Search NASA⌕ Search

SEARCH · Search NASA

Results for “Model selection”

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 1,045 records · Page 58

Pressure-Sensitive Paint and Video Model Deformation Systems at the NASA Langley Unitary Plan Wind Tunnel

Pressure-sensitive paint (PSP) and video model deformation (VMD) systems have been installed in the Unitary Plan Wind Tunnel at the NASA Langley Research Center to support the supersonic wind tunnel testing requirements of the High Speed Research (HSR) program. The PSP and VMD systems have been operational since early 1996 and provide the capabilities of measuring global surface static pressures and wing local twist angles and deflections (bending). These techniques have been successfully applied to several HSR wind tunnel models for wide ranges of the Mach number, Reynolds number, and angle of attack. A review of the UPWT PSP and VMD systems is provided, and representative results obtained on selected HSR models are shown. A promising technique to streamline the wind tunnel testing process, Modern Experimental Design, is also discussed in conjunction with recently-completed wing deformation measurements at UPWT.

Erickson, G. E.↗

Testing and Analytical Modeling for Purging Process of a Cryogenic Line

To gain confidence in developing analytical models of the purging process for the cryogenic main propulsion systems of upper stage, two test series were conducted. The test article, a 3.35 m long with the diameter of 20 cm incline line, was filled with liquid or gaseous hydrogen and then purged with gaseous helium (GHe). Total of 10 tests were conducted. The influences of GHe flow rates and initial temperatures were evaluated. The Generalized Fluid System Simulation Program (GFSSP), an in-house general-purpose fluid system analyzer computer program, was utilized to model and simulate selective tests. The test procedures, modeling descriptions, and the results are presented in the following sections.

Hedayat, A.↗

Testing and Analytical Modeling for Purging Process of a Cryogenic Line

To gain confidence in developing analytical models of the purging process for the cryogenic main propulsion systems of upper stage, two test series were conducted. The test article, a 3.35 m long with the diameter of 20 cm incline line, was filled with liquid or gaseous hydrogen and then purged with gaseous helium (GHe). Total of 10 tests were conducted. The influences of GHe flow rates and initial temperatures were evaluated. The Generalized Fluid System Simulation Program (GFSSP), an in-house general-purpose fluid system analyzer computer program, was utilized to model and simulate selective tests. The test procedures, modeling descriptions, and the results are presented in the following sections.

Hedayat, A.↗

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↗

MLtool Python Code

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↗

Multiscale Simulation of Precipitation Over East Asia by Variable Resolution CAM-MPAS

This study evaluates the precipitation over East Asia simulated by the variable resolution CAM-MPAS featuring the Model for Prediction Across Scales (MPAS) atmospheric dynamical core coupled with the physics parameterizations of the Community Atmosphere Model (CAM) version 5.4. Two CAM-MPAS experiments, one with a global quasi-uniform resolution mesh of 120 km (MPAS-UR) and the other with a variable resolution mesh of 30 to 120 km refined over East Asia (MPAS-VR), are conducted from 1989 to 2005 following the Atmospheric Model Intercomparison Project protocol. Two regional climate model simulations at ~25 km resolution from the Coordinated Regional Downscaling Experiment East Asia second plan are also analyzed for comparison. Results show that CAM-MPAS performs better than the selected regional models in simulating the precipitation climatology over East Asia. Compared with MPAS-UR, MPAS-VR with refinement over East Asia better simulates the precipitation over and around the Tibetan Plateau (TP), the frequency distribution of light, moderate, and heavy precipitation in various regions of China, and the interannual precipitation pattern associated with the East Asian summer monsoon. The improvements in MPAS-VR relative to MPAS-UR largely come from the decreased grid spacing rather than the use of finer-scale terrain information. Particularly, decreased grid spacing is more essential for simulating the precipitation magnitude over TP and the location of peak precipitation south of TP, likely due to the better resolved physical and dynamical processes associated with orographic precipitation. The 30-120 km CAM-MPAS model is shown to be a promising tool for precipitation simulation over East Asia.

