Results for “prediction.” — Page 8

NASA NTRS2001

Analysis of Measured and Predicted Acoustics from an XV-15 Flight Test

Flight acoustic and vehicle state data from an XV-15 acoustic flight test are examined. Flight predictions using TRAC are performed for a level flight (repeated) and four descent conditions (including a BVI). The assumptions and procedures used for TRAC flight predictions as well as the variability in flight measurements, which are used for input and comparison to predictions, are investigated in detail. Differences were found in the measured vehicle airspeed, altitude, glideslope, and vehicle orientation (yaw, pitch and roll angle) between each of the repeat runs. These differences violate some of the prediction assumptions and significantly impacted the resulting acoustic predictions. Multiple acoustic pulses, with a variable time between the pulses, were found in the measured acoustic time histories for the repeat runs. These differences could be attributed in part to the variability in vehicle orientation. Acoustic predictions that used the measured vehicle orientation for the repeat runs captured this multiple pulse variability. Thickness noise was found to be dominant on approach for all the cases, except the BVI condition. After the aircraft passed overhead, broadband noise and low frequency loading noise were dominant. The predicted LowSPL time histories compared well with measurement on approach to the array for the non-BVI conditions and poorly for the BVI condition. Accurate prediction of the lift share between the rotor and fuselage must be known in order to improve predictions. At a minimum, measurements of the rotor thrust and tip-path-plane angle are critical to further develop accurate flight acoustic prediction capabilities.

Boyd, D. Douglas, Jr.↗

NASA NTRS2007

Prediction of Geomagnetic Activity and Key Parameters in High-latitude Ionosphere

Prediction of geomagnetic activity and related events in the Earth's magnetosphere and ionosphere are important tasks of US Space Weather Program. Prediction reliability is dependent on the prediction method, and elements included in the prediction scheme. Two of the main elements of such prediction scheme are: an appropriate geomagnetic activity index, and an appropriate coupling function (the combination of solar wind parameters providing the best correlation between upstream solar wind data and geomagnetic activity). We have developed a new index of geomagnetic activity, the Polar Magnetic (PM) index and an improved version of solar wind coupling function. PM index is similar to the existing polar cap PC index but it shows much better correlation with upstream solar wind/IMF data and other events in the magnetosphere and ionosphere. We investigate the correlation of PM index with upstream solar wind/IMF data for 10 years (1995-2004) that include both low and high solar activity. We also have introduced a new prediction function for the predicting of cross-polar-cap voltage and Joule heating based on using both PM index and upstream solar wind/IMF data. As we show such prediction function significantly increase the reliability of prediction of these important parameters. The correlation coefficients between the actual and predicted values of these parameters are approx. 0.9 and higher.

Khazanov, George V.↗

NASA NTRS2008

Uncertainty and Sensitivity Analyses of a Two-Parameter Impedance Prediction Model

This paper presents comparisons of predicted impedance uncertainty limits derived from Monte-Carlo-type simulations with a Two-Parameter (TP) impedance prediction model and measured impedance uncertainty limits based on multiple tests acquired in NASA Langley test rigs. These predicted and measured impedance uncertainty limits are used to evaluate the effects of simultaneous randomization of each input parameter for the impedance prediction and measurement processes. A sensitivity analysis is then used to further evaluate the TP prediction model by varying its input parameters on an individual basis. The variation imposed on the input parameters is based on measurements conducted with multiple tests in the NASA Langley normal incidence and grazing incidence impedance tubes; thus, the input parameters are assigned uncertainties commensurate with those of the measured data. These same measured data are used with the NASA Langley impedance measurement (eduction) processes to determine the corresponding measured impedance uncertainty limits, such that the predicted and measured impedance uncertainty limits (95% confidence intervals) can be compared. The measured reactance 95% confidence intervals encompass the corresponding predicted reactance confidence intervals over the frequency range of interest. The same is true for the confidence intervals of the measured and predicted resistance at near-resonance frequencies, but the predicted resistance confidence intervals are lower than the measured resistance confidence intervals (no overlap) at frequencies away from resonance. A sensitivity analysis indicates the discharge coefficient uncertainty is the major contributor to uncertainty in the predicted impedances for the perforate-over-honeycomb liner used in this study. This insight regarding the relative importance of each input parameter will be used to guide the design of experiments with test rigs currently being brought on-line at NASA Langley.

