Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “regression models”

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 541 records · Page 30
All sourcesNASA NTRSDOE OSTI
NASA NTRS1974

Nonlinear models for estimating GSFC travel requirements

A methodology is presented for estimating travel requirements for a particular period of time. Travel models were generated using nonlinear regression analysis techniques on a data base of FY-72 and FY-73 information from 79 GSFC projects. Although the subject matter relates to GSFX activities, the type of analysis used and the manner of selecting the relevant variables would be of interest to other NASA centers, government agencies, private corporations and, in general, any organization with a significant travel budget. Models were developed for each of six types of activity: flight projects (in-house and out-of-house), experiments on non-GSFC projects, international projects, ART/SRT, data analysis, advanced studies, tracking and data, and indirects.

Buffalano, C.↗
NASA NTRS1981

Evaluation of the CEAS model for barley yields in North Dakota and Minnesota

The CEAS yield model is based upon multiple regression analysis at the CRD and state levels. For the historical time series, yield is regressed on a set of variables derived from monthly mean temperature and monthly precipitation. Technological trend is represented by piecewise linear and/or quadriatic functions of year. Indicators of yield reliability obtained from a ten-year bootstrap test (1970-79) demonstrated that biases are small and performance as indicated by the root mean square errors are acceptable for intended application, however, model response for individual years particularly unusual years, is not very reliable and shows some large errors. The model is objective, adequate, timely, simple and not costly. It considers scientific knowledge on a broad scale but not in detail, and does not provide a good current measure of modeled yield reliability.

Barnett, T. L.↗
NASA NTRS2007

A Comparison of Two Balance Calibration Model Building Methods

Simulated strain-gage balance calibration data is used to compare the accuracy of two balance calibration model building methods for different noise environments and calibration experiment designs. The first building method obtains a math model for the analysis of balance calibration data after applying a candidate math model search algorithm to the calibration data set. The second building method uses stepwise regression analysis in order to construct a model for the analysis. Four balance calibration data sets were simulated in order to compare the accuracy of the two math model building methods. The simulated data sets were prepared using the traditional One Factor At a Time (OFAT) technique and the Modern Design of Experiments (MDOE) approach. Random and systematic errors were introduced in the simulated calibration data sets in order to study their influence on the math model building methods. Residuals of the fitted calibration responses and other statistical metrics were compared in order to evaluate the calibration models developed with different combinations of noise environment, experiment design, and model building method. Overall, predicted math models and residuals of both math model building methods show very good agreement. Significant differences in model quality were attributable to noise environment, experiment design, and their interaction. Generally, the addition of systematic error significantly degraded the quality of calibration models developed from OFAT data by either method, but MDOE experiment designs were more robust with respect to the introduction of a systematic component of the unexplained variance.

DeLoach, Richard↗
NASA NTRS1973

An investigation to improve the Menhaden fishery prediction and detection model through the application of ERTS-A data

The author has identified the following significant results. Linear regression of secchi disc visibility against number of sets yielded significant results in a number of instances. The variability seen in the slope of the regression lines is due to the nonuniformity of sample size. The longer the period sampled, the larger the total number of attempts. Further, there is no reason to expect either the influence of transparency or of other variables to remain constant throughout the season. However, the fact that the data for the entire season, variable as it is, was significant at the 5% level, suggests its potential utility for predictive modeling. Thus, this regression equation will be considered representative and will be utilized for the first numerical model. Secchi disc visibility was also regressed against number of sets for the three day period September 27-September 29, 1972 to determine if surface truth data supported the intense relationship between ERTS-1 identified turbidity and fishing effort previously discussed. A very negative correlation was found. These relationship lend additional credence to the hypothesis that ERTS imagery, when utilized as a source of visibility (turbidity) data, may be useful as a predictive tool.

Maughan, P. M.↗
NASA NTRSDate not supplied

Application of Artificial Intelligence for Surface PM2.5 Estimations from Geostationary Satellite and Atmospheric Numerical Model Data

PM2.5, particulate matter (PM) with a diameter less than or equal to 2.5 μm, is emitted from anthropogenic fuel combustion and forest fires. Due to their small size, PM2.5 can penetrate into respiratory systems and cause or exacerbate serious illness. The US Environmental Protection Agency (EPA) regulates the levels of surface PM2.5 but surface monitoring has spatial and temporal limitations. The Aerosol Optical Depth (AOD) retrievals from the Geostationary Operational Environmental Satellite (GOES) missions and meteorological factors can be utilized as an alternative technique to estimate surface PM2.5 levels at a higher spatial and temporal resolution compared to surface monitors. Traditional estimation approaches rely on linear regression techniques and have limitations modeling the nonlinear relationship between the meteorological factors, AOD retrievals, and surface PM2.5. We compare different machine learning techniques and identify the best-suited model that can represent the nonlinearity between the factors affecting PM2.5 levels

