SEARCH · Engineering Papers

Results for “Model error”

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 1,459 records · Page 81

All sources NASA NTRS DOE OSTI

NASA NTRS2015

Pressure Loss Predictions of the Reactor Simulator Subsystem at NASA GRC

Testing of the Fission Power System (FPS) Technology Demonstration Unit (TDU) is being conducted at NASA GRC. The TDU consists of three subsystems: the Reactor Simulator (RxSim), the Stirling Power Conversion Unit (PCU), and the Heat Exchanger Manifold (HXM). An Annular Linear Induction Pump (ALIP) is used to drive the working fluid. A preliminary version of the TDU system (which excludes the PCU for now), is referred to as the RxSim subsystem and was used to conduct flow tests in Vacuum Facility 6 (VF 6). In parallel, a computational model of the RxSim subsystem was created based on the CAD model and was used to predict loop pressure losses over a range of mass flows. This was done to assess the ability of the pump to meet the design intent mass flow demand. Measured data indicates that the pump can produce 2.333 kg/sec of flow, which is enough to supply the RxSim subsystem with a nominal flow of 1.75 kg/sec. Computational predictions indicated that the pump could provide 2.157 kg/sec (using the Spalart-Allmaras turbulence model), and 2.223 kg/sec (using the k- turbulence model). The computational error of the predictions for the available mass flow is -0.176 kg/sec (with the S-A turbulence model) and -0.110 kg/sec (with the k-epsilon turbulence model) when compared to measured data.

Turbulence Models↗

NASA NTRS2016

Pressure Loss Predictions of the Reactor Simulator Subsystem at NASA Glenn Research Center

Testing of the Fission Power System (FPS) Technology Demonstration Unit (TDU) is being conducted at NASA Glenn Research Center. The TDU consists of three subsystems: the reactor simulator (RxSim), the Stirling Power Conversion Unit (PCU), and the heat exchanger manifold (HXM). An annular linear induction pump (ALIP) is used to drive the working fluid. A preliminary version of the TDU system (which excludes the PCU for now) is referred to as the "RxSim subsystem" and was used to conduct flow tests in Vacuum Facility 6 (VF 6). In parallel, a computational model of the RxSim subsystem was created based on the computer-aided-design (CAD) model and was used to predict loop pressure losses over a range of mass flows. This was done to assess the ability of the pump to meet the design intent mass flow demand. Measured data indicates that the pump can produce 2.333 kg/sec of flow, which is enough to supply the RxSim subsystem with a nominal flow of 1.75 kg/sec. Computational predictions indicated that the pump could provide 2.157 kg/sec (using the Spalart-Allmaras (S‒A) turbulence model) and 2.223 kg/sec (using the k- turbulence model). The computational error of the predictions for the available mass flow is ‒0.176 kg/sec (with the S-A turbulence model) and -0.110 kg/sec (with the k- turbulence model) when compared to measured data.

