SEARCH · Search NASA

Results for “regression testing”

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 505 records · Page 28

All sources NASA NTRS DOE OSTI

NASA NTRS1992

Cortical localization of cognitive function by regression of performance on event-related potentials

This paper demonstrates a new method of mapping cortical localization of cognitive function, using electroencephalographic data. Cross-subject regression analyses are used to identify cortical sites and post-stimulus latencies where there is a high correlation between subjects' performance and their cognitive event-related potential amplitude. The procedure was tested using a mental arithmetic task and was found to identify essentially the same cortical regions that have been associated with such tasks on the basis of research with patients suffering localized cortical lesions. Thus, it appears to offer an inexpensive, noninvasive tool for exploring the dynamics of localization in neurologically normal subjects.

Montgomery, R. W.↗

DOE OSTI2021

AI-Batt (Autonomous Identification of Battery Life Models) [SWR 21-36]

Autonomous Identification of Battery Life Models (AI-Batt) AI-Batt is a MATLAB code base for developing lifetime models for batteries from accelerated aging data. The code base provides many functions for processing, visualizing, and modeling battery aging data, making the data processing, exploration, and modeling workflow substantially faster. These tools are tailored for working with battery aging data sets, which usually consist of many separate time-series for each cell, with many test conditions and possible replicates at each condition, which makes it difficult to simply process or visualize the data set. Complex modeling tasks, such as cross-validation, sensitivity analysis, and uncertainty quantification have been implemented to enable thorough statistical investigation of model predictions. Additionally, several machine-learning algorithms are implemented to autonomously identify suitable models via symbolic regression. Data processing functions automatically cast data from the struct data type, which is commonly used to store experimental data, but is not an acceptable input for most algorithms, to the table data type, which can be easily used as input to any optimization algorithm. Also, the data can be separated into time-invariant and time-variant data tables, which is helpful for exploring the data set as well as developing separate models for time-variant and time-invariant aging mechanisms. For example, in aging tests with constant temperature, temperature is a time-invariant experimental condition. Visualization tools enable plotting of data, model fits, and model simulations possible with single-line function calls, empowering data exploration of complex data sets with both time-varying and time-invariant trends. Plots can be automatically generated for the whole data set, or separated by data group (groups of test replicates) or individual data series. Data points or data series can be automatically colored by the value of a variable with a variety of color maps, and model predictions can also be colored by the value of a fit statistic. Comparisons between data sets and the predictions/simulations of different models on the same data set can be easily plotted as well. Distributions of parameter values from bootstrap resampling can be plotted to visualize the reliability of parameter estimation, or determine any correlations between parameters. Modeling tools handle the complex task of creating and parsing symbolic equations for modeling battery lifetime. Equations are parsed to grab relevant data variables, parameter values, or specified sub-models for input into optimization, evaluation, or simulation functions. Models can be optimized locally (one set of parameters for each data series), bi-level (some parameters shared across the data set), or globally (single set of parameters for all data). Functions implementing symbolic regression algorithms help users to discover effective model equations, even in poorly sampled, high-dimensional data.

Smith, Kandler [National Renewable Energy Lab. (NR↗

DOE OSTI2024

Impurity gas detection for SNF canisters using probabilistic deep learning and acoustic sensing *

Abstract Monitoring impurity gases in spent nuclear fuel (SNF) canisters is a novel structural health monitoring approach for SNF in dry storage. The SNF canisters are sealed containers that do not facilitate visual access to the inside. Acoustic sensing can be deployed by taking advantage of the pathways unobstructed by internal hardware. Although the ultrasonic time-of-flight measurement can provide valuable information, it is limited in its ability to discern the concentration of only one impurity gas. As such, deep learning algorithms, particularly convolutional neural networks (CNNs), offer a promising solution. In this study, CNN-based probabilistic deep learning models were implemented to detect and quantify multiple impurity gases in helium. An experimental platform was established to simulate canister conditions, and ultrasonic test data were collected. The presence of argon and air in helium at concentrations ranging from 0% to 1.2% at increments of 0.05% was considered. The multi-layer perceptron, decision tree, and logistic regression classifiers achieved high accuracies when distinguishing pure helium from helium with impurities. CNN with dropout layers and CNN using maximum likelihood estimation showed a similar performance, indicating their ability to capture uncertainties. The ensemble CNN model exhibited improved predictions and the ability to balance individual gas concentration by integrating 1D- and 2D-CNN models. These findings contribute probabilistic deep learning solutions for impurity gas detection and analysis within SNF canisters, thus ensuring safe storage and management of SNFs.