Manisha Khatri↗
NASA NTRS2023

Hybrid CFD Engineering Model of Plume Induced Erosion and Crater Formation During Descent of Lunar Landers

With rapidly increased worldwide interest in landing on the moon, the issue of Plume Surface Interactions (PSI) is gaining attention. Hazards posed by lander plume induced dust and debris, as well as landing site deformation can be mitigated when better understood through predictive simulations. As simulation enabling computational power continues to increase, hybrid Computational Fluid Dynamics (CFD)/Engineering models provide the immediate ability to conduct parametric/trade studies driving design decisions for landers. Reduced order erosion engineering models apply correlations of the surface erosion rate to the plume induced surface forces with the correlations anchored to flight observations from Apollo LM landings. The MSFC propulsion fluid dynamics branch has developed such a hybrid model for predicting the plume induced viscous erosive regression of the Lunar surface beneath a landing vehicle. The (Descent Interpolated Gas Granular Erosion Model) DIGGEM was originally implemented as a post processing tool to calculate induced erosion rates through vehicle descent using CFD solutions of the vehicle plume at several fixed altitudes. This capability has since been advanced in the Loci/Chem-DIGGEM model to allow transient, moving vehicle, fully coupled viscous erosion modeling of vehicle descent PSI. This model has recently been used to make preflight predictions of the erosive regression of the ground beneath a Commercial Lunar Payload Services (CLPS) vehicle in support of measurements to be made by the Stereo Cameras for Lunar Plume Surface Studies (SCALPSS) instrument.

Plume Surface Interaction↗
NASA NTRS2023

Hybrid CFD Engineering Model of Plume Induced Erosion and Crater Formation During Descent of Lunar Landers

With rapidly increased worldwide interest in landing on the moon, the issue of Plume Surface Interactions (PSI) is gaining attention. Hazards posed by lander plume induced dust and debris, as well as landing site deformation can be mitigated when better understood through predictive simulations. As simulation enabling computational power continues to increase, hybrid Computational Fluid Dynamics (CFD)/Engineering models provide the immediate ability to conduct parametric/trade studies driving design decisions for landers. Reduced order erosion engineering models apply correlations of the surface erosion rate to the plume induced surface forces with the correlations anchored to flight observations from Apollo LM landings. The MSFC propulsion fluid dynamics branch has developed such a hybrid model for predicting the plume induced viscous erosive regression of the Lunar surface beneath a landing vehicle. The (Descent Interpolated Gas Granular Erosion Model) DIGGEM was originally implemented as a post processing tool to calculate induced erosion rates through vehicle descent using CFD solutions of the vehicle plume at several fixed altitudes. This capability has since been advanced in the Loci/Chem-DIGGEM model to allow transient, moving vehicle, fully coupled viscous erosion modeling of vehicle descent PSI. This model has recently been used to make preflight predictions of the erosive regression of the ground beneath a Commercial Lunar Payload Services (CLPS) vehicle in support of measurements to be made by the Stereo Cameras for Lunar Plume Surface Studies (SCALPSS) instrument.

Plume Surface Interaction↗
NASA NTRS1981

Evaluation of the Williams-type model for barley yields in North Dakota and Minnesota

The Williams-type yield model is based on multiple regression analysis of historial time series data at CRD level pooled to regional level (groups of similar CRDs). Basic variables considered in the analysis include USDA yield, monthly mean temperature, monthly precipitation, soil texture and topographic information, and variables derived from these. Technologic trend is represented by piecewise linear and/or quadratic functions of year. Indicators of yield reliability obtained from a ten-year bootstrap test (1970-1979) demonstrate that biases are small and performance based on root mean square appears to be acceptable for the intended AgRISTARS large area applications. The model is objective, adequate, timely, simple, and not costly. It consideres scientific knowledge on a broad scale but not in detail, and does not provide a good current measure of modeled yield reliability.