54 ENVIRONMENTAL SCIENCES↗

Infrared-Fused Vision-Based Thermoregulation Performance Estimation for Personal Thermal Comfort-Driven HVAC System Controls

Thermal comfort is one of the primary factors influencing occupant health, well-being, and productivity in buildings. Existing thermal comfort systems require occupants to frequently communicate their comfort vote via a survey which is impractical as a long-term solution. Here, we present a novel thermal infrared-fused computer vision sensing method to capture thermoregulation performance in a non-intrusive and non-invasive manner. In this method, we align thermal and visible images, detect facial segments (i.e., nose, eyes, face boundary), and accordingly read the temperatures from the appropriate coordinates in the thermal image. We focus on the human face since it is often clearly visible to cameras and is not merged into a hot background (unlike hands). We use a regularized Gaussian Mixture model to track the thermoregulation changes over time and apply a heuristic algorithm to extract hot and cold indices. We present a personalized and a generalized comfort modeling method, selected based on the availability of the occupant historical indices measurements in a neutral environment, and use the time-series of the hot and cold indices to define corrections to HVAC system operations in the form of setpoint constraints. To evaluate the efficacy of our proposed approach in responding to thermal stimuli, we designed a series of controlled experiments to simulate exposure to cold and hot environments. While applying personalized modeling showed an acceptable average accuracy of 91.3%, the generalized model’s average accuracy was only 65.2%. This shows the importance of having access to physiological records in modeling and assessing comfort. We also found that individual differences should be considered in selecting the cooling and heating rates when some knowledge of the occupant’s overall thermal preference is available.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Single-Walled Aluminosilicate Nanotube/Poly(vinyl alcohol) Nanocomposite Membranes

The fabrication, detailed characterization, and molecular transport properties of nanocomposite membranes containing high fractions (up to 40 vol %) of individually-dispersed aluminosilicate single-walled nanotubes (SWNTs) in poly(vinyl alcohol) (PVA), are reported. The microstructure, SWNT dispersion, SWNT dimensions, and intertubular distances within the composite membranes are characterized by scanning and transmission electron microscopy (SEM and TEM), energy-dispersive spectroscopy (EDS), X-ray diffraction (XRD), XRD rocking curve analysis, small-angle X-ray scattering (SAXS), and solid-state NMR. PVA/SWNT nanocomposite membranes prepared from SWNT gels allow uniform dispersion of individual SWNTs in the PVA matrix with a random distribution of orientations. SAXS analysis reveals the length ({approx}500 nm) and outer diameter ({approx}2.2 nm) of the dispersed SWNTs. Electron microscopy indicates good adhesion between the SWNTs and the PVA matrix without the occurrence of defects such as voids and pinholes. The transport properties of the PVA/SWNT membranes are investigated experimentally by ethanol/water mixture pervaporation measurements, computationally by grand canonical Monte Carlo and molecular dynamics, and by a macroscopic transport model for anisotropic permeation through nanotube-polymer composite membranes. The nanocomposite membranes substantially enhance the water throughput with increasing SWNT volume fraction, which leads to a moderate reduction of the water/ethanol selectivity. The model is parameterized purely from molecular simulation data with no fitted parameters, and shows reasonably good agreement with the experimental water permeability data.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Lattice Boltzmann model for compressible fluids

A lattice Boltzmann model is derived which simulates compressible fluids. By choosing the parameters of the equilibrium distribution appropriately, the sound speed (which may be set arbitrarily low), bulk viscosity, and kinematic viscosity can be selected. This model simulates compressible flows and can include shocks. With a proper rescaling and zero-sound speed, this model simulates Burgers's equation. The viscosity determined by a Chapman-Enskog expansion compares well with that measured form simulations. The exact solutions of Burgers's equation on the unit circle are compared to solutions of lattice Boltzmann model finding reasonable agreement.

Alexander, F. J.↗

Structural Analysis and Performance-Based Validation of a Composite Wing Spar

Electric-motor powered aircraft possess the ability to operate with efficient energy delivery, but lack the operational range of internal combustion engine powered aircraft. This range limitation requires the use of high aspect ratio, thin-chord wings to minimize aerodynamic drag losses, which results in highly loaded composite spar structures. High aspect ratio wings are required to increase mission durations for a NASA-developed experimental multi-rotor electric powered aircraft denoted as the Scalable Convergent Electric Propulsion Technology and Operations Research (SCEPTOR) or “X-57”. This paper examines the structural performance of the composite main wing spars to validate spar strength using ply-based laminate finite element methods. Geometric scaling of a main spar test-section was initially proposed for proof-testing but sacrificed stability. Ply-based structures modeling with local structural features was implemented as a risk-reduction methodology. Ply-based modeling was selected to augment the conventional “building block” approach to reduce risk, and leverage a “performance-based” approval processes encouraged in Federal Aviation Administration (FAA) design guidance. Therefore, ply-based laminate modeling of the full-scale main spar and forward spar shear-web attachments were subsequently undertaken to determine load path complexity with predicted flight loads. Ply-based modeling included stress concentrations and interlaminate behavior at interface locations that can be obscured in traditional finite element sizing models. Analysis of the wing spar laminate ply-based models compared with bearing test coupon performance was used to reduce future wing assembly proof-testing burden and facilitate performance-based flight hardware safety for the X-57 experimental aircraft.