47 OTHER INSTRUMENTATION↗

NASA NTRS1985

Pattern recognition in the satellite temperature retrieval problem

Pattern recognition procedures have been developed in order to improve the first-guess fields for satellite temperature retrievals. The first procedure is used to select one or more historical radiosonde temperature profiles as analog estimates of ambient thermal structure. The second procedure is used to organize a priori data into shape-coherent pattern libraries using structural information inherent in the data itself. On the basis of independent tests of about 800 temperature retrievals, it was found that: (1) the pattern recognition techniques reduced first-guess profile errors by nearly 50 percent in comparison with traditional partitioning schemes; and (2) with regression and physical-iterative retrieval algorithms, however, the effect of pattern recognition on temperature retrieval error was insignificant. Analysis of individual retrieval errors showed that poor retrievals may outweigh the potential benefits of both pattern recognition techniques.

Thompson, O. E.↗

NASA NTRS1986

Acousto-ultrasonic verification of the strength of filament wound composite material

The concept of acousto-ultrasonic (AU) waveform partitioning was applied to nondestructive evaluation of mechanical properties in filament wound composites (FWC). A series of FWC test specimens were subjected to AU analysis and the results were compared with destructively measured interlaminar shear strengths (ISS). AU stress-wave factor (SWF) measurements gave greater than 90 percent correlation coefficient upon regression against the ISS. This high correlation was achieved by employing the appropriate time and frequency domain partitioning as dictated by wave propagation path analysis. There is indication that different SWF frequency partitions are sensitive to ISS at different depths below the surface.

Kautz, H. E.↗

NASA NTRS1991

Experimental investigation of a solid rocket combustion simulator

The response of solid rocket motor materials to high-temperature corrosive gases is usually accomplished by testing the materials in a subscale solid rocket motor. While this imposes the proper thermal and chemical environment, a solid rocket motor does not provide practical features that would enhance systematic evaluations such as: the ability to throttle for margin testing, on/off capability, low test cost, and a low-hazards test article. Solid Rocket Combustion Simulators (SRCS) are being evaluated by NASA to test solid rocket nozzle materials and incorporate these essential practical features into the testing of rocket materials. The SRCS is designed to generate the thermochemical environment of a solid rocket. It uses hybrid rocket motor technology in which gaseous oxygen (Gox) is injected into a chamber containing a solid fuel grain. Specific chemicals are injected in the aft mixing chamber so that the gases entering the test section match the temperature and a non-dimensional erosion factor B' to insure similarity with a solid motor. Because the oxygen flow can be controlled, this approach allows margin testing, the ability to throttle, and an on/off capability. The fuel grains are inert which makes the test article very safe to handle. The objective of this work was to establish the baseline operating characteristics of a Labscale Solid Rocket Combustion Simulator (LSRCS). This included establishing the baseline burning rates of plexiglass fuels and the evaluation of a combustion instability for hydroxy-terminated polybutadyene (HTPB) propellants. The scope of the project included: (1) activation of MSFC Labscale Hybrid Combustion Simulator; (2) testing of plexiglass fuel at Gox ranges from 0.025 to 0.200 lb/s; (3) burning HTPB fuels at a Gox rate of 0.200 lb/s using four different mixing chamber configurations; and (4) evaluating the fuel regression and chamber pressure responses of each firing.