Mass Flow↗

NASA NTRSDate not supplied

Enhancing Air Traffic Control Planning with Automatic Speech Recognition

The decisions made during the Federal Aviation Administration Air Traffic Control System Command Center's planning teleconferences hold significant sway over the National Airspace System. Held every two hours, these teleconferences convene air traffic managers and stakeholders from across the nation to discuss airspace conditions, weather, and constraints, leading to the formulation and adjustment of traffic management initiatives. Given the critical nature of these decisions, the need for accurate and efficient record-keeping is paramount. In recent years, the application of automatic speech recognition has gained popularity across diverse industries, including aviation. While traditional applications focus on transcribing air traffic control communication, this paper explores a unique application of automatic speech recognition by converting the audio from planning teleconferences into text transcriptions. This innovative approach addresses key challenges in the field, presenting potential benefits for quality assurance, real-time participation, and downstream natural language processing tasks. A notable breakthrough in the machine learning community, namely the transformer neural network architecture, forms the backbone of the proposed solution in this paper. The transformer architecture's role in this research represents a paradigm shift in the efficiency of automatic speech recognition models. By reducing the amount of in-domain training data required, this architecture allows for the fine-tuning of such models like Whisper, originally pretrained on vast English speech datasets. The adaptability of the transformer architecture proves invaluable in capturing the nuances of aviation terminology and specific language used in planning teleconferences. Leveraging the Whisper model as a baseline, our research details the fine-tuning and validation using a dataset comprising 20 hours of meticulously transcribed planning teleconferences. Notably, the baseline pretrained Whisper model exhibited a word error rate of 18.77%. Through the fine-tuning process, the model achieved a substantial improvement, demonstrating an impressive performance with a reduced word error rate of 6.82%. This substantial decrease in WER not only highlights the effectiveness of the transformer architecture but also emphasizes the practical advancements achieved through the application of automatic speech recognition in this specific domain. The utilization of automatic speech recognition in planning teleconferences in this work introduces several novelties. Firstly, the creation of text transcriptions offers a valuable tool for quality assurance and facilitates the efficient review of teleconferences. This is an important aspect of the proposed solution, given the time-sensitive and high-stakes nature of decisions made during these meetings. Furthermore, text-searchable transcriptions provide a streamlined approach for locating and validating critical information, potentially saving hours of manual effort in searching through audio recordings. Moreover, our research identifies a key use case for external facilities and stakeholders. In situations where attendance at the planning teleconference is not feasible, having access to text transcriptions in real-time or shortly after the teleconference ends, proves to be a time-saving and informative resource. This feature enhances collaboration and ensures that stakeholders can stay abreast of important discussions and decisions even in their absence. Despite the efficiency gains facilitated by the transformer architecture in automatic speech recognition technology, it is essential to acknowledge the human factors in data creation. Subject matter experts play a crucial role in accurately transcribing planning teleconferences due to the specificity and complexity of the information discussed. The research dataset, consisting of 20 hours of transcribed planning teleconferences, forms the foundation for fine-tuning and validating the Whisper model. The achieved word error rate of 6.82% demonstrates promising advancements, particularly in recognizing essential aviation terminology within the teleconferences. In conclusion, this paper presents a comprehensive exploration of the application of automatic speech recognition in Air Traffic Control System Command Center planning teleconferences, leveraging the transformer architecture for enhanced efficiency. The novel contributions lie in the improved accessibility of decision-making records, real-time participation opportunities for external stakeholders, and the potential for downstream natural language processing advancements. As the aviation industry continues to evolve, the integration of automatic speech recognition technologies holds the promise of revolutionizing decision-making processes and contributing to the overall safety and efficiency of air traffic management.

ATM↗

NASA NTRS1996

Uncertainty Modeling for Structural Control Analysis and Synthesis

The development of an accurate model of uncertainties for the control of structures that undergo a change in operational environment, based solely on modeling and experimentation in the original environment is studied. The application used throughout this work is the development of an on-orbit uncertainty model based on ground modeling and experimentation. A ground based uncertainty model consisting of mean errors and bounds on critical structural parameters is developed. The uncertainty model is created using multiple data sets to observe all relevant uncertainties in the system. The Discrete Extended Kalman Filter is used as an identification/parameter estimation method for each data set, in addition to providing a covariance matrix which aids in the development of the uncertainty model. Once ground based modal uncertainties have been developed, they are localized to specific degrees of freedom in the form of mass and stiffness uncertainties. Two techniques are presented: a matrix method which develops the mass and stiffness uncertainties in a mathematical manner; and a sensitivity method which assumes a form for the mass and stiffness uncertainties in macroelements and scaling factors. This form allows the derivation of mass and stiffness uncertainties in a more physical manner. The mass and stiffness uncertainties of the ground based system are then mapped onto the on-orbit system, and projected to create an analogous on-orbit uncertainty model in the form of mean errors and bounds on critical parameters. The Middeck Active Control Experiment is introduced as experimental verification for the localization and projection methods developed. In addition, closed loop results from on-orbit operations of the experiment verify the use of the uncertainty model for control analysis and synthesis in space.

Campbell, Mark E.↗

NASA NTRS1983

Distributed control of large space antennas

A systematic way to choose control design parameters and to evaluate performance for large space antennas is presented. The structural dynamics and control properties for a Hoop and Column Antenna and a Wrap-Rib Antenna are characterized. Some results of the effects of model parameter uncertainties to the stability, surface accuracy, and pointing errors are presented. Critical dynamics and control problems for these antenna configurations are identified and potential solutions are discussed. It was concluded that structural uncertainties and model error can cause serious performance deterioration and can even destabilize the controllers. For the hoop and column antenna, large hoop and long meat and the lack of stiffness between the two substructures result in low structural frequencies. Performance can be improved if this design can be strengthened. The two-site control system is more robust than either single-site control systems for the hoop and column antenna.

