Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Noisy data”

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 631 records · Page 35
All sourcesNASA NTRSDOE OSTI
NASA NTRS1984

Landsat-4 sensor performance

Preflight and in-orbit sensor and data measurements indicate that TM meets or exceeds most specifications. Measured spectral band edges meet instrument specifications in 12 out of 14 cases; there is ample dynamic range. The signal-to-noise ratio exceeds specifications, except for band 3, channel 4; and band 7 channel 7 is very noisy but still meets specifications. The modulation transfer function of channel 4, band 2, is smaller than specified. Registration errors between the primary focal plane (PFP) and the cold focal plane (CFP) are about 0.75 pixels along-scan and 0.2 pixels across scan. Forward and reverse scan discontinuities are well within ground-processing capabilities to rectify. Instrument gain variability, up to 7 percent for band 5, requires use of the internal calibration (IC) system to assure radiometric accuracy. Preliminary applications evaluation of image contents indicates that TM provides much better definition of edges than MSS.

Barker, J. L.↗
NASA NTRS2014

TPSAS-NF1676L-19259-DND

In climate and Earth energy budget studies, understanding the presence and distribution of clouds is an essential first step. AVHRR measurements comprise the longest record of imager data from a common instrument for remote sensing of cloud and radiation properties. The NASA LaRC AVHRR cloud mask is a global cloud detection algorithm used as part of the Climate Data Record (CDR) project initiated by NOAA's National Climatic Data Center (NCDC). The objective of CDR program is to provide climate information derived from weather satellite data for more than 30 years. Recent improvements in the LaRC AVHRR cloud mask include detecting more daytime small cumulus and thin Cirrus in tropical oceans, especially over Sun Glint regions, better discrimination between polar clouds and snow ice surfaces including nighttime over Antarctica where 3.7 μm signals are noisy, and enhancement of nighttime ocean thin Cirrus and low clouds detection. Monthly averaged global and zonal cloud fractions from improved NOAA 18 AVHRR cloud mask compare well with CERES-MODIS Edition 4 (CM Ed4), MODIS team, and CALIPSO cloud mask. Spatially and temporally matched cloud detection statistics between NOAA 18 AVHRR and CALIPSO cloud mask will also be presented.

Qing Z Trepte↗
DOE OSTI2021

VERIFICATION OF TRISO FUEL BURNUP USING MACHINE LEARNING ALGORITHMS

Pebble Bed Reactors are fueled with fuel pebbles that are circulated multiple times through the reactor vessel before discharge. During the normal operation of a PBR, ejected pebbles are returned to the reactor or discharged depending on the fuel burnup and physical condition of the pebbles. The burnup measurement is usually based on detected radiation signatures of fission products accumulated in the pebble fuel over burnup. Previous research has shown that height of photopeaks of fission products, such as 134Cs, 137Cs, 154Eu, etc., can be used independently or in combination to infer or predict the level of burnup in the fuel. However, it remains challenging to measure such complex sources due to self-shielding effects, strong radiation background and intervening materials. Another operational challenge is the required high throughput of burnup measurement, which necessitates limited measurement time and thus impacts quality of measured gamma-ray spectra. Hence, advanced spectral analysis methods are needed to analyze the noisy gamma spectra and predict the burnup values. We propose to use machine learning (ML) method to interpret gamma-ray spectra and predict the burnup values of the pebbles. ML has achieved widespread success and adoption across a few domains that require pattern recognition and analysis in varied data types. In this work, we apply three proven ML approaches - multilayer perceptrons, convolutional neural networks, and transformers - to the task of predicting fuel burnup from measured gamma spectra, and compile a dataset of simulated spectra for training and validation of the ML models. In this paper, we will discuss the network architecture of these three ML approaches and compare the performance of the simplest of these (MLP) to a standard linear regression.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗
NASA NTRS1988

