of how multi-modality remote sensing aligns with the physiological stages of maize. The proposed architecture shows promise in improving predictions and offering interpretable insights into the factors affecting maize crop yields, while demonstrating the impact of data collection by different modalities through the growing season. By identifying relevant factors and critical growth stages, the model's attention weights provide valuable information that can be used in both plant breeding and crop management. The consistency of attention weights with biological growth stages reinforces the potential of deep learning networks in agricultural applications, particularly in leveraging remote sensing data for yield prediction. To the best of our knowledge, this is the first study that investigates the use of hyperspectral and LiDAR UAV time series data for explaining/interpreting plant growth stages within deep learning networks and forecasting plot-level maize grain yield using late fusion modalities with attention mechanisms.

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRS2002

Preliminary Thermal-Mechanical Sizing of Metallic TPS: Process Development and Sensitivity Studies

The purpose of this research was to perform sensitivity studies and develop a process to perform thermal and structural analysis and sizing of the latest Metallic Thermal Protection System (TPS) developed at NASA LaRC (Langley Research Center). Metallic TPS is a key technology for reducing the cost of reusable launch vehicles (RLV), offering the combination of increased durability and competitive weights when compared to other systems. Accurate sizing of metallic TPS requires combined thermal and structural analysis. Initial sensitivity studies were conducted using transient one-dimensional finite element thermal analysis to determine the influence of various TPS and analysis parameters on TPS weight. The thermal analysis model was then used in combination with static deflection and failure mode analysis of the sandwich panel outer surface of the TPS to obtain minimum weight TPS configurations at three vehicle stations on the windward centerline of a representative RLV. The coupled nature of the analysis requires an iterative analysis process, which will be described herein. Findings from the sensitivity analysis are reported, along with TPS designs at the three RLV vehicle stations considered.

Poteet, Carl C.↗
NASA NTRS1991

Frequency weighted system identification and linear quadratic controller design

Application of filters for frequency weighting of Markov parameters (pulse response functions) is described in relation to system/observer identification. The time domain identification approach recovers a model which has a pulse response weighted according to frequency. The identified model is composed of the original system and filters. The augmented system is in a form which can be used directly for frequency weighted linear quadratic controller design. Data from either single or multiple experiments can be used to recover the Markov parameters. Measured acceleration signals from a truss structure are used for system identification and the model obtained is used for frequency weighted controller design. The procedure makes the identification and controler design complementary problems.

Horta, Lucas G.↗
NASA NTRS1984

Optimization of space structures

Computational methods for the design of structures for specified transient response, truss beam units with specified attached vibration absorbers, and laminates for structural components of large space structures are examined. Equations for the measurement of structural stiffness that are maximized for a specific total mass and that will reduce the structural weight are presented. A model for a cantilevered space truss beam of a specific mass and with a specified tip vibration absorber is explained. Design criteria of the laminates include minimizing the weight as well as frequency, buckling, and global stiffness constraints. Other variables include orientation of the lamina and the thickness of each layer.

Reiss, R.↗
NASA NTRS1989

Fabrication of light-weighted Si/SiC lidar mirrors

Small models of lightweight Si/SiC lidar mirrors have been fabricated via a scalable and cost-effective chemical vapor deposition (CVD) process. These mirrors are about 7.5 cm in diameter and consist of a Si-cladded SiC faceplate and lightweight structure made of SiC. The lightweight structure consists of an outer hexagonal cell with six triangular inner cells. The mirrors were polished to a figure better than 1/5 of a wave and a finish of better than 5A rms. The CVD route to fabrication of mirrors is fast and has the potential to yield several mirrors in a few weeks time from a signel reactor.

Goela, Jitendra S.↗
NASA NTRS2012

Effects of Tropospheric Spatio-Temporal Correlated Noise on the Analysis of Space Geodetic Data

The standard VLBI analysis models the distribution of measurement noise as Gaussian. Because the price of recording bits is steadily decreasing, thermal errors will soon no longer dominate. As a result, it is expected that troposphere and instrumentation/clock errors will increasingly become more dominant. Given that both of these errors have correlated spectra, properly modeling the error distributions will become increasingly relevant for optimal analysis. We discuss the advantages of modeling the correlations between tropospheric delays using a Kolmogorov spectrum and the frozen flow assumption pioneered by Treuhaft and Lanyi. We then apply these correlated noise spectra to the weighting of VLBI data analysis for two case studies: X/Ka-band global astrometry and Earth orientation. In both cases we see improved results when the analyses are weighted with correlated noise models vs. the standard uncorrelated models. The X/Ka astrometric scatter improved by approx.10% and the systematic Delta delta vs. delta slope decreased by approx. 50%. The TEMPO Earth orientation results improved by 17% in baseline transverse and 27% in baseline vertical.

instrumentation errors↗
NASA NTRS1993

System identification and model reduction using modulating function techniques

Weighted least squares (WLS) and adaptive weighted least squares (AWLS) algorithms are initiated for continuous-time system identification using Fourier type modulating function techniques. Two stochastic signal models are examined using the mean square properties of the stochastic calculus: an equation error signal model with white noise residuals, and a more realistic white measurement noise signal model. The covariance matrices in each model are shown to be banded and sparse, and a joint likelihood cost function is developed which links the real and imaginary parts of the modulated quantities. The superior performance of above algorithms is demonstrated by comparing them with the LS/MFT and popular predicting error method (PEM) through 200 Monte Carlo simulations. A model reduction problem is formulated with the AWLS/MFT algorithm, and comparisons are made via six examples with a variety of model reduction techniques, including the well-known balanced realization method. Here the AWLS/MFT algorithm manifests higher accuracy in almost all cases, and exhibits its unique flexibility and versatility. Armed with this model reduction, the AWLS/MFT algorithm is extended into MIMO transfer function system identification problems. The impact due to the discrepancy in bandwidths and gains among subsystem is explored through five examples. Finally, as a comprehensive application, the stability derivatives of the longitudinal and lateral dynamics of an F-18 aircraft are identified using physical flight data provided by NASA. A pole-constrained SIMO and MIMO AWLS/MFT algorithm is devised and analyzed. Monte Carlo simulations illustrate its high-noise rejecting properties. Utilizing the flight data, comparisons among different MFT algorithms are tabulated and the AWLS is found to be strongly favored in almost all facets.

Shen, Yan↗
NASA NTRS1992

Fitting a three-component scattering model to polarimetric SAR data

A new technique for fitting a three-component scattering mechanism model to the polarimetric synthetic aperture radar (SAR) data itself, without utilizing any ground truth measurements, is presented. The three scattering mechanism components included in the model are volume scatter from randomly oriented dipoles, first-order Bragg surface scatter and a dihedral scattering mechanism for two surfaces with different dielectric constants. The model fit yields an estimate of the contribution to the total backscatter of each of the three components. The backscatter contributions can also be compared to give the relative percentage weight of each. The model fit has an equal number of input parameters (the polarimetric radar backscatter measurements) and output parameters (the backscatter parameters describing them). The model can be applied to entire images or to small areas within an image to give a first-order estimate of the relevant scattering mechanisms. The model was applied to many C-, L- and P-band Airborne SAR (AIRSAR) images of different types of terrain. Results were presented at the workshop.

Freeman, A.↗