Moore, James B.↗

Machine Learning Emulators and Empirical Models Combining Climate and Global Crop Models for Seasonal Agricultural Production

We present results from several connected efforts to apply machine learning methods to estimates of seasonal agricultural production anomalies around the world. First, we apply the XGBoost Random Forest method to fit emulators that mimic global crop models participating in the Agricultural Model Intercomparison and Improvement Project (AgMIP) Global Gridded Crop Model Intercomparison (GGCMI). These are the same models used in the agricultural sector simulations of the Inter-Sectoral Impacts Model Intercomparison Project (ISIMIP). These emulators use 8 climate variables split across 5 sub-seasonal representations of the growing season for each ½ degree grid cell around the world for maize, wheat, rice and soybeans. Emulators are useful for estimating conditions that have not already been simulated by GGCMI (e.g., in a seasonal prediction model) and also to diagnose model differences and capabilities. For example, emulators of the pDSSAT maize model tend to be more reliant on mean temperatures than the LPJmL model, and few models have strong responses to cold extremes. Second, we use a similar XGBoost approach to fit empirical models for national production data for the top 20 producing countries according to the United Nations Food and Agricultural Organization (FAO). Models utilize both climate observations and the GGCM models as predictors, resulting in skillful models for many (but not all) top producing-countries. The patterns of climate and crop model features selected indicate regions and systems that are better or worse simulated by the GGCMs. For example, information in cold extreme predictors is often combined with GGCM output predictors to provide sensitivity that models may underrepresent.

machine learning↗

Simulation Model of Microsphere Distribution for Selective Internal Radiation Therapy Agrees With Observations

Purpose: To perform a detailed analysis of microsphere distribution in biopsy material from a patient treated with {sup 90}Y-labeled resin spheres and characterize microsphere distribution in the hepatic artery tree, and to construct a novel dichotomous bifurcation model for microsphere deposits and evaluate its accuracy in simulating the observed microsphere deposits. Methods and Materials: Our virtual model consisted of arteries that successively branched into 2 new generations of arteries at 20 nodes. The artery diameter exponentially decreased from the lowest generation to the highest generation. Three variable parameters were optimized to obtain concordance between simulations and measure microsphere distributions: an artery coefficient of variation (ACV) for the diameter of all artery generations and the microsphere flow distribution at the nodes; a hepatic tree distribution volume (HDV) for the artery tree; and an artery diameter reduction (ADR) parameter. The model was tested against previously measured activity concentrations in 84 biopsies from the liver of 1 patient. In 16 of 84 biopsies, the microsphere distribution regarding cluster size and localization in the artery tree was determined via light microscopy of 30-μm sections (mean concentration, 14 microspheres/mg; distributions divided into 3 groups with mean microsphere concentrations of 4.6, 14, and 28 microspheres/mg). Results: Single spheres and small clusters were observed in terminal arterioles, whereas large clusters, up to 450 microspheres, were observed in larger arterioles. For 14 microspheres/mg, the optimized parameter values were ACV=0.35, HDV = 50 cm{sup 3}, and ADR=6 μm. For 4.6 microspheres/mg, ACV and ADR decreased to 0.26 and 0 μm, respectively, whereas HDV increased to 130 cm{sup 3}. The opposite trend was observed for 28 microspheres/mg: ACV = 0.49, HDV = 20 cm{sup 3}, and ADR = 8 μm. Conclusion: Simulations and measurements reveal that microsphere clusters are larger and more common in volumes with high microsphere concentrations and indicate that the spatial distribution of the artery tree must be considered in estimates of microsphere distributions.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Harnessing large language models’ zero-shot and few-shot learning capabilities for regulatory research