Chaos in a spatially-developing plane mixing layer

A spatially-developing plane mixing layer was analyzed for chaotic behavior. A direct numerical simulation of the Navier-Stokes equations in a 2-D domain infinite in y and having inflow-outflow boundary conditions in x was used for data. Spectra, correlation dimension and the largest Lyapunov exponent were computed as functions of downstream distance x. When forced at a single (fundamental) frequency with maximum amplitude, the flow is periodic at the inflow but becomes aperiodic with increasing x. The aperiodic behavior is caused by the presence of a noisy subharmonic caused by the feedback between the necessarily nonphysical inflow and outflow boundary conditions. In order to overshadow this noise the flow was also studied with the same fundamental forcing and added random forcing of amplitude upsilon prime sub R/delta U = 0.01 at the inlet. Results were qualitatively the same in both cases: for small x, spectral peaks were sharp and dimension was nearly 1, but as x increased a narrowband spectral peak grew, spectra decayed exponentially at high frequencies and dimension increased to greater than 3. Based on these results, the flow appears to exhibit deterministic chaos. However, at no location was the largest Lyapunov exponent found to be significantly greater than zero.

Broze, J. G.↗
NASA NTRS2003

Extraction of Qualitative Features from Sensor Data Using Windowed Fourier Transform

In this paper, we use Matlab to model the health monitoring of a system through the information gathered from sensors. This implies assessment of the condition of the system components. Once a normal mode of operation is established any deviation from the normal behavior indicates a change. This change may be due to a malfunction of an element, a qualitative change, or a change due to a problem with another element in the network. For example, if one sensor indicates that the temperature in the tank has experienced a step change then a pressure sensor associated with the process in the tank should also experience a step change. The step up and step down as well as sensor disturbances are assumed to be exponential. An RC network is used to model the main process, which is step-up (charging), drift, and step-down (discharging). The sensor disturbances and spike are added while the system is in drift. The system is allowed to run for a period equal to three time constant of the main process before changes occur. Then each point of the signal is selected with a trailing data collected previously. Two trailing lengths of data are selected, one equal to two time constants of the main process and the other equal to two time constants of the sensor disturbance. Next, the DC is removed from each set of data and then the data are passed through a window followed by calculation of spectra for each set. In order to extract features the signal power, peak, and spectrum are plotted vs time. The results indicate distinct shapes corresponding to each process. The study is also carried out for a number of Gaussian distributed noisy cases.

Amini, Abolfazl M.↗
DOE OSTI2022

Performance Evaluations of Noisy Approximate Quantum Fourier Arithmetic

The Quantum Fourier Transform (QFT) grants competitive advantages, especially in resource usage and circuit approximation, for performing arithmetic operations on quantum computers, and offers a potential route towards a numerical quantum-computational paradigm. In this paper, we utilize efficient techniques to implement QFT-based integer addition and multiplications. These operations are fundamental to various quantum applications including Shor’s algorithm, weighted sum optimization problems in data processing and machine learning and quantum algorithms requiring inner products. We carry out performance evaluations of these implementations based on IBM’s superconducting qubit architecture using different compatible noise models. We isolate the sensitivity of the component quantum circuits on both one-/two-qubit gate error rates, and the number of the arithmetic operands’ superposed integer states. We analyze performance, and identify the most effective approximation depths for quantum add and quantum multiply within the given context. We observe significant dependency of the optimal approximation depth on the degree of machine noise and the number of superposed states in certain performance regimes. Finally, we elaborate on the algorithmic challenges - relevant to signed, unsigned, modular and non-modular versions - that could also be applied to current implementations of QFT-based subtraction, division, exponentiation, and their potential tensor extensions. Here, we analyze performance trends in our results and speculate on possible future development within this computational paradigm.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1990

Spacecraft Doppler tracking with a VLBI antenna

Preliminary results are reported from Doppler-shift measurements to the Voyager-2 spacecraft at a distance of 26 AU, obtained using the 32-m VLBI antenna at Medicina (Italy) during July and August 1988. The apparatus comprises the el-az antenna, an S-X-band receiver, a hydrogen maser to generate the reference signal, a Mark III VLBI terminal, and a digital tone extractor capable of isolating a tone of known frequency from a noisy signal and giving its phase and amplitude. A signal transmitted in S-band from the NASA Deep Space Network (DSN) station in Australia and retransmitted coherently in X-band by Voyager, was received 7 h 6 min later at Medicina and at the DSN station in Madrid. Sample data are presented graphically and shown to be of generally high quality; further in-depth analysis is under way.