Frederick, Robert A., Jr.↗

NASA NTRS2006

Excitations for Rapidly Estimating Flight-Control Parameters

A flight test on an F-15 airplane was performed to evaluate the utility of prescribed simultaneous independent surface excitations (PreSISE) for real-time estimation of flight-control parameters, including stability and control derivatives. The ability to extract these derivatives in nearly real time is needed to support flight demonstration of intelligent flight-control system (IFCS) concepts under development at NASA, in academia, and in industry. Traditionally, flight maneuvers have been designed and executed to obtain estimates of stability and control derivatives by use of a post-flight analysis technique. For an IFCS, it is required to be able to modify control laws in real time for an aircraft that has been damaged in flight (because of combat, weather, or a system failure). The flight test included PreSISE maneuvers, during which all desired control surfaces are excited simultaneously, but at different frequencies, resulting in aircraft motions about all coordinate axes. The objectives of the test were to obtain data for post-flight analysis and to perform the analysis to determine: 1) The accuracy of derivatives estimated by use of PreSISE, 2) The required durations of PreSISE inputs, and 3) The minimum required magnitudes of PreSISE inputs. The PreSISE inputs in the flight test consisted of stacked sine-wave excitations at various frequencies, including symmetric and differential excitations of canard and stabilator control surfaces and excitations of aileron and rudder control surfaces of a highly modified F-15 airplane. Small, medium, and large excitations were tested in 15-second maneuvers at subsonic, transonic, and supersonic speeds. Typical excitations are shown in Figure 1. Flight-test data were analyzed by use of pEst, which is an industry-standard output-error technique developed by Dryden Flight Research Center. Data were also analyzed by use of Fourier-transform regression (FTR), which was developed for onboard, real-time estimation of the derivatives.

Moes, Tim↗

NASA NTRS2022

Predicting the Penetration of a Shielded TPS Tile

All spacecraft are subject to the possibility of high-speed particle impacts during their mission life. In low earth orbit, those impacts could be the result of collisions with pieces of orbital debris or with meteorites. Beyond LEO, and especially beyond GEO, those impacts will likely be caused by meteorites. Such high-speed impacts on spacecraft surfaces create debris clouds that travel towards and eventually impact other downstream spacecraft components. In addition to the impulsive load that such debris clouds would impart to the spacecraft elements with which they subsequently collide, the largest fragment in these debris clouds poses a significant threat on its own to those spacecraft elements. In order to be able to assess the severity of the threat posed by such a fragment, it is important to be able to predict the extent of damage sustained by the impacted spacecraft element. In this paper, we present a new process for determining the penetration depth in a shielded TPS tile system. The requirements for the process were that it be applicable across a full spectrum of impact low velocities (i.e. from ~ 7 km/s to above 50 km/s), that it be applicable over a wide range of projectile materials densities (i.e. ranging from water to steel, for example), and that it be applicable over a wide range of trajectory obliquities (i.e. not just normal impacts). Furthermore, the process developed should be sufficiently flexible so that it can be used for shield design parameter and / or system configuration trade studies (i.e. be written in terms of as many shield and TPS tile system parameters as possible). Following the initial impact of the projectile on the outer wall of the dual-wall shield, a debris cloud (i.e. the primary debris cloud) is created that travels towards and impacts the inner wall of the shield. Within this debris cloud is a combination of solid, liquid, and vaporized material, depending on the impact velocity and the impedance mismatch of the projectile and outer wall materials. The impact of this debris cloud on the inner wall creates another debris cloud (i.e. the secondary debris cloud) that then travels towards and impacts the TPS tile. Within the particulate distribution of the material in each of these debris clouds is a so-called largest fragment. The process used to calculate the penetration depth in a shielded TPS tile system developed herein consists of the following three steps: 1. Does the initial projectile perforate the outer wall of the shield? If NO, the TPS tile remains undamaged. If YES, calculate the size and speed of the largest fragment exiting the outer wall of the shield and traveling towards the shield’s inner wall. 2. Does the largest fragment in the debris cloud exiting the outer wall of the shield perforate the inner wall of the shield? If NO, the TPS tile again remains undamaged. If YES, calculate the size and speed of the largest fragment exiting the rear of the shield inner wall. 3. Calculate the penetration depth of the largest fragment in the debris cloud exiting the inner wall into the TPS. Perforation of the inner and outer shield walls is predicted using the Cour-Palais penetration depth equation and an associated failure criterion [1]. The largest debris cloud fragment diameters and associated velocities are calculated using regression equations developed from a mix of hydrocode and experimental diameter and velocity information [2]. Finally, TPS penetration depths are calculated using an empirical penetration depth predictor equation based on high-speed impact test data and are calibrated using a limited number of hydrocode penetration depth predictions for water, nylon, aluminum, Dunite, and steel projectiles The predictions of the penetration depth calculation process are compared against the predictions of more than 60 SPHC hydrocode runs. These comparisons show that the process used to calculate TPS tile penetration depths usually yields values that are within 1 mm of the predictions of hydrocode simulations. If a maximum allowable TPS penetration depth is known, the process developed herein can now be used to develop a failure limit equation that would predict whether or not an impacting particle would penetrate deeply enough into the TPS to result in a “failed” or “not failed” end state. References [1] E.L. Christiansen, “Design and performance equations for advanced meteoroid and debris shields”, International Journal of Impact Engineering, Vol. 14, 1993, pp. 145-156. [2] W.P. Schonberg, “Predicting the size of the largest particle fragment in a debris cloud created by an orbital debris impact and its associated velocity”, 2021 Applied Space Environments Conference, NASA Jet Propulsion Laboratory, Pasadena, California, November, 2021.