Cameron, J. M.↗

DOE OSTI2026

Tula: Optimizing Time, Cost, and Generalization in Distributed Large-Batch Training

Distributed training increases the number of batches processed per iteration either by scaling-out (adding more nodes) or scaling-up (increasing the batch-size). However, the largest configuration does not necessarily yield the best performance. Horizontal scaling introduces additional communication overhead, while vertical scaling is constrained by computation cost and device memory limits. Thus, simply increasing the batch-size leads to diminishing returns: training time and cost decrease initially but eventually plateaus, creating a knee-point in the time/cost vs. batch-size pareto curve. The optimal batch-size therefore depends on the underlying model, data and available compute resources. Large batches also suffer from worse model quality due to the well-known “generalization gap”. In this paper, we present Tula, an online service that automatically optimizes time, cost, and convergence quality for large-batch training of convolutional models. It combines parallel-systems modeling with statistical performance prediction to identify the optimal batchsize. Tula predicts training time and cost within 7.5−14% error across multiple models, and achieves up to 20× overall speedup and improves test accuracy by ≈9% on average over standard large-batch training on various vision tasks, thus successfully mitigating the generalization gap and accelerating training at the same time.

Tyagi, Sahil [ORNL] (ORCID:0009000783144745)↗

NASA NTRS1988

Impact of device level faults in a digital avionic processor

This paper describes an experimental analysis of the impact of gate and device-level faults in the processor of a flight control system. Via mixed mode simulation faults were injected both at the gate (stuck-at) and at the transistor levels, and their propagation through the chip to the output pins was measured. The results show that there is little correspondence between a stuck-at and a device-level fault model insofar as error activity or detection within a functional unit is concerned. Insofar as error activity outside the injected unit and at the output pins are concerned, the stuck-at and device models track each other, although the stuck-at model overestimates, by over one hundred percent, the probability of fault propagation to the output pins. The stuck-at model significantly underestimates the impact of an internal chip fault on the output pins.

Kim, S.↗

NASA NTRS1994

Anticipatory Neurofuzzy Control

Technique of feedback control, called "anticipatory neurofuzzy control," developed for use in controlling flexible structures and other dynamic systems for which mathematical models of dynamics poorly known or unknown. Superior ability to act during operation to compensate for, and adapt to, errors in mathematical model of dynamics, changes in dynamics, and noise. Also offers advantage of reduced computing time. Hybrid of two older fuzzy-logic control techniques: standard fuzzy control and predictive fuzzy control.

Mccullough, Claire L.↗

NASA NTRS2000

Tidal Implications for the GRACE Satellite Mission

A fundamental part of the processing of the GRACE satellite data will be correction for the gravitational effect of tides. This includes solid earth, ocean, and atmospheric tides. The most problematic are ocean tides in shallow seas and in high latitudes (above the 66$^{\circ}$ inclination of Topex). Somewhat smaller, but still important, errors are associated with the S$_1$ air tide and the ocean's dynamic response to it. In this paper we explore several facets of the tidal error budget. The ocean tide error spectrum is estimated (a) by inverse methods and (b) by simple differencing of several of the best tidal models. The tidal error spectrum is flatter than the tidal signal spectrum, and it exceeds 10\% of the signal at degrees 15 and above. The tide errors also exceed the anticipated sensitivity of GRACE for all degrees below 40, and possibly below 50. The assumption of constant seawater density (of about 1035 kg\,M$^{-3}$) is a minor error source. Owing to GRACE's high inclination, solar tides are aliased to long periods, which is potentially problematic for geophysical interpretation of the gravity signal. Simulations of the effect of high-latitude solar-tide errors on gravity estimates around Antarctica will be discussed.

Ray, Richard D.↗

NASA NTRS2000

Sampling Errors of SSM/I and TRMM Rainfall Averages: Comparison with Error Estimates from Surface Data and a Sample Model

Quantitative use of satellite-derived maps of monthly rainfall requires some measure of the accuracy of the satellite estimates. The rainfall estimate for a given map grid box is subject to both remote-sensing error and, in the case of low-orbiting satellites, sampling error due to the limited number of observations of the grid box provided by the satellite. A simple model of rain behavior predicts that Root-mean-square (RMS) random error in grid-box averages should depend in a simple way on the local average rain rate, and the predicted behavior has been seen in simulations using surface rain-gauge and radar data. This relationship was examined using satellite SSM/I data obtained over the western equatorial Pacific during TOGA COARE. RMS error inferred directly from SSM/I rainfall estimates was found to be larger than predicted from surface data, and to depend less on local rain rate than was predicted. Preliminary examination of TRMM microwave estimates shows better agreement with surface data. A simple method of estimating rms error in satellite rainfall estimates is suggested, based on quantities that can be directly computed from the satellite data.

