Search NASA⌕ Search

SEARCH · Search NASA

Results for “multiple regression”

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 613 records · Page 34

Factors Affecting Route Selection and Survival of Steelhead Kelts at Snake River Dams in 2012 and 2013

In 2012 and 2013, Pacific Northwest National Laboratory (PNNL) conducted a study that summarized the passage route proportions and route-specific survival rates of steelhead kelts that passed through Federal Columbia River Power System (FCRPS) dams. To accomplish this, a total of 811 steelhead kelts were tagged with Juvenile Salmon Acoustic Telemetry System (JSATS) transmitters. Acoustic receivers, both autonomous and cabled, were deployed throughout the FCRPS to monitor the downstream movements of tagged kelts. Kelts were also tagged with passive integrated transponder tags to monitor passage through juvenile bypass systems (JBS) and detect returning fish. The current study evaluated data collected in 2012 and 2013 to identify environmental, temporal, operational, individual, and behavioral variables that were related to forebay residence time, route of passage, and survival of steelhead kelts at FCRPS dams on the Snake River. Multiple approaches, including 3-D tracking, bivariate and multivariable regression modeling, and decision tree analyses were used to identify the environmental, temporal, operational, individual, and behavioral variables that had the greatest effect on forebay residence time, route of passage, and route-specific and overall dam passage survival probabilities for tagged kelts at Lower Granite (LGR), Little Goose (LGS), and Lower Monumental (LMN) dams. In general, kelt behavior and discharge appeared to work independently to affect forebay residence times. Kelt behavior, primarily approach location, migration depth, and “searching” activities in the forebay, was found to have the greatest influence on their route of passage. The condition of kelts was the single most important factor affecting their survival. The information gathered in this study may be used by dam operators and fisheries managers to identify potential management actions to improve in-river survival of kelts or collection methods for kelt reconditioning programs to aid the recovery of Snake River steelhead populations.

13 HYDRO ENERGY↗

wa-hls4ml and lui-gnn: A benchmark and GNN-based surrogate model for hls4ml resource and latency estimation

As machine learning (ML) increasingly serves as a tool for addressing real-time challenges in scientific applications, the development of advanced tooling has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as model synthesis, are now becoming limiting factors in the rapid iteration of designs. To reduce these emerging constraints, multiple efforts are being launched toward designing an ML-based surrogate model that estimates resource usage of synthesized accelerator architectures. This model would reduce the design iteration time, especially when designing within a set of given hardware constraints. This approach shows considerable potential, but as it stands, the effort is early and would benefit from coordination and standardization to assist future work as it emerges. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of more than 100,000 fully connected neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. In addition to the resource utilization and latency data provided, the dataset includes generated artifacts and log files for many of the synthesized neural networks, in order to support future research in ML-based code generation. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, as well as the average performance across a subset of the dataset. We measure the performance of a given predictor model through multiple metrics, including $R^2$ score and SMAPE on regression tasks, as well as inference time to further characterize the estimator under test. Additionally, we introduce the latency/utilization inference graph neural network (lui-gnn), a surrogate model that uses a graph neural network to represent input architectures in the form of a directed graph. This graph representation allows for a diverse set of model architectures to all be effectively handled by a surrogate model. We present the architecture and performance of the model, as evaluated by the new proposed benchmark, including SMAPE, $R^2$ score, and inference times, and find that lui-gnn generally predicts latency and utilization for the 75\% quantile within several percent of the synthesized resources on the synthetic test dataset, indicating that this approach of estimating resource and latency via a surrogate models has promise and warrants further research.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

ERTS data user investigation to develop a multistage forest sampling inventory system

The author has identified the following significant results. A unique digital timber volume estimation system was developed for use with the MSS CCT tapes. The system was tested on a 64-square mile area in Northern California's Trinity Alps. The outcome of a systematic experiment, in which several possible combinations of bands 5 and 7 and a contrast measure were tried, showed that an estimated gain in precision of 50% can be obtained in a multistage sampling design. The difference between bands 5 and 7 proved to be of special importance for the estimation of biomass in the form of timber volume. In addition, an interpretation model for high flight U2 photographs was developed. A maximum multiple correlation coefficient of 0.74 was obtained for the regression model, explaining 55% of the variation in timber volume as estimated from aerial photos and ground measurments. An interpretation model for MSS color composites is in the testing stage.