Meteroid↗

NASA NTRS2022

MLtool: Universal Supervised Machine Learning Tool to Model Tabulated Data

Machine Learning (ML) is a subfield of Artificial Intelligence that gives computers the ability to learn from past data without being explicitly programmed. The predictive capabilities of ML models have already been used to facilitate several scientific breakthroughs. However, the practical application of ML is often limited due to the gaps in technical knowledge of its users. The common issue faced by many scientific researchers is the inability to choose the appropriate ML pipelines that are needed to treat real-world data, which is often sparse and noisy. To solve this problem, we have developed an automated Machine Learning tool (MLtool) that includes a set of ML algorithms and approaches to aid scientific researchers. The current version of MLtool is implemented as an object-oriented Python code that is easily extensible. It includes 44 different regression algorithms used to model data. MLtool helps users select the best model for their data, based on the scoring metrics used. Besides regression algorithms, MLtool also includes a suite of pre- and post-processing techniques such as missing value imputation, categorical variable encoding, input feature normalization, uncertainty quantification, exploratory data analysis (EDA), etc. MLtool was tested on several publicly available multi-dimensional data sets and was found capable of making accurate predictions.

Machine learning↗

NASA NTRS2022

MLtool Python Code

Machine Learning↗

NASA NTRS1997

Modeling the effects of exercise during 100% oxygen prebreathe on the risk of hypobaric decompression sickness

BACKGROUND: Several previous studies indicated that exercise during prebreathe with 100% O2 decreased the incidence of hypobaric decompression sickness (DCS). We report a meta-analysis of these investigations combined with a new study in our laboratory to develop a statistical model as a predictive tool for DCS. HYPOTHESIS: Exercise during prebreathe increases N2 elimination in a theoretical 360-min half-time compartment decreasing the incidence of DCS. METHODS: A dose-response probability tissue ratio (TR) model with 95% confidence limits was created for two groups, prebreathe with exercise (n = 113) and resting prebreathe (n = 113), using nonlinear regression analysis with maximum likelihood optimization. RESULTS: The model predicted that prebreathe exercise would reduce the residual N2 in a 360-min half-time compartment to a level analogous to that in a 180-min compartment. This finding supported the hypothesis. The incidence of DCS for the exercise prebreathe group was significantly decreased (Chi-Square = 17.1, p < 0.0001) from the resting prebreathe group. CONCLUSIONS: The results suggested that exercise during prebreathe increases tissue perfusion and N2 elimination approximately 2-fold and markedly lowers the risk of DCS. Based on the model, the prebreathe duration may be reduced from 240 min to a predicted 91 min for the protocol in our study, but this remains to be verified. The model provides a useful planning tool to develop and test appropriate prebreathe exercise protocols and to predict DCS risks for astronauts.