Bell, Thomas L.↗

DOE OSTI2026

Benchmarking the performance of uncertainty quantification methods for neural network-based interatomic potentials

Machine-learned interatomic potentials (ML-IAPs) continue to gain popularity as accurate, computationally efficient replacements for traditional, physics-based interatomic potentials and expensive ab initio methods. Uncertainty quantification (UQ) of ML-IAPs is a growing area of research as UQ is critical in many applications of IAPs, such as developing curated datasets, active learning-based data augmentation, self-improving models, and estimating the uncertainty of molecular dynamics simulations. In this paper, we construct and benchmark a series of different neural network potentials (NNPs) with varying network architectures to determine the performance of these models with respect to both the mean and uncertainty calibration error. Each NNP method is specifically designed to predict either epistemic or aleatoric uncertainty with particular focus on the differences in behavior between the epistemic and aleatoric uncertainty estimates. We benchmark these methods using multiple datasets common in the ML-IAP literature. The results show that the aleatoric uncertainty from single-shot model architectures is a competitive alternative to ensemble-based epistemic uncertainty predictions in regions of sufficient data-density. However, in regions where the representative data is sparse, aleatoric uncertainty models tend to overpredict and epistemic methods tend to underpredict the actual model error. We conclude that the type of UQ is crucial when discussing performance of probabilistic model results as different methods have different performance characteristics depending on the regime in which they are evaluated. Therefore, the type of UQ method should be carefully evaluated against both the data characteristics and requirements for the intended application.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2021

Exploring Autoencoder-based Error-bounded Compression for Scientific Data

Error-bounded lossy compression is becoming an indispensable technique for the success of today's scientific projects with vast volumes of data produced during the simulations or instrument data acquisitions. Not only can it significantly reduce data size, but it also can control the compression errors based on user-specified error bounds. Autoencoder (AE) models have been widely used in image compression, but few AE-based compression approaches support error-bounding features, which are highly required by scientific applications. To address this issue, we explore using convolutional autoencoders to improve error-bounded lossy compression for scientific data, with the following three key contributions. (1) We provide an in-depth investigation of the characteristics of various autoencoder models and develop an error-bounded autoencoder-based framework in terms of the SZ model. (2) We optimize the compression quality for main stages in our designed AE-based error-bounded compression framework, fine-tuning the block sizes and latent sizes and also optimizing the compression efficiency of latent vectors. (3) We evaluate our proposed solution using five real-world scientific datasets and comparing them with six other related works. Experiments show that our solution exhibits a very competitive compression quality from among all the compressors in our tests. In absolute terms, it can obtain a much better compression quality (100%similar to 800% improvement in compression ratio with the same data distortion) compared with SZ2.1 and ZFP in cases with a high compression ratio.

Liu, Jinyang↗

NASA NTRS1986

Tracking and Data Relay Satellite (TDRS) tracking and orbit determination

Analysis results were derived using the Goddard Trajectory Determination System to determine the orbit of TDRS-East. A batch weighted least-squares algorithm is used to fit the Bilateration Ranging Transponder system tracking measurements. Six state parameters (position and velocity components) and the solar radiation pressure coefficient are estimated. The solutions are obtained by fitting the data over 34 hr arcs that overlap by 10 hr. Ephemerides are generated over the 34 hr arcs and then compared over the overlapping regions. The position differences indicate the consistency of the solutions and give a measure of their accuracy. The effect of the modeling of perturbative forces (solar radiation pressure, nonspherical geopotential field) on these results are studied. Error analysis is performed using the ORAN program to estimate the effect of force-model errors and measurement-related errors on the TDRS-East ephemeris.