Comoretto, G.↗
NASA NTRS1994

A Study of the Relationship Between Anthropogenic Sulfate and Cloud Drop Nucleation

The characteristics of the cloud drop size distribution near cloud base are initially determined by the aerosol particles that serve as CCN and by the local updraft velocity. Chemical reactions of the emitted gaseous sulfur compounds due to human activities will alter, through gas-to-particle conversion, the aerosol size distribution, total number, and its chemical composition. Recently, Boucher and Rodhe and Jones et.al have each developed parameterizations relating cloud drop concentration to sulfate mass or aerosol number concentration, respectively, and used them to develop estimates of the indirect forcing by anthropogenic sulfate aerosols. THese parameterizations made use of measure relationships in continental and maritime clouds. However, these relationships are inherently noisy, yielding more than a factor of 2 variation in cloud drop concentration for a given aerosol number (or for a given sulfate mass) concentration. The large spatial and temporal variabilities in the concentration, chemical characteristics, and size distribution of aerosols have made it difficult to develop such a parameterization from data. In this paper, our focus is to develop a means for relating the predicted anthropogenic sulfate mass to cloud drop number concentration over the range of expected conditions associated with continental and marine aerosol. We start with an assumed pre-existing particle size distribution and develop an approximation of the altered distribution after addition of anthropogenic sulfate. We thereby develop a conservative estimate of the possible change in cloud drop number concentration due to anthropogenic sulfate.

Chuang, Catherine C.↗
NASA NTRS2004

Airborne Spectral Measurements of Ocean Directional Reflectance

During summer of 2001 NASA's Cloud Absorption Radiometer (CAR) obtained measurement of ocean angular distribution of reflected radiation or BRDF (bidirectional reflectance distribution function) aboard the University of Washington Convair CV-580 research aircraft under cloud-free conditions. The measurements took place aver the Atlantic Ocean off the eastern seaboard of the U.S. in the vicinity of the Chesapeake Light Tower and at nearby National Oceanic and Atmospheric Administration (NOAA) Buoy Stations. The measurements were in support of CLAMS, Chesapeake Lighthouse and Aircraft Measurements for Satellites, field campaign that was primarily designed to validate and improve NASA's Earth Observing System (EOS) satellite data products being derived from three sensors: MODIS (MODerate Resolution Imaging Spectro-Radiometer), MISR (Multi-angle Imaging Spectro-Radiometer) and CERES (Clouds and Earth s Radiant Energy System). Because of the high resolution of the CAR measurements and its high sensitivity to detect weak ocean signals against a noisy background, results of radiance field above the ocean are seen in unprecedented detail. The study also attempts to validate the widely used Cox-Munk model for predicting reflectance from a rough ocean surface.

Gatebe, Charles K.↗
DOE OSTI2025

Performance evaluations of signed and unsigned noisy approximate quantum Fourier arithmetic

The Quantum Fourier Transform (QFT) grants competitive advantages, especially in resource usage and circuit approximation, for performing arithmetic operations on quantum computers, and offers a potential route toward a numerical quantum-computational paradigm. In this paper, we utilize efficient techniques to implement QFT-based integer addition and multiplications. These operations are fundamental to various quantum applications including Shor’s algorithm, weighted-sum optimization problems in data processing and machine learning, and quantum algorithms requiring inner products. We carry out performance evaluations of these implementations based on IBM’s superconducting-qubit architecture using different compatible noise models. We isolate the sensitivity of the component quantum circuits on both one-/two-qubit gate error rates, and the number of the arithmetic operands’ superposed integer states. We analyze performance and identify the most effective approximation depths for unsigned quantum addition and quantum multiplication within the given context. We then perform a similar analysis of signed addition and compare to the unsigned results. We observe significant dependency of the optimal approximation depth on the degree of machine noise and the number of superposed states in certain performance regimes. Finally, we elaborate on the algorithmic challenges—relevant to signed, unsigned, modular and non-modular versions—that could also be applied to current implementations of QFT-based subtraction, division, exponentiation, and their potential tensor extensions. Here, we analyze the performance trends in our results and speculate on possible future developments within this computational paradigm.