Jones, M. G.↗

NASA NTRS2008

Predictive Information: Status or Alert Information?

Previous research investigating the efficacy of predictive information for detecting and diagnosing aircraft system failures found that subjects like to have predictive information concerning when a parameter would reach an alert range. This research focused on where the predictive information should be located, whether the information should be more closely associated with the parameter information or with the alert information. Each subject saw 3 forms of predictive information: (1) none, (2) a predictive alert message, and (3) predictive information on the status display. Generally, subjects performed better and preferred to have predictive information available although the difference between status and alert predictive information was minimal. Overall, for detection and recalling what happened, status predictive information is best; however for diagnosis, alert predictive information holds a slight edge.

Trujillo, Anna C.↗

NASA NTRS2021

A Recursive Multi-step Machine Learning Approach for Airport Configuration Prediction

Airport configuration selection is a complex decision-making process that involves several operational and human factors. In this paper we propose a novel recursive multi-step machine learning (ML) approach to predict airport configuration. The multi-step approach guarantees stability of the predicted configuration by taking as input the configuration predicted at the previous time step. The features of the proposed model include weather data, future arrival and departure counts and current configuration. Due to the importance of arrival and departure counts in predicting the airport configuration, arrival counts are calculated using landing time predictions selected from physics-based landing time predictions available in FAA System Wide Information Management data feeds for each flight. The selection rules were developed and refined to select the most accurate time for different phases of flight. The proposed model predicts the airport configurations up to 6 hours ahead. In this paper we show the predictive performance of the proposed model for six major US airports, including Charlotte Douglas International Airport (CLT), Dallas/Fort Worth International Airport (DFW), John F. Kennedy International Airport (JFK), Newark Liberty International Airport (EWR), LaGuardia Airport (LGA) and Dallas Love Field Airport (DAL). We trained and evaluated models on 2019 and 2020 data in order to study the effect of the pandemic and how changes in traffic patterns affected the performance of the proposed model. Results are compared with a baseline assuming no airport configuration changes. In our results for DFW, we obtained a prediction accuracy of 89.3% for 3 hours ahead prediction, and 82.8% for 6 hours ahead when applied on 2019 data.

machine learning↗

NASA NTRS2021

A Recursive Multi-step Machine Learning Approach for Airport Configuration Prediction

Airport configuration selection is a complex decision-making process that involves several operational and human factors. In this paper we propose a novel recursive multi-step machine learning (ML) approach to predict airport configuration. The multi-step approach guarantees stability of the predicted configuration by taking as input the configuration predicted at the previous time step. The features of the proposed model include weather data, future arrival and departure counts and current configuration. Due to the importance of arrival and departure counts in predicting the airport configuration, arrival counts are calculated using landing time predictions selected from physics-based landing time predictions available in FAA System Wide Information Management data feeds for each flight. The selection rules were developed and refined to select the most accurate time for different phases of flight. The proposed model predicts the airport configurations up to 6 hours ahead. In this paper we show the predictive performance of the proposed model for six major US airports, including Charlotte Douglas International Airport (CLT), Dallas/Fort Worth International Airport (DFW), John F. Kennedy International Airport (JFK), Newark Liberty International Airport (EWR), LaGuardia Airport (LGA) and Dallas Love Field Airport (DAL). We trained and evaluated models on 2019 and 2020 data in order to study the effect of the pandemic and how changes in traffic patterns affected the performance of the proposed model. Results are compared with a baseline assuming no airport configuration changes. In our results for DFW, we obtained a prediction accuracy of 89.3% for 3 hours ahead prediction, and 82.8% for 6 hours ahead when applied on 2019 data.