NASA Program Environmental Health↗

NASA NTRS2005

Regression Analysis of Long-Term Profile Ozone Data Set from BUV Instruments

We have produced a profile merged ozone data set (MOD) based on the SBUV/SBUV2 series of nadir-viewing satellite backscatter instruments, covering the period from November 1978 - December 2003. In 2004, data from the Nimbus 7 SBUV and NOAA 9, ll, and 16 SBUV/2 instruments were reprocessed using the Version 8 (V8) algorithm and most recent calibrations. More recently, data from the Nimbus 4 BUT instrument, which was operational from 1970 - 1977, were also reprocessed using the V8 algorithm. As part of the V8 profile calibration, the Nimbus 7 and NOAA 9 (1993-1997 only) instrument calibrations have been adjusted to match the NOAA 11 calibration, which was established based on comparisons with SSBUV shuttle flight data. Differences between NOAA 11, Nimbus 7 and NOAA 9 profile zonal means are within plus or minus 5% at all levels when averaged over the respective periods of data overlap. NOAA 16 SBUV/2 data have insufficient overlap with NOAA 11, so its calibration is based on pre-flight information. Mean differences over 4 months of overlap are within plus or minus 7%. Given the level of agreement between the data sets, we simply average the ozone values during periods of instrument overlap to produce the MOD profile data set. Initial comparisons of coincident matches of N4 BUV and Arosa Umkehr data show mean differences of 0.5 (0.5)% at 30km; 7.5 (0.5)% at 35 km; and 11 (0.7)% at 40 km, where the number in parentheses is the standard error of the mean. In this study, we use the MOD profile data set (1978-2003) to estimate the change in profile ozone due to changing stratospheric chlorine levels. We use a standard linear regression model with proxies for the seasonal cycle, solar cycle, QBO, and ozone trend. To account for the non-linearity of stratospheric chlorine levels since the late 1990s, we use a time series of Effective Chlorine, defined as the global average of Chlorine + 50 * Bromine at 1 hPa, as the trend proxy. The Effective Chlorine data are taken from the 3-D Goddard CTM. We will show the latest trend results using this statistical model. In addition, the Nimbus 4 BUV data offer an opportunity to test the physical properties of our statistical model. From ground-based comparisons we will establish an uncertainty range for the Nimbus 4 data. We then extrapolate our statistical model fit backwards in time and compare to the Nimbus 4 data. We compare the characteristics of the residual, defined as the difference between the data and statistical regression fit, during the Nimbus 4 time period and the 1978-2003 period over which the statistical model coefficients were estimated, and present these results.

Stolarski, Richard S.↗

NASA NTRS2004

Analysis of Wind Tunnel Lateral Oscillatory Data of the F-16XL Aircraft

Static and dynamic wind tunnel tests were performed on an 18% scale model of the F-16XL aircraft. These tests were performed over a wide range of angles of attack and sideslip with oscillation amplitudes from 5 deg. to 30 deg. and reduced frequencies from 0.073 to 0.269. Harmonic analysis was used to estimate Fourier coefficients and in-phase and out-of-phase components. For frequency dependent data from rolling oscillations, a two-step regression method was used to obtain unsteady models (indicial functions), and derivatives due to sideslip angle, roll rate and yaw rate from in-phase and out-of-phase components. Frequency dependence was found for angles of attack between 20 deg. and 50 deg. Reduced values of coefficient of determination and increased values of fit error were found for angles of attack between 35 deg. and 45 deg. An attempt to estimate model parameters from yaw oscillations failed, probably due to the low number of test cases at different frequencies.