Barnett, T. L.↗
NASA NTRS1989

Analysis of flight data from a High-Incidence Research Model by system identification methods

Data partitioning and modified stepwise regression were applied to recorded flight data from a Royal Aerospace Establishment high incidence research model. An aerodynamic model structure and corresponding stability and control derivatives were determined for angles of attack between 18 and 30 deg. Several nonlinearities in angles of attack and sideslip as well as a unique roll-dominated set of lateral modes were found. All flight estimated values were compared to available wind tunnel measurements.

Batterson, James G.↗
NASA NTRS1992

Aerodynamic parameters of the X-31 drop model estimated from flight-data at high angles of attack

Lateral aerodynamic parameters of the X-31 drop model were estimated from flight data at angles of attack between 25 deg and 45 deg. Partitioned data from an ensemble of 12 maneuvers and data from 13 single maneuvers were analyzed by a stepwise regression technique to obtain an aerodynamic model structure and least squares parameter estimates. Because of data collinearity in several maneuvers, these maneuvers were reanalyzed by two biased estimation techniques, mixed estimation and fractional rank regression. The final parameter estimates in the form of stability and control derivatives were plotted against the angle of attack and compared with wind tunnel results and a limited number of estimates from full-scale aircraft data. There was no significant disagreement between parameters from the two sets of drop model data and the full-scale aircraft data. Some differences, however, existed between the dihedral, damping-in-roll, and aileron-effectiveness parameters from flight and wind tunnel data.

Klein, Vladislav↗
NASA NTRSDate not supplied

Towards an Aviation Large Language Model by Fine-tuning and Evaluating Transformers

In the aviation domain, there are many applications for machine learning and artificial intelligence tools that utilize natural language. For example, there is a desire to know the commonalities in written safety reports such as voluntary post incidents reports or aerial wildfire operations reports to better understand the risks present. Another use-case is the possibility of extracting airspace procedures and constraints currently written in documents such as Letters of Agreement. These applications can benefit from the use of state-of-the-art natural language processing techniques when adapted to the language/phraseology specific to the aviation domain. This paper evaluates the viability of adaptation of NLP tools to the aviation domain by fine-tuning transformer based models using aviation data sets. In 2018, a novel language model based on neural units (also called transformers) was created and became known as “Bidirectional Encoder Representations from Transformers” or BERT. This architecture combined with large amounts of English training data and innovative semi-supervised training tasks set the standard for what would later emerge as Large Language Models. The performance of these models was further improved by hyperparameter tuning and refinement of the semi-supervised training task and resulted in “Robustly Optimized BERT Pre-training Approach through hyperparameter tuning” or RoBERTa models. These pre-trained Large Language Models proved to be useful for a wide variety of natural language processing tasks such as text classification and question answering through a process called fine-tuning. The transformer architecture with pre-trained weights served as the basis with the last few layers replaced with layers fine-tuned to perform a new task e.g., a layer that provides a label for the entire input text. This process of fine-tuning can also be used to adapt the models to new domains; e.g., BioBERT started with the pre-trained BERT model and was completed by additional fine-tuning and training on biomedical documents. Transformer-based architectures can also be used to create rich representations of text called embeddings which can serve as the input to other machine learning models. This allows simpler algorithms such as logistic regression to use context-rich representations of the text while still remaining quick to train and evaluate. In the world of aviation, there is a growing demand for natural language processing and understanding but the domain presents unique challenges. Due to the technical content (and specialized language) of most aviation documents, fine-tuning pre-trained Large Language Models to specific tasks has not met the benchmark on natural language processing tasks set by simpler models trained from scratch on the data. To address this deficiency, this paper evaluates the improvements from fine-tuning a Large Language Model on a large set of aviation documents using the original semi-supervised training tasks before performing specific natural language tasks. In fine-tuning, a domain-specific dataset is used on the original training task but with the pre-trained Large Language Model instead of starting from a random initialization. This approach allows the model to be adapted to the specific domain language without discarding the information gained from training on general English data. This paper utilized two major dataset types to train and assess the RoBERTa fine-tuning performance. The first are 7,057 Letters of Agreement which are Federal Aviation Administration (FAA) documents that formalize airspace operations across the national airspace system. They contain many examples of ‘aviation English’ using domain specific terminology and phrasing which serves as a representative basis to perform the semi-supervised fine-tuning. The second type is the 494 document classification labels to be used for evaluation. This down-stream evaluation aims to show the performance of the fine-tuned model, better understand how much data is needed for an effective fine-tuning, and how fine-tuning can be adapted for different applications in-the domain. After semi-supervised training, evaluation begins by encoding the documents for classification using the fine-tuned RoBERTa model. Then a logistic regression classifier is trained to label the document type and compared against our ground truth labels. This currently leads to a 82.8% accuracy on 10-fold cross validation showing improvement over baseline RoBERTa which achieved 81.0%. We plan to measure the improvements on additional tasks and it is expected that these improvements will lead to more robust models that can tackle the natural language processing challenges present in aviation datasets.