machine learning↗

NASA NTRSDate not supplied

Assimilation of SMAP Observations Over Land Improves the Simulation and Prediction of Tropical Cyclone Idai

This work is focused on the role of soil moisture in the prediction of tropical cyclones (TCs) approaching land and after landfall. Soil moisture conditions can impact the circulation and structure of an existing tropical cyclone (TC) when part or all of the circulation is over land. For example, dry land surface conditions may lead to faster dissipation of a TC over land (often associated with changes in precipitation structure), whereas very wet conditions may help sustain or in rare cases re-intensify a TC. Moreover, the presence of strong soil moisture gradients may affect the symmetry and development of the TC circulation leading to changes in its over-land track. While the link between soil moisture conditions and TC evolution in proximity to land is relatively well understood in theory, applications of these findings in the context of numerical weather prediction (NWP) have been limited. Here we present a case study that explores the potential of improving TC predictions through an improved soil moisture initialization in an NWP framework. Specifically, we examine the impact of assimilating observations from the NASA Soil Moisture Active Passive (SMAP) mission into the NASA Goddard Earth Observing System (GEOS) global weather model on the prediction of South-West Indian Ocean TC Idai (2019). SMAP provides accurate L-band (1.4 GHz) brightness temperatures (Tb) observations that are sensitive to soil moisture globally and at high revisit times of 2-3 days. It has previously been shown that the assimilation of SMAP Tb observations significantly improves modeled land surface states. Thus, it is expected that SMAP can be used to constrain land surface initial conditions and potentially benefit TC forecasts. Here we present two sets of retrospective forecasts of TC Idai that are compared in an Observing System Experiment framework at ¼ degree resolution: (i) forecasts initialized from an analysis that is comparable to the GEOS operational analysis (without SMAP Tb assimilation) and (ii) forecasts initialized from an analysis that additionally assimilates SMAP brightness temperature observations over land using a weakly-coupled land analysis. We find that the assimilation of SMAP meaningfully improves the representation of TC Idai’s structure as well as the prediction of its intensity and track. The analyzed TC size, as measured by the wind speed radius, is improved by up to 18% in the analysis with SMAP assimilation relative to the control run. The forecast intensity error, measured against the observed intensity, is reduced by up to 23%. At the 1/4-degree resolution used here, GEOS unavoidably under-estimates TC intensity and over-estimates TC size. The SMAP assimilation therefore corrects the model in the right direction, leading to a storm that is more energetic and more compact. Furthermore, we find that the along-track forecast error is reduced by up to 34%, indicating a more accurate propagation speed, which is consistent with the fact that TC speed over land is strongly affected by surface processes. The impact of SMAP assimilation on the forecast cross-track error is neutral. Across the TC forecast skill metrics used here, the improvements from SMAP DA are largest at lead times of 36 to 72 hours, suggesting that the predictability of forecasts at shorter lead times may be dominated by short-term convective processes, while the land and its longer memory gains in importance as a source of predictability on a 2-3 day timescale. We further investigated the underlying mechanisms leading to the skill improvements from SMAP data assimilation by isolating the land areas that directly influence TC Idai using a back trajectory analysis. We find that the assimilation of SMAP leads to wetter soil moisture conditions that cause an increased latent heat flux, which ultimately results in TC analyzed representation that has higher column-integrated total moisture content and total energy compared to the analysis in the control run without SMAP assimilation. Overall, the results highlight that the assimilation of SMAP observations into a global numerical weather prediction model can lead to pronounced improvements of TC predictions. This is a crucial step towards a better mitigation of the socio-economic impact of landfalling TCs and thus safeguarding human lives. Finally, our study presents an event-based approach that assesses the impact of land data assimilation for a particular weather event rather than by globally averaging differences in skill. We argue that global skill assessments – while necessary – can mute the impact of land data assimilation, because the land’s influence on the atmosphere is constrained to certain locations and certain times. Instead, the event-based approach better highlights the true potential of land data assimilation in the context of NWP, especially for extreme events when accurate predictions are critical.