Computational models↗
NASA NTRS2016

Identification of Fixations in Noisy Eye Movements via Recursive Subdivision

When solving problems, multi-person airline crews can choose whether to work together, or to address different aspects of a situation with a divide and conquer strategy. Knowing which of these strategies is most effective may help airlines develop better procedures and training. This paper concentrates on joint attention as a measure of crew coordination. We report results obtained by applying cross recurrence analysis to eye movement data from two-person crews, collected in a flight simulator experiment. The analysis shows that crews exhibit coordinated gaze roughly one sixth of the time, with a tendency for the captain to lead the first officers visual attention. The degree to which crews coordinate their gaze is not significantly correlated with performance ratings assigned by instructors; further research questions and approaches are discussed.

signal processing↗
DOE OSTI2021

High-Resolution Imaging of Unstained Polymer Materials

Electron microscopy has played an important role in polymer characterization. Traditionally, electron diffraction is used to study crystalline polymers while transmission electron microscopy is used to study microphase separation in stained block copolymers and other multiphase systems. We describe developments that eliminate the barrier between these two approaches - it is now possible to image polymer crystals with atomic resolution. The focus of this Review is on high-resolution imaging (30 Å and smaller) of unstained polymers. Recent advances in hardware allow for capturing numerous (as many as 105) low-dose images from an unperturbed specimen; beam damage is a significant barrier to high-resolution electron microscopy of polymers. Machine-learning-based software is then used to sort and average the images to retrieve pristine structural information from a collection of noisy images. Acknowledging the heterogeneity in polymer samples prior to averaging is essential. Molecular conformations in a wide range of amphiphilic block copolymers, polymerized ionic liquids, and conjugated polymers can be gleaned from two-dimensional projections (2D), three-dimensional (3D) tomograms, and four-dimensional (4D) scanning transmission electron microscopy (STEM) data sets where 2D diffraction patterns are taken as a function of position. Some methods such as phase contrast STEM have been used to image closely related materials such as metal-organic frameworks but not polymers. With improvements in hardware and software, such methods may soon be applied to polymers. Our goal is to provide a comprehensive understanding of the strategies toward the high-resolution imaging of radiation sensitive polymer materials at different length scales.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Regularization via f -Divergence: An Application to Multi-Oxide Spectroscopic Analysis

In this paper, we explore the application of convolutional neural networks (CNNs) for predicting the chemical composition of complex geologic samples in a simulated Martian atmospheric environment. Specifically, we aim to characterize oxide weight percentages (wt.%) of rock samples analyzed by remote Laser-Induced Breakdown Spectroscopy (LIBS), framing the problem as a multi-target regression task . Neural networks trained on LIBS spectra are prone to overfitting due to high spectral complexity, limited labeled data, and measurement noise. While regularization is critical for improving generalization, common methods (e.g., ℓ 2 regularization) impose constraints not directly tied to data distribution properties. We propose a novel regularization method based on a specific ƒ-divergence induced by a graph-based estimator, designed to constrain the distributional discrepancy between predictions and targets. This regularizer serves a dual purpose: (a) mitigating overfitting by enforcing a constraint on the distributional difference between predictions and noisy targets, and (b) acting as an auxiliary loss that penalizes large divergences. To enable backpropagation, we develop a differentiable approximation of this particular ƒ-divergence, making the method feasible for neural networks. Experiments on ChemCam and SuperCam LIBS calibration spectra show that mathematical equation-divergence regularization outperforms or matches standard regularization methods (ℓ 1 , ℓ 2 , dropout) and the classical baseline, partial least squares (PLS). Combining ƒ-divergence regularization with standard regularization yields further performance gains, indicating that distributional regularization is useful in this context giving a promising direction for robust model training in planetary science applications. Source code is publicly available at Klein and Li (2025), https://doi.org/10.11578/dc.20250530.7.

58 GEOSCIENCES↗
DOE OSTI2023

BigNeuron: a resource to benchmark and predict performance of algorithms for automated tracing of neurons in light microscopy datasets