ATM↗
NASA NTRS1987

Comparison of two numerical techniques for aerodynamic model identification

An algorithm, called the Minimal Residual QR algorithm, is presented to solve subset regression problems. It is shown that this scheme can be used as a numerically reliable implementation of the stepwise regression technique, which is widely used to identify an aerodynamic model from flight test data. This capability as well as the numerical superiority of this scheme over the stepwise regression technique is demonstrated in an experimental simulation study.

Verhaegen, M. H.↗
NASA NTRS2011

Sensitivity of C-Band Polarimetric Radar-Based Drop Size Distribution Measurements to Maximum Diameter Assumptions

The estimation of rain drop size distribution (DSD) parameters from polarimetric radar observations is accomplished by first establishing a relationship between differential reflectivity (Z(sub dr)) and the central tendency of the rain DSD such as the median volume diameter (D0). Since Z(sub dr) does not provide a direct measurement of DSD central tendency, the relationship is typically derived empirically from rain drop and radar scattering models (e.g., D0 = F[Z (sub dr)] ). Past studies have explored the general sensitivity of these models to temperature, radar wavelength, the drop shape vs. size relation, and DSD variability. Much progress has been made in recent years in measuring the drop shape and DSD variability using surface-based disdrometers, such as the 2D Video disdrometer (2DVD), and documenting their impact on polarimetric radar techniques. In addition to measuring drop shape, another advantage of the 2DVD over earlier impact type disdrometers is its ability to resolve drop diameters in excess of 5 mm. Despite this improvement, the sampling limitations of a disdrometer, including the 2DVD, make it very difficult to adequately measure the maximum drop diameter (D(sub max)) present in a typical radar resolution volume. As a result, D(sub max) must still be assumed in the drop and radar models from which D0 = F[Z(sub dr)] is derived. Since scattering resonance at C-band wavelengths begins to occur in drop diameters larger than about 5 mm, modeled C-band radar parameters, particularly Z(sub dr), can be sensitive to D(sub max) assumptions. In past C-band radar studies, a variety of D(sub max) assumptions have been made, including the actual disdrometer estimate of D(sub max) during a typical sampling period (e.g., 1-3 minutes), D(sub max) = C (where C is constant at values from 5 to 8 mm), and D(sub max) = M*D0 (where the constant multiple, M, is fixed at values ranging from 2.5 to 3.5). The overall objective of this NASA Global Precipitation Measurement Mission (GPM/PMM Science Team)-funded study is to document the sensitivity of DSD measurements, including estimates of D0, from C-band Z(sub dr) and reflectivity to this range of D(sub max) assumptions. For this study, GPM Ground Validation 2DVD's were operated under the scanning domain of the UAHuntsville ARMOR C-band dual-polarimetric radar. Approximately 7500 minutes of DSD data were collected and processed to create gamma size distribution parameters using a truncated method of moments approach. After creating the gamma parameter datasets the DSD's were then used as input to a T-matrix model for computation of polarimetric radar moments at C-band. All necessary model parameterizations, such as temperature, drop shape, and drop fall mode, were fixed at typically accepted values while the D(sub max) assumption was allowed to vary in sensitivity tests. By hypothesizing a DSD model with D(sub max) (fit) from which the empirical fit to D0 = F[Z(sub dr)] was derived via non-linear least squares regression and a separate reference DSD model with D(sub max) (truth), bias and standard error in D0 retrievals were estimated in the presence of Z(sub dr) measurement error and hypothesized mismatch in D(sub max) assumptions. Although the normalized standard error for D0 = F[Z(sub dr)r] can increase slightly (as much as from 11% to 16% for all 7500 DSDs) when the D(sub max) (fit) does not match D(sub max) (truth), the primary impact of uncertainty in D(sub max) is a potential increase in normalized bias error in D0 (from 0% to as much as 10% over all 7500 DSDs, depending on the extent of the mismatch between D(sub max) (fit) and D(sub max) (truth)). For DSDs characterized by large Z(sub dr) (Z(sub dr) > 1.5 to 2.0 dB), the normalized bias error for D0 estimation at C-band is sometimes unacceptably large (> 10%), again depending on the extent of the hypothesized D(sub max) mismatch. Modeled errors in D0 retrievals from Z(sub dr) at C-band are demonstrated in detail and comparedo similar modeled retrieval errors at S-band and X-band where the sensitivity to D(sub max) is expected to be less. The impact of D(sub max) assumptions to the retrieval of other DSD parameters such as Nw, the liquid water content normalized intercept parameter, are also explored. Likely implications for DSD retrievals using C-band polarimetric radar for GPM are assessed by considering current community knowledge regarding D(sub max) and quantifying the statistical distribution of Z(sub dr) from ARMOR over a large variety of meteorological conditions. Based on these results and the prevalence of C-band polarimetric radars worldwide, a call for more emphasis on constraining our observational estimate of D(sub max) within a typical radar resolution volume is made