Jana Kolassa↗

NASA NTRSDate not supplied

Assimilation of SMAP Observations Over Land Improves the Simulation and Prediction of Tropical Cyclone Idai

This work is focused on the role of soil moisture in the prediction of tropical cyclones (TCs) approaching land and after landfall. Soil moisture conditions can impact the circulation and structure of an existing tropical cyclone (TC) when part or all of the circulation is over land. For example, dry land surface conditions may lead to faster dissipation of a TC over land (often associated with changes in precipitation structure), whereas very wet conditions may help sustain or in rare cases re-intensify a TC. Moreover, the presence of strong soil moisture gradients may affect the symmetry and development of the TC circulation leading to changes in its over-land track. While the link between soil moisture conditions and TC evolution in proximity to land is relatively well understood in theory, applications of these findings in the context of numerical weather prediction (NWP) have been limited. Here we present a case study that explores the potential of improving TC predictions through an improved soil moisture initialization in an NWP framework. Specifically, we examine the impact of assimilating observations from the NASA Soil Moisture Active Passive (SMAP) mission into the NASA Goddard Earth Observing System (GEOS) global weather model on the prediction of South-West Indian Ocean TC Idai (2019). SMAP provides accurate L-band (1.4 GHz) brightness temperatures (Tb) observations that are sensitive to soil moisture globally and at high revisit times of 2-3 days. It has previously been shown that the assimilation of SMAP Tb observations significantly improves modeled land surface states. Thus, it is expected that SMAP can be used to constrain land surface initial conditions and potentially benefit TC forecasts. Here we present two sets of retrospective forecasts of TC Idai that are compared in an Observing System Experiment framework at ¼ degree resolution: (i) forecasts initialized from an analysis that is comparable to the GEOS operational analysis (without SMAP Tb assimilation) and (ii) forecasts initialized from an analysis that additionally assimilates SMAP brightness temperature observations over land using a weakly-coupled land analysis. We find that the assimilation of SMAP meaningfully improves the representation of TC Idai’s structure as well as the prediction of its intensity and track. The analyzed TC size, as measured by the wind speed radius, is improved by up to 18% in the analysis with SMAP assimilation relative to the control run. The forecast intensity error, measured against the observed intensity, is reduced by up to 23%. At the 1/4-degree resolution used here, GEOS unavoidably under-estimates TC intensity and over-estimates TC size. The SMAP assimilation therefore corrects the model in the right direction, leading to a storm that is more energetic and more compact. Furthermore, we find that the along-track forecast error is reduced by up to 34%, indicating a more accurate propagation speed, which is consistent with the fact that TC speed over land is strongly affected by surface processes. The impact of SMAP assimilation on the forecast cross-track error is neutral. Across the TC forecast skill metrics used here, the improvements from SMAP DA are largest at lead times of 36 to 72 hours, suggesting that the predictability of forecasts at shorter lead times may be dominated by short-term convective processes, while the land and its longer memory gains in importance as a source of predictability on a 2-3 day timescale. We further investigated the underlying mechanisms leading to the skill improvements from SMAP data assimilation by isolating the land areas that directly influence TC Idai using a back trajectory analysis. We find that the assimilation of SMAP leads to wetter soil moisture conditions that cause an increased latent heat flux, which ultimately results in TC analyzed representation that has higher column-integrated total moisture content and total energy compared to the analysis in the control run without SMAP assimilation. Overall, the results highlight that the assimilation of SMAP observations into a global numerical weather prediction model can lead to pronounced improvements of TC predictions. This is a crucial step towards a better mitigation of the socio-economic impact of landfalling TCs and thus safeguarding human lives. Finally, our study presents an event-based approach that assesses the impact of land data assimilation for a particular weather event rather than by globally averaging differences in skill. We argue that global skill assessments – while necessary – can mute the impact of land data assimilation, because the land’s influence on the atmosphere is constrained to certain locations and certain times. Instead, the event-based approach better highlights the true potential of land data assimilation in the context of NWP, especially for extreme events when accurate predictions are critical.