Klein, Vladislav↗

NASA NTRS2003

Age affects severity of venous gas emboli on decompression from 14.7 to 4.3 psia

INTRODUCTION: Variables that define who we are, such as age, weight and fitness level influence the risk of decompression sickness (DCS) and venous gas emboli (VGE) from diving and aviation decompressions. We focus on age since astronauts that perform space walks are approximately 10 yr older than our test subjects. Our null hypothesis is that age is not statistically associated with the VGE outcomes from decompression to 4.3 psia. METHODS: Our data are from 7 different NASA tests where 188 men and 50 women performed light exercise at 4.3 psia for planned exposures no less than 4 h. Prebreathe (PB) time on 100% oxygen ranged from 150-270 min, including ascent time, with exercise of different intensity and length being performed during the PB in four of the seven tests with 150 min of PB. Subjects were monitored for VGE in the pulmonary artery using a Doppler ultrasound bubble detector for a 4-min period every 12 min. There were six design variables; the presence or absence of lower body adynamia and five PB variables; plus five concomitant variables on physical characteristics: age, weight height, body mass index, and gender that were available for logistic regression (LR). We used LR models for the probability of DCS and VGE, and multinomial logit (ML) models for the probability of Spencer VGE Grades 0-IV at exposure times of 61, 95, 131, 183 min, and for the entire exposure. RESULTS: Age was significantly associated with VGE in both the LR and ML models, so we reject the null hypothesis. Lower body adynamia was significant for all responses. CONCLUSIONS: Our selection of tests produced a wide range of the explanatory variables, but only age, lower body adynamia, height, and total PB time was helpful in various combinations to model the probability of DCS and VGE.

Clinical Trial↗

NASA NTRS1975

A digital algorithm for characteristic film curves

The task of establishing a film calibration scheme for magnitude studies of Skylab photographic images of Comet Kohoutek is examined. Since the data are recorded in terms of film density and have to be used in terms of exposure, the conversion from density to exposure is critical. In this film calibration scheme, the hardware deals with the data sources, recording medium, and data conversion to a computer compatible program, whereas the software deals with signal to noise enhancement, stepwedge calibration curve and leads to modeling of the film characteristic curves. A mathematical model of the characteristic curve is obtained using a modified version of Efroymson's (1960) stepwise multiple linear regression algorithm, which gives log exposure as a function of density. The difference in the calibration curves from pre- and postflight exposures is well accounted for in the model as a result of sensitive statistical tests. The characteristic curve modeling program requires about 4K of core and is executed in about 3 min.

Buckner, J.↗

NASA NTRS1978

Dynamics of cell degradation

The use of chemical and physical data as a supplement to linear regression models in the prediction of cell failure is discussed. Principal factors to be considered are the positive thickness and weight, and the negative thickness. A model for cell degradation and failure in accelerated life test cells is presented and predictions based on a teardown analysis are included.

Mcdermott, P. P.↗

NASA NTRS1988

A nonparametric software reliability growth model

Miller and Sofer have presented a nonparametric method for estimating the failure rate of a software program. The method is based on the complete monotonicity property of the failure rate function, and uses a regression approach to obtain estimates of the current software failure rate. This completely monotone software model is extended. It is shown how it can also provide long-range predictions of future reliability growth. Preliminary testing indicates that the method is competitive with parametric approaches, while being more robust.

Miller, Douglas R.↗

NASA NTRS2000

An Evaluation Technique for an F/A-18 Aircraft Loads Model Using F/A-18 Systems Research Aircraft Flight Data

A limited evaluation of the F/A-18 baseline loads model was performed on the Systems Research Aircraft at NASA Dryden Flight Research Center (Edwards, California). Boeing developed the F/A-18 loads model using a linear aeroelastic analysis in conjunction with a flight simulator to determine loads at discrete locations on the aircraft. This experiment was designed so that analysis of doublets could be used to establish aircraft aerodynamic and loads response at 20 flight conditions. Instrumentation on the right outboard leading edge flap, left aileron, and left stabilator measured the hinge moment so that comparisons could be made between in-flight-measured hinge moments and loads model-predicted values at these locations. Comparisons showed that the difference between the loads model-predicted and in-flight-measured hinge moments was up to 130 percent of the flight limit load. A stepwise regression technique was used to determine new loads derivatives. These derivatives were placed in the loads model, which reduced the error to within 10 percent of the flight limit load. This paper discusses the flight test methodology, a process for determining loads coefficients, and the direct comparisons of predicted and measured hinge moments and loads coefficients.

Olney, Candida D.↗