Carey, Lawrence D.↗
NASA NTRS1993

The analytical representation of viscoelastic material properties using optimization techniques

This report presents a technique to model viscoelastic material properties with a function of the form of the Prony series. Generally, the method employed to determine the function constants requires assuming values for the exponential constants of the function and then resolving the remaining constants through linear least-squares techniques. The technique presented here allows all the constants to be analytically determined through optimization techniques. This technique is employed in a computer program named PRONY and makes use of commercially available optimization tool developed by VMA Engineering, Inc. The PRONY program was utilized to compare the technique against previously determined models for solid rocket motor TP-H1148 propellant and V747-75 Viton fluoroelastomer. In both cases, the optimization technique generated functions that modeled the test data with at least an order of magnitude better correlation. This technique has demonstrated the capability to use small or large data sets and to use data sets that have uniformly or nonuniformly spaced data pairs. The reduction of experimental data to accurate mathematical models is a vital part of most scientific and engineering research. This technique of regression through optimization can be applied to other mathematical models that are difficult to fit to experimental data through traditional regression techniques.

Hill, S. A.↗
NASA NTRS1985

Crop area estimation based on remotely-sensed data with an accurate but costly subsample

Research activities conducted under the auspices of National Aeronautics and Space Administration Cooperative Agreement NCC 9-9 are discussed. During this contract period research efforts are concentrated in two primary areas. The first are is an investigation of the use of measurement error models as alternatives to least squares regression estimators of crop production or timber biomass. The secondary primary area of investigation is on the estimation of the mixing proportion of two-component mixture models. This report lists publications, technical reports, submitted manuscripts, and oral presentation generated by these research efforts. Possible areas of future research are mentioned.

Gunst, R. F.↗
NASA NTRS1987

Comparison of two numerical techniques for aerodynamic model identification

A new algorithm, called the Minimal Residual QR algorithm, is presented to solve subset regression problems. It is shown that this new scheme can be used as a numerically reliable implementation of the stepwise regression technique, which is widely used to identify an aerodynamic model from flight test data. This capability as well as the numerical superiority of this scheme over the stepwise regression technique is demonstrated in an experimental simulation study.

Verhaegen, M. H.↗
NASA NTRS1983

On the determination of airplane model structure form flight data

A procedure based on a modified stepwise regression and several selection criteria is presented for the determination of airplane model structure from flight data. The aerodynamic force and moment coefficients in an airplane model are expresed either as polynomials in output and input variables or as a combination of splines. The procedure is demonstrated in three examples by attempting to determine a local, extended and global model. Some of the resulting models are verified by using the maximum likelihood estimation or by examining model prediction capabilities.

Klein, V.↗
NASA NTRS2002

Modeling and control for closed environment plant production systems

A computer program was developed to study multiple crop production and control in controlled environment plant production systems. The program simulates crop growth and development under nominal and off-nominal environments. Time-series crop models for wheat (Triticum aestivum), soybean (Glycine max), and white potato (Solanum tuberosum) are integrated with a model-based predictive controller. The controller evaluates and compensates for effects of environmental disturbances on crop production scheduling. The crop models consist of a set of nonlinear polynomial equations, six for each crop, developed using multivariate polynomial regression (MPR). Simulated data from DSSAT crop models, previously modified for crop production in controlled environments with hydroponics under elevated atmospheric carbon dioxide concentration, were used for the MPR fitting. The model-based predictive controller adjusts light intensity, air temperature, and carbon dioxide concentration set points in response to environmental perturbations. Control signals are determined from minimization of a cost function, which is based on the weighted control effort and squared-error between the system response and desired reference signal.

NASA Discipline Life Support Systems↗