Teles, J.↗

NASA NTRS1995

In situ measurements of BrO during AASE II

BrO measured from the NASA ER-2 during Airborne Arctic Stratospheric Expedition (AASE) II exhibited a mean value (for 20-minute averages) of 5.4 +/- 0.3 pptv, with a standard deviation of 3.1 pptv. Ratios of BrO to available inorganic bromine (Br(sub y)) show only slight increases in polar regions relative to midlatitudes. A comparison between observed latitudinal and diurnal variations of this same ratio and that calculated by photochemical models shows reasonable agreement in behavior, but significant discrepancies in magnitude. It is unclear whether this difference is due to errors in measurements, models or both.

Avallone, L. M.↗

DOE OSTI2022

Verification of the 3-Region Advanced Test Reactor MCNP Model

The verification of the 3-region homogenized fuel Advanced Test Reactor MCNP model. The 3-region model was compared to the 19-plate model found in the 94-CIC report. Flux tallies, energy deposition tallies, and quarter core mesh tallies were used to compare the two models. The 3-region model needed updating in order to make good comparisons between the models. The percent error from the flux and energy deposition tallies data shows that experiment positions inside the flux trap have higher errors than positions outside the fuel ring. The standard deviation data obtained from the mesh tallies shows that the two models agree within two standard deviations throughout the reactor. It is concluded that the model works adequately for what it is used for.

99 GENERAL AND MISCELLANEOUS↗

DOE OSTI2022

Verification of the 3-Region Advanced Test Reactor MCNP Model

99 GENERAL AND MISCELLANEOUS↗

NASA NTRS2019

Evaluation Analysis of NASA SMAP L3 and L4 and SPoRT-LIS Soil Moisture Data in the United States

Soil moisture has a critical role in the development, frequency and persistence of climatic and hydrologic extremes such as drought, heat wave and flooding events. In situ soil moisture data are uneven and sparse in time and space. This highlights the need to utilize other soil moisture sources to fill this spatiotemporal gap. The goal of this study is to validate one satellite-based and two model-based soil moisture datasets with in situ data across the United States. Soil moisture information from the Soil Moisture Active Passive (SMAP) enhanced level 3 (L3) (SMAP L3) and modeled level 4 (L4) (SMAP L4) data at 9-km resolution and Short-term Prediction Research and Transition-Land Information System (SPoRT-LIS) modeled at 3-km resolution were selected for evaluation. SPoRT-LIS is a near real-time, high resolution operational land analysis data. Ground-based data were obtained from the North American Soil Moisture Database (NASMD) for 362 stations. Seven statistical indicators including anomaly, Spearman, and Pearson correlation coefficients, the systematic error (Bias), root mean square error (RMSE), unbiased root mean square error (ubRMSE), and normalized standard deviation (SDV) were used to evaluate the satellite- and model-derived soil moisture data. In addition, the triple collocation (TC) error model was used to measure the error among SMAP L4, SPoRT-LIS and ground-based data. This study assesses which satellite or modeled dataset is most appropriate for specific times and locations to use as a surrogate for in situ observations. Temporal and spatial analysis demonstrated that, overall, SMAP L4 performed better than SMAP L3 and SPoRT-LIS. Strong agreement was observed between SMAP L4 and in situ observations (ρ = 0.53, Bias = −0.006) in all seasons and most regions with various land covers, especially in winter and in the central regions of the United States. For croplands, SMAP L4 presented the best agreement with in situ data, analyzing all period (ρ = 0.60) and non-winter period (ρ = 0.61) separately.

Tavakol, Ameneh↗

NASA NTRS1999

Improved Earth Sensor Performance Using a Sequentially Correlated Noise Model

Spacecraft attitude estimation by means of an extended Kalman filter requires a reasonably true model of the inherent noise of each sensor. For some sensors, the largest uncorrected noise comes not from the sensor itself, but from errors in the model for the predicted observations. This is certainly the case for Earth horizon sensors. The Earth horizon as seen from low Earth orbit is nearly a circle whose radius depends primarily on altitude. A straightforward latitude-dependent correction is added to this to account for the oblateness of the Earth. There also are both seasonal and stochastic variations in the horizon height. The seasonal variations can be predicted to some limited degree based on models derived from historical data. The stochastic component characteristically shows variations that are correlated both in time and space but which are unpredictable over long time spans. This work investigates whether Earth horizon sensor performance can be improved by solving for its systematic error as an augmentation of an attitude Kalman filter. It is found that using only Earth and Sun sensors, the augmented state is not fully observable. Even when magnetometer data is included, only the pitch axis component of the error can be improved; the roll component is unobservable.

Sedlak, J.↗