Langley, P. G.↗

Health Monitoring System for the SSME-fault detection algorithms

A Health Monitoring System (HMS) Framework for the Space Shuttle Main Engine (SSME) has been developed by United Technologies Corporation (UTC) for the NASA Lewis Research Center. As part of this effort, fault detection algorithms have been developed to detect the SSME faults with sufficient time to shutdown the engine. These algorithms have been designed to provide monitoring coverage during the startup, mainstage and shutdown phases of the SSME operation. The algorithms have the capability to detect multiple SSME faults, and are based on time series, regression and clustering techniques. This paper presents a discussion of candidate algorithms suitable for fault detection followed by a description of the algorithms selected for implementation in the HMS and the results of testing these algorithms with the SSME test stand data.

Tulpule, S.↗

Refining fast simulation using machine learning

At the CMS experiment, a growing reliance on the fast Monte Carlo application (FastSim) will accompany the high luminosity and detector granularity expected in Phase 2. The FastSim chain is roughly 10 times faster than the application based on the GEANT4 detector simulation and full reconstruction referred to as FullSim. However, this advantage comes at the price of decreased accuracy in some of the final analysis observables. In this contribution, a machine learning-based technique to refine those observables is presented. We employ a regression neural network trained with a sophisticated combination of multiple loss functions to provide post-hoc corrections to samples produced by the FastSim chain. The results show considerably improved agreement with the FullSim output and an improvement in correlations among output observables and external parameters. This technique is a promising replacement for existing correction factors, providing higher accuracy and thus contributing to the wider usage of FastSim.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Landsat 8 monitoring of multi-depth suspended sediment concentrations in Lake Erie’s Maumee River using machine learning

Satellite remote sensing has been widely used to map suspended sediment concentration (SSC) in waterbodies. However, due to the complexity of sediment-water interactions, it has been difficult to derive linear and non-linear regression equations to reliably predict SSC, especially when trying to estimate depth of integrated sediment. Herein, this study uses Landsat 8 OLI (Operational Land Imager) sensor to map SSC within the Maumee River in Ohio, USA, at multiple depth intervals (15, 61, 91, and 182 cm). Simple linear least squares regression (LLSR), and three common machine learning models: random forest (RF), support vector regression (SVR), and model averaged neural network (MANN) were used to estimate SSC at the depth intervals. All machine learning models significantly outperformed LLSR while RF performed the best. In both RF and MANN, R2 (coefficient of determination) increases with depth with a maximum R2 of 0.89 and 0.83, respectively, at a depth of 0–182 cm. The results show that machine learning models can implement nonlinear relationships that produce better predictions than traditional linear regression methods in estimating depth integrated SSC, especially when samples are limited.

47 OTHER INSTRUMENTATION↗

Influence of amplitude-related perfusion parameters in the parotid glands by non-fat-saturated dynamic contrast-enhanced magnetic resonance imaging

Purpose: To verify whether quantification of parotid perfusion is affected by fat signals on non-fat-saturated (NFS) dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and whether the influence of fat is reduced with fat saturation (FS). Methods: This study consisted of three parts. First, a retrospective study analyzed DCE-MRI data previously acquired on different patients using NFS (n = 18) or FS (n = 18) scans. Second, a phantom study simulated the signal enhancements in the presence of gadolinium contrast agent at six concentrations and three fat contents. Finally, a prospective study recruited nine healthy volunteers to investigate the influence of fat suppression on perfusion quantification on the same subjects. Parotid perfusion parameters were derived from NFS and FS DCE-MRI data using both pharmacokinetic model analysis and semiquantitative parametric analysis. T tests and linear regression analysis were used for statistical analysis with correction for multiple comparisons. Results: NFS scans showed lower amplitude-related parameters, including parameter A, peak enhancement (PE), and slope than FS scans in the patients (all with P < 0.0167). The relative signal enhancement in the phantoms was proportional to the dose of contrast agent and was lower in NFS scans than in FS scans. The volunteer study showed lower parameter A (6.75 ± 2.38 a.u.), PE (42.12% ± 14.87%), and slope (1.43% ± 0.54% s{sup −1}) in NFS scans as compared to 17.63 ± 8.56 a.u., 104.22% ± 25.15%, and 9.68% ± 1.67% s{sup −1}, respectively, in FS scans (all with P < 0.005). These amplitude-related parameters were negatively associated with the fat content in NFS scans only (all with P < 0.05). Conclusions: On NFS DCE-MRI, quantification of parotid perfusion is adversely affected by the presence of fat signals for all amplitude-related parameters. The influence could be reduced on FS scans.