Abstract Large language models (LLMs) are sophisticated AI-driven models trained on vast sources of natural language data. They are adept at generating responses that closely mimic human conversational patterns. One of the most notable examples is OpenAI's ChatGPT, which has been extensively used across diverse sectors. Despite their flexibility, a significant challenge arises as most users must transmit their data to the servers of companies operating these models. Utilizing ChatGPT or similar models online may inadvertently expose sensitive information to the risk of data breaches. Therefore, implementing LLMs that are open source and smaller in scale within a secure local network becomes a crucial step for organizations where ensuring data privacy and protection has the highest priority, such as regulatory agencies. As a feasibility evaluation, we implemented a series of open-source LLMs within a regulatory agency’s local network and assessed their performance on specific tasks involving extracting relevant clinical pharmacology information from regulatory drug labels. Our research shows that some models work well in the context of few- or zero-shot learning, achieving performance comparable, or even better than, neural network models that needed thousands of training samples. One of the models was selected to address a real-world issue of finding intrinsic factors that affect drugs' clinical exposure without any training or fine-tuning. In a dataset of over 700 000 sentences, the model showed a 78.5% accuracy rate. Our work pointed to the possibility of implementing open-source LLMs within a secure local network and using these models to perform various natural language processing tasks when large numbers of training examples are unavailable.

Biochemistry & Molecular Biology↗

An integrated approach for determining plutonium mass in spent fuel assemblies with nondestructive assay

There are a variety of reasons for quantifying plutonium (Pu) in spent fuel. Below, five motivations are listed: (1) To verify the Pu content of spent fuel without depending on unverified information from the facility, as requested by the IAEA ('independent verification'). New spent fuel measurement techniques have the potential to allow the IAEA to recover continuity of knowledge and to better detect diversion. (2) To assure regulators that all of the nuclear material of interest leaving a nuclear facility actually arrives at another nuclear facility ('shipper/receiver'). Given the large stockpile of nuclear fuel at reactor sites around the world, it is clear that in the coming decades, spent fuel will need to be moved to either reprocessing facilities or storage sites. Safeguarding this transportation is of significant interest. (3) To quantify the Pu in spent fuel that is not considered 'self-protecting.' Fuel is considered self-protecting by some regulatory bodies when the dose that the fuel emits is above a given level. If the fuel is not self-protecting, then the Pu content of the fuel needs to be determined and the Pu mass recorded in the facility's accounting system. This subject area is of particular interest to facilities that have research-reactor spent fuel or old light-water reactor (LWR) fuel. It is also of interest to regulators considering changing the level at which fuel is considered self-protecting. (4) To determine the input accountability value at an electrochemical processing facility. It is not expected that an electrochemical reprocessing facility will have an input accountability tank, as is typical in an aqueous reprocessing facility. As such, one possible means of determining the input accountability value is to measure the Pu content in the spent fuel that arrives at the facility. (5) To fully understand the composition of the fuel in order to efficiently and safely pack spent fuel into a long-term repository. The NDA of spent fuel can be part of a system that cost-effectively meets the burnup credit needs of a repository. Behind each of these reasons is a regulatory structure with MC&A requirements. In the case of the IAEA, the accountable quantity is elemental plutonium. The material in spent fuel (fissile isotopes, fission products, etc.) emits signatures that provide information about the content and history of the fuel. A variety of nondestructive assay (NDA) techniques are available to quantify these signatures. The effort presented in this paper is investigation of the capabilities of 12 NDA techniques. For these 12, none is conceptually capable of independently determining the Pu content in a spent fuel assembly while at the same time being able to detect the diversion of a significant quantity of rods. For this reason the authors are investigating the capability of 12 NDA techniques with the end goal of integrating a few techniques together into a system that is capable of measuring Pu mass in an assembly. The work described here is the beginning of what is anticipated to be a five year effort: (1) two years of modeling to select the best technologies, (2) one year fabricating instruments and (3) two years measuring spent fuel. This paper describes the first two years of this work. In order to cost effectively and robustly model the performance of the 12 NDA techniques, an 'assembly library' was created. The library contains the following: (a) A diverse range of PWR spent fuel assemblies (burnup, enrichment, cooling time) similar to that which exists in spent pools today and in the future. (b) Diversion scenarios that capture a range of possible rod removal options. (c) The spatial and isotopic detail needed to accurately quantify the capability of all the NDA techniques so as to enable integration. It is our intention to make this library available to other researchers in the field for inter-comparison purposes. The performance of each instrument will be quantified for the full assembly library for measurements in three different media: air, water and borated water. The 12 NDA techniques being researched are the following: Delayed Gamma, Delayed Neutrons, Differential Die-Away, Lead Slowing Down Spectrometer, Neutron Multiplicity, Nuclear Resonance Fluorescence, Passive Prompt Gamma, Passive Neutron Albedo Reactivity, Self-integration Neutron Resonance Densitometry, Total Neutron (Gross Neutron), X-Ray Fluorescence, {sup 252}Cf Interrogation with Prompt Neutron Detection.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