BigNeuron is an open community bench-testing platform with the goal of setting open standards for accurate and fast automatic neuron tracing. We gathered a diverse set of image volumes across several species that is representative of the data obtained in many neuroscience laboratories interested in neuron tracing. Here, we report generated gold standard manual annotations for a subset of the available imaging datasets and quantified tracing quality for 35 automatic tracing algorithms. The goal of generating such a hand-curated diverse dataset is to advance the development of tracing algorithms and enable generalizable benchmarking. Together with image quality features, we pooled the data in an interactive web application that enables users and developers to perform principal component analysis, t-distributed stochastic neighbor embedding, correlation and clustering, visualization of imaging and tracing data, and benchmarking of automatic tracing algorithms in user-defined data subsets. The image quality metrics explain most of the variance in the data, followed by neuromorphological features related to neuron size. Furthermore, we observed that diverse algorithms can provide complementary information to obtain accurate results and developed a method to iteratively combine methods and generate consensus reconstructions. The consensus trees obtained provide estimates of the neuron structure ground truth that typically outperform single algorithms in noisy datasets. However, specific algorithms may outperform the consensus tree strategy in specific imaging conditions. Finally, to aid users in predicting the most accurate automatic tracing results without manual annotations for comparison, we used support vector machine regression to predict reconstruction quality given an image volume and a set of automatic tracings.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1977

TDRSS performance evaluation for TDRSS services using analytical simulation

An analytical simulation package which can predict bit error rate and tracking performance for nonlinear, bandlimited satellite links was developed for the Tracking and Data Relay Satellite System (TDRSS). It is based on analytical simulation, i.e., a combination of simulation techniques and analytical performance evaluation. This allows the use of a more realistic signal format than is possible with a purely analytical approach and maintains the advantage of short computation time on bandlimited, nonlinear channels with a noisy uplink. The link models and the main features of the program are described and an overview of the mathematical approach is given. Sample results of computer runs are provided for current power budgets and system design parameters for both forward and return links.

Braun, W. R.↗
NASA NTRS1987

Photon noise and constant-volume operators

In a previous paper, a class of constant volume (CV) operators designed to maximize spatial resolution in the presence of photon noise was introduced. In this paper, the statistical properties are derived of the output images created by CV operators applied to photon-noisy versions of input images consisting of edges, spots, and gratings. The results are used to calculate the input image parameters that would permit any test image to be discriminated from a uniform field with a given level of reliability and consequently to compare the predictions of a CV operator model with psychophysical data. These results also allow the potential usefulness of CV operators in artificial image processing applications to be assessed.

Yellott, John I., Jr.↗
NASA NTRS2008

Estimating Mixing Heights Using Microwave Temperature Profiler

A paper describes the Microwave Temperature Profiler (MTP) for making measurements of the planetary boundary layer thermal structure data necessary for air quality forecasting as the Mixing Layer (ML) height determines the volume in which daytime pollution is primarily concentrated. This is the first time that an airborne temperature profiler has been used to measure the mixing layer height. Normally, this is done using a radar wind profiler, which is both noisy and large. The MTP was deployed during the Texas 2000 Air Quality Study (TexAQS-2000). An objective technique was developed and tested for estimating the ML height from the MTP vertical temperature profiles. In order to calibrate the technique and evaluate the usefulness of this approach, estimates from a variety of measurements during the TexAQS-2000 were compared. Estimates of ML height were used from radiosondes, radar wind profilers, an aerosol backscatter lidar, and in-situ aircraft measurements in addition to those from the MTP.

Nielson-Gammon, John↗
DOE OSTI2026

Graph Identification of Proteins in Tomograms (GRIP-Tomo) 2.0: Topologically aware classification for proteins

Cryo-electron tomography (cryo-ET) enables structural characterization of biomolecules under near-native conditions. Existing approaches for interpreting the resulting three-dimensional volumes are computationally expensive and have difficulty interpreting density associated with small proteins/complexes. To explore alternate approaches for identifying proteins in cryo-ET data we pursued a Graph Network and topologically invariant approach. Here, we report on a fast algorithm that classifies particles by searching for nuances of evolutionarily conversed motifs and the geometrical characteristics of protein structure. GRIP-Tomo 2.0 is a machine-learning pipeline that extracts interpretable topological features of protein structures within noisy experimental backgrounds. Compared to version 1.0, the new pipeline includes three upgrades that significantly improve performance including synthetic tomogram generation simulating realistic noise, graph-based persistent feature extraction as protein fingerprints, and high-performance computing acceleration. GRIP-Tomo 2.0 achieves over 90% accuracy in classifying between proteins and noise using both real and synthetic datasets which represents a foundational step toward advancing cryo-ET workflows and empowering automated visual proteomics.

Li, Chengxuan↗