60 APPLIED LIFE SCIENCES↗

Inference for Nonparanormal Partial Correlation via Regularized Rank-Based Nodewise Regression

Abstract Partial correlation is a common tool in studying conditional dependence for Gaussian distributed data. However, partial correlation being zero may not be equivalent to conditional independence under non-Gaussian distributions. In this paper, we propose a statistical inference procedure for partial correlations under the high-dimensional nonparanormal (NPN) model where the observed data are normally distributed after certain monotone transformations. The NPN partial correlation is the partial correlation of the normal transformed data under the NPN model, which is a more general measure of conditional dependence. We estimate the NPN partial correlations by regularized nodewise regression based on the empirical ranks of the original data. A multiple testing procedure is proposed to identify the nonzero NPN partial correlations. The proposed method can be carried out by a simple coordinate descent algorithm for lasso optimization. It is easy-to-implement and computationally more efficient compared to the existing methods for estimating NPN graphical models. Theoretical results are developed to show the asymptotic normality of the proposed estimator and to justify the proposed multiple testing procedure. Numerical simulations and a case study on brain imaging data demonstrate the utility of the proposed procedure and evaluate its performance compared to the existing methods. Data used in preparation of this article were obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database.

97 MATHEMATICS AND COMPUTING↗

Machine learning models for volumetric swelling in uranium nitride

Machine learning methods are applied to predict the volumetric swelling rate of the nuclear fuel uranium nitride (UN) over various temperatures, irradiation conditions, and power densities. Both kernel-based methods and symbolic regression models for UN swelling are developed and compared with multiple experimental datasets. We find that the UN pellet geometry and dimensions must be taken into account to accurately model swelling behavior. Strong agreement is observed between the developed machine learning models and the data. The predictive error generated by the machine learning models improves on empirical models taken from the literature. Sensitivity analysis is performed to determine which properties such as temperature, burnup, and power density, are most important in the swelling process. We find that machine learning can be used to quickly develop accurate swelling models for nuclear materials. In conclusion, the presented results illustrate the potential of machine learning to determine volumetric swelling in UN.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Low-dose gold nanoparticles exert subtle endocrine-modulating effects on the ovarian steroidogenic pathway ex vivo independent of oxidative stress

Gold nanoparticles (GNPs) have gained considerable attention for application in science and industry. However, the untoward effects of such particles on female fertility remain unclear. The objectives of this study were to (1) examine the effects of 10-nm GNPs on progesterone and estradiol-17b accumulation by rat ovaries ex vivo and (2) to identify the locus/loci whereby GNPs modulate steroidogenesis via multiple-reference gene quantitative real-time RT-PCR. Regression analyses indicated a positive relationship between both Star (p < 0.05, r2 = 0.278) and Cyp11a1 (p < 0.001, r2 = 0.366) expression and P4 accumulation. upon exposure to 1.43 * 106 GNPs/mL. Additional analyses showed that E2 accumulation was positively associated with Hsd3b1 (p < 0.05, r2 = 0.181) and Cyp17a1 (p < 0.01, r2 = 0.301) expression upon exposure to 1.43 * 13 and 1.43 * 109 GNPs/mL, respectively. These results suggest a subtle treatmentdependent impact of low-dose GNPs on the relationship between progesterone or estradiol-17b and specific steroidogenic target genes, independent of oxidative stress or inhibin.