A mathematical approach to survey electrochemical impedance spectroscopy for aging in lithium-ion batteries

Electrochemical impedance spectroscopy (EIS) is a valuable technique to detect the health status and aging phenomena in lithium-ion batteries (LiB). Equivalent circuit modeling (ECM) is conventionally used when interpreting EIS data and gaining physical insights into the aging mechanisms. However, performing ECM is resource intensive and expert-level of knowledge is usually required to select suitable models and fitting parameters. This article presents a quick and user-friendly data analysis algorithm as an alternative to ECM by mathematically fitting geometric features in Nyquist plots and obtaining the growth trends of the features. The evolving trends in the Nyquist plots, such as chord lengths of the arcs and interception points, are consistent with the growth of resistance components obtained using ECM with R 2 values from 0.67 to 0.99, and therefore can be used as indicators of battery aging. Our results show that the quick-fitting approach is suitable for analyzing a series of EIS data acquired during battery cycling and identifying the underlying aging mechanisms.

Chen, Bor-Rong↗

Evaluating the strength of the land$-$atmosphere moisture feedback in Earth system models using satellite observations

The relationship between terrestrial water storage (TWS) and atmospheric processes has important implications for predictability of climatic extremes and projection of future climate change. In places where moisture availability limits evapotranspiration (ET), variability in TWS has the potential to influence surface energy fluxes and atmospheric conditions. Where atmospheric conditions, in turn, influence moisture availability, a full feedback loop exists. Here we developed a novel approach for measuring the strength of both components of this feedback loop, i.e., the forcing of the atmosphere by variability in TWS and the response of TWS to atmospheric variability, using satellite observations of TWS, precipitation, solar radiation, and vapor pressure deficit during 2002–2014. Our approach defines metrics to quantify the relationship between TWS anomalies and climate globally on a seasonal to interannual timescale. Metrics derived from the satellite data were used to evaluate the strength of the feedback loop in 38 members of the Community Earth System Model (CESM) Large Ensemble (LENS) and in six models that contributed simulations to phase 5 of the Coupled Model Intercomparison Project (CMIP5). We found that both forcing and response limbs of the feedback loop in LENS were stronger than in the satellite observations in tropical and temperate regions. Feedbacks in the selected CMIP5 models were not as strong as those found in LENS, but were still generally stronger than those estimated from the satellite measurements. Consistent with previous studies conducted across different spatial and temporal scales, our analysis suggests that models may overestimate the strength of the feedbacks between the land surface and the atmosphere. Lastly, we describe several possible mechanisms that may contribute to this bias, and discuss pathways through which models may overestimate ET or overestimate the sensitivity of ET to TWS.

54 ENVIRONMENTAL SCIENCES↗

ABF Past Accomplishments - TEA/LCA

The Integrated Analysis task in the Agile BioFoundry (ABF) conducts techno-economic analysis (TEA) and life cycle assessment (LCA) to quantify the economics and environmental impacts of bioprocesses under development. The team from Argonne National Laboratory (ANL) and the National Renewable Energy Laboratory (NREL) develops TEA and LCA models of selected compound targets and hosts of interest to the ABF in an effort to provide an analysis-based foundation to the R&D. To date, the Integrated Analysis task has performed process modeling and analyses on multiple pairings of bioproducts and hosts, centered around major metabolic "beachheads" and "exemplar" molecules chosen for each of the metabolic pathways of interest. This presentation provides an overview of the approach used for TEA and LCA in the ABF and of the progress, outcomes, and impacts of TEA and LCA to support ABF goals, with a focus towards developing bio-based products that are both environmentally and economically viable.

BIOMASS FUELS,INORGANIC, ORGANIC, PHYSICAL, AND AN↗

Nanoparticle Injection Technology for Remediating Leaks of CO 2 Storage Formation (Final Report)

Leakage of a well can occur through the wellbore, the annulus between well tubing and steel casing, or on the outside of the casing. The overall goal of the project is to develop a new technology that can repair or reduce the leakage of wellbore. The new technology is based on simultaneously injecting nanoparticles electrochemically and removing some of the harmful ions in the system so the distressed cement materials can be densified and the corrosion risk of steel casing can be reduced. There were two objectives. One is the development of the injection technology for leakage repair and the other is the development of new numerical simulation models that can simulate and predict the performance of the new repair technology. An electro-migration testing unit and a in-house wellbore testing system were developed; healing agents with nanoparticles were selected, numerical models were developed to simulate the electro-migration technology, and the effectiveness of the new technology were evaluated.

42 ENGINEERING↗