Larson, Jeremy K.↗

Simultaneous quantification of uranium( VI ), samarium, nitric acid, and temperature with combined ensemble learning, laser fluorescence, and Raman scattering for real-time monitoring

In this work, laser-induced fluorescence spectroscopy (LIFS), Raman spectroscopy, and a stacked regression ensemble was developed for near real-time quantification of uranium(VI) (1–100 μg mL –1 ), samarium (0–200 μg mL –1 ) and nitric acid (0.1–4 M) with varying temperature (20 °C–45 °C). LIFS applications range from fundamental lab-scale studies to real-time process monitoring at industrial levels, such as nuclear reprocessing applications, provided the phenomena affecting the fluorescence spectrum are accounted for (e.g., absorption, quenching, complexation). Multiple chemometric models were examined and compared to a more traditional multivariate regression approach called partial least squares (PLS). Results obtained on synthetic samples selected using D-optimal experimental design indicated that a stacked regression method, which included ridge regression, random forest, PLS, and an eXtreme gradient boost algorithm, successfully measured uranium(VI) concentrations directly in nitric acid without measuring luminescence lifetimes or standard addition. The top model resulted in percent root-mean-square error of prediction values of 5.2, 1.9, 3.0, and 2.3% for U(VI), Sm 3+ , HNO 3 , and temperature, respectively. The approach may be useful for quantifying fluorescent fission products (e.g., Sm 3+ ) to provide information on burnup of irradiated nuclear fuel. This novel framework reinforces the applicability of LIFS for real-time applications in nuclear fuel cycle applications.

38 RADIATION CHEMISTRY, RADIOCHEMISTRY, AND NUCLEA↗

Relative effect of anthropogenic warming and natural climate variability to changes in Compound drought and heatwaves

Compound drought and heatwave (CDHW) events can be influenced by large scale teleconnections and anthropogenic warming, leading to severe socio-economic impacts across various climate regions. Here, the relative influence of six different teleconnection patterns and anthropogenic global warming on the global CDHW occurrences is quantified systematically using the instrumental data period, 1982–2016. The results from the study suggest a substantial increase in the CDHW events (1–5 events per year) across various parts of the globe at the beginning of 21st century (2000–2016). A Bayesian approach is implemented to identify the most vulnerable climate regions based on the degree of susceptibility of heatwaves (DSHW) towards drought. As such, top ten most vulnerable regions are selected based on the DSHW magnitude, and a partial correlation analysis is performed to select the natural and anthropogenic drivers of CDHW in those regions, separately. A logistic regression model is then used to determine significant changes in the odds of CDHW due to changes in the selected drivers that suggest a significantly positive, and multiplicative effect of anthropogenic global warming in the top ten most vulnerable climate regions. Finally, the same logistic regression model, integrated with an analytical framework, is applied to determine the relative influence of anthropogenic global warming on the changes in odds of CDHW for the future, 1.5 °C and 2 °C warming limits. Finally, the results suggest that relative to the 2 °C global warming, constraining to the 1.5 °C global warming limit may conduce about 17-fold reduction in the odds of CDHW in the most vulnerable climate region, East Asia, 5–8-fold reduction in Western North America, Northern Australia, Central North America, Central Europe, South Asia, and the Mediterranean region, and 3–4-fold reduction in Northeastern Brazil, Eastern North America, and West Asia.

54 ENVIRONMENTAL SCIENCES↗

Interpretable Net Load Forecasting Using Smooth Multiperiodic Features

We consider the problem of forecasting net load over a horizon such as one day, using a trailing window of past net load values as well as date and time. We focus on three variations on this problem: point forecasts, marginal quantile forecasts, and generating conditional samples of the future value. We propose a method that relies on linear regression using some custom engineered time-based features to capture multiple periodicities, such as daily, weekly, and seasonal, and their interactions. Our proposed models are readily interpretable, and rely on efficient and reliable convex optimization [1] to fit. We illustrate our method on four years worth of hourly net load data, comparing predictions made with various subsets of the features.

Ogut, Mehmet G↗

Modeling Longitudinal Data Containing Non-Normal Within Subject Errors

The mission of the National Aeronautics and Space Administration’s (NASA) human research program is to advance safe human spaceflight. This involves conducting experiments, collecting data, and analyzing data. The data are longitudinal and result from a relatively few number of subjects; typically 10 – 20. A longitudinal study refers to an investigation where participant outcomes and possibly treatments are collected at multiple follow-up times. Standard statistical designs such as mean regression with random effects and mixed–effects regression are inadequate for such data because the population is typically not approximately normally distributed. Hence, more advanced data analysis methods are necessary. This research focuses on four such methods for longitudinal data analysis: the recently proposed linear quantile mixed models (lqmm) by Geraci and Bottai (2013), quantile regression, multilevel mixed–effects linear regression, and robust regression. This research also provides computational algorithms for longitudinal data that scientists can directly use for human spaceflight and other longitudinal data applications, then presents statistical evidence that verifies which method is best for specific situations. This advances the study of longitudinal data in a broad range of applications including applications in the sciences, technology, engineering and mathematics fields.

Feiveson, Alan↗

Variance Decomposition of MEDLI2 Reconstructed Heating Using Neural Networks

The Mars Entry, Descent, and Landing Instrumentation (MEDLI2) sensor suite collected data during entry of the Mars 2020 Perseverance rover into Mars’ atmosphere. This suite included a network of MEDLI2 Instrumented Sensor Plugs (MISPs). Each MISP was comprised of a cylinder made of Thermal Protection System (TPS) material with 1-3 embedded thermocouples (TCs), and it was flush mounted into the heatshield or backshell. Data from these in-depth TCs were used to reconstruct the aeroheating environment of the vehicle throughout entry. Surface heating was posed as an inverse problem, with the goal of estimating the surface heating by minimizing an objective function of the difference between MISP temperature measurements during flight and the temperature predictions derived from the Fully Implicit Ablation and Thermal response (FIAT) program. Given an aerothermal environment, FIAT calculates the material response and provides in-depth temperatures throughout the TPS material. To achieve the reverse, an internal tool called FIAT_Opt runs through multiple different environments until the output temperature at the TC depth closely matches the flight data. 95% confidence intervals on the reconstructed surface heating were obtained using Monte Carlo analysis, in which uncertainties in the thermocouple depth and the TPS material properties (e.g., density, thermal conductivity, heat capacity, emissivity) based on flight-lot material testing were included. A variance decomposition method using Sobol indices was employed to assess the sensitivity of the reconstructed peak heating to the TC placement and material property uncertainties. Variance decomposition was found to require tens of thousands of FIAT_Opt runs in order for the Sobol indices to converge. With a single FIAT_Opt run taking on the order of 40 minutes, the required number of computations would take months to complete, even if using multiple CPUs. To mitigate this problem, three machine learning models (ridge regression with cross-validation, random forest regression, and a deep neural network) were trained and tested using the 2000 Monte Carlo runs that were already completed. A subset of 1600 runs were used to train the model (i.e., training set), while the remaining 400 runs were used as the test set. The predictions from the deep neural network (DNN) on the test set showed nearly perfect agreement to the actual values computed with FIAT_Opt (R2 > 0.99). Using the DNN as a surrogate model, the variance decomposition using 50,000 runs was completed within minutes. The resulting Sobol indices showed that the reconstructed peak surface heating was most sensitive to the uncertainties in the thermal conductivity (ST = 0.37) and heat capacity (ST = 0.26). This method can be leveraged to provide requirements for material property measurements needed to improve the accuracy of surface heating prediction and ultimately lead to the reduction of design margins in the future. This presentation will include background on the MEDLI2 suite; the method used for inverse heating estimation; the way that material property uncertainties were accounted for using Monte Carlo analysis; a brief background on variance decomposition; the motivation for using machine learning in this context; how a neural network was trained on the data to enable variance decomposition in a fraction of the time; and the variance decomposition results for one of the MISPs.

Hannah Alpert↗

Spirometric traits show quantile-dependent heritability, which may contribute to their gene-environment interactions with smoking and pollution

Background: “Quantile-dependent expressivity” refers to a genetic effect that is dependent upon whether the phenotype (e.g., spirometric data) is high or low relative to its population distribution. Forced vital capacity (FVC), forced expiratory volume in 1 second (FEV 1 ), and the FEV 1 /FVC ratio are moderately heritable spirometric traits. The aim of the analyses is to test whether their heritability (h 2 ) is constant over all quantiles of their distribution. Methods: Quantile regression was applied to the mean age, sex, height and smoking-adjusted spirometric data over multiple visits in 9,993 offspring-parent pairs and 1,930 sibships from the Framingham Heart Study to obtain robust estimates of offspring-parent (β OP ), offspring-midparent (β OM ), and full-sib regression slopes (β FS ). Nonparametric significance levels were obtained from 1,000 bootstrap samples. β OP s were used as simple indicators of quantile-specific heritability (i.e.,h 2 = 2β OP /(1+r spouse ), where r spouse was the correlation between spouses). Results: β OP ± standard error (SE) decreased by 0.0009 ± 0.0003 (P = 0.003) with every one-percent increment in the population distribution of FEV 1 /FVC, i.e., β OP ± SE were: 0.182 ± 0.031, 0.152 ± 0.015; 0.136 ± 0.011; 0.121 ± 0.013; and 0.099 ± 0.013 at the 10th, 25th, 50th, 75th, and 90th percentiles of the FEV 1 /FVC distribution, respectively. These correspond to h 2 ± SEs of 0.350 ± 0.060 at the 10th, 0.292 ± 0.029 at the 25th, 0.262 ± 0.020 at the 50th, 0.234 ± 0.025 at the 75th, and 0.191 ± 0.025 at the 90th percentiles of the FEV 1 /FVC ratio. Maximum mid-expiratory flow (MMEF) h 2 ± SEs increased 0.0025 ± 0.0007 (P = 0.0004) with every one-percent increment in its distribution, i.e.: 0.467 ± 0.046, 0.467 ± 0.033, 0.554 ± 0.038, 0.615 ± 0.042, and 0.675 ± 0.060 at the 10th, 25th, 50th, 75th, and 90th percentiles of its distribution. This was due to forced expiratory flow at 75% of FVC (FEF75%), whose quantile-specific h 2 increased an average of 0.0042 ± 0.0008 for every one-percent increment in its distribution. It is speculated that previously reported gene-environment interactions may be partially attributable to quantile-specific h 2 , i.e., greater heritability in individuals with lower FEV 1 /FVC due to smoking or airborne particles exposure vs. nonsmoking, unexposed individuals. Conclusion: Heritabilities of FEV 1 /FVC, MMEF, and FEF75% from quantile-regression of offspring-parent and sibling spirometric data suggest their quantile-dependent expressivity.

59 BASIC BIOLOGICAL SCIENCES↗

A Tutorial on Bayesian analysis of linear shock compression data

Gas gun and other shock compression experiments often produce shock wave velocity measurements that are linearly associated with particle velocity. Traditionally, this empirical relationship is quantified with a single Hugoniot curve that is estimated using least squares regression. However, for downstream modeling and simulation tasks, it is often more useful to have multiple Hugoniot curves in the pressure–volume plane that are consistent with the data. We employ Bayesian uncertainty quantification methods as a framework for propagating measurement uncertainty through to model parameters and predictions. Specifically, this Tutorial shows how to sample multiple Hugoniot curves in the pressure–volume plane that are consistent with the shock wave-particle velocity measurements in a two-step Bayesian approach. First, we obtain an analytical expression for the posterior distribution of the linear model parameters using Bayesian linear regression. Second, we propagate samples from the posterior distribution through the Rankine–Hugoniot equations to yield Hugoniot curves in the pressure–volume plane. The procedure is demonstrated with publicly available data on argon, copper, and nickel, and compared against bootstrapping and linear regression. The Bayesian procedure is shown to be interpretable, computationally inexpensive, and less sensitive than an alternative bootstrapping approach to the removal of the point in the copper dataset that has the largest particle velocity. As a Tutorial on Bayesian methodology for the shock compression community, we provide several derivations and explanations that make this paper self-contained, and make all code and data available at github.com/llnl/BALSCD.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Combined Data and Deep Learning Model Uncertainties: An Application to the Measurement of Solid Fuel Regression Rate

In complex physical process characterization, such as the measurement of the regression rate for solid hybrid rocket fuels, where both the observation data and the model used have uncertainties originating from multiple sources, combining these in a systematic way for quantities of interest (QoI) remains a challenge. In this paper, we present a forward propagation uncertainty quantification (UQ) process to produce a probabilistic distribution for the observed regression rate r. We characterized two input data uncertainty sources from the experiment (the distortion from the camera U c and the non-zero-angle fuel placement U Y ), the prediction and model form uncertainty from the deep neural network (U m ), as well as the variability from the manually segmented images used for training it (U s ). Here, we conducted seven case studies on combinations of these uncertainty sources with the model form uncertainty. The main contribution of this paper is the investigation and inclusion of the experimental image data uncertainties involved, and how to include them in a workflow when the QoI is the result of multiple sequential processes.

42 ENGINEERING↗