Jana Kolassa↗

NASA NTRS2023

Comparing Magnetopause Predictions From Two MHD Models During A Geomagnetic Storm and A Quiet Period

Magnetopause location is an important prediction of numerical simulations of the magnetosphere, yet the models can err, either under-predicting or over-predicting the motion of the boundary. This study compares results from two of the most widely used magnetohydrodynamic (MHD) models, the Lyon–Fedder–Mobarry (LFM) model and the Space Weather Modeling Framework (SWMF), to data from the GOES 13 and 15 satellites during the geomagnetic storm on 22 June 2015, and to THEMIS A, D, and E during a quiet period on 31 January 2013. The models not only reproduce the magnetopause crossings of the spacecraft during the storm, but they also predict spurious magnetopause motion after the crossings seen in the GOES data. We investigate the possible causes of the over-predictions during the storm and find the following. First, using different ionospheric conductance models does not significantly alter predictions of the magnetopause location. Second, coupling the Rice Convection Model (RCM) to the MHD codes improves the SWMF magnetopause predictions more than it does for the LFM predictions. Third, the SWMF produces a stronger ring current than LFM, both with and without the RCM and regardless of the LFM spatial resolution. During the non-storm event, LFM predicts the THEMIS magnetopause crossings due to the southward interplanetary magnetic field better than the SWMF. Additionally, increasing the LFM spatial grid resolution improves the THEMIS predictions, while increasing the SWMF grid resolutions does not.

magnetohydrodynamics↗

DOE OSTI2024

Machine Learning Framework for Conotoxin Class and Molecular Target Prediction

Conotoxins are small and highly potent neurotoxic peptides derived from the venom of marine cone snails which have captured the interest of the scientific community due to their pharmacological potential. These toxins display significant sequence and structure diversity, which results in a wide range of specificities for several different ion channels and receptors. Despite the recognized importance of these compounds, our ability to determine their binding targets and toxicities remains a significant challenge. Predicting the target receptors of conotoxins, based solely on their amino acid sequence, remains a challenge due to the intricate relationships between structure, function, target specificity, and the significant conformational heterogeneity observed in conotoxins with the same primary sequence. We have previously demonstrated that the inclusion of post-translational modifications, collisional cross sections values, and other structural features, when added to the standard primary sequence features, improves the prediction accuracy of conotoxins against non-toxic and other toxic peptides across varied datasets and several different commonly used machine learning classifiers. Here, we present the effects of these features on conotoxin class and molecular target predictions, in particular, predicting conotoxins that bind to nicotinic acetylcholine receptors (nAChRs). We also demonstrate the use of the Synthetic Minority Oversampling Technique (SMOTE)-Tomek in balancing the datasets while simultaneously making the different classes more distinct by reducing the number of ambiguous samples which nearly overlap between the classes. In predicting the alpha, mu, and omega conotoxin classes, the SMOTE-Tomek PCA PLR model, using the combination of the SS and P feature sets establishes the best performance with an overall accuracy (OA) of 95.95%, with an average accuracy (AA) of 93.04%, and an f1 score of 0.959. Using this model, we obtained sensitivities of 98.98%, 89.66%, and 90.48% when predicting alpha, mu, and omega conotoxin classes, respectively. Similarly, in predicting conotoxins that bind to nAChRs, the SMOTE-Tomek PCA SVM model, which used the collisional cross sections (CCSs) and the P feature sets, demonstrated the highest performance with 91.3% OA, 91.32% AA, and an f1 score of 0.9131. The sensitivity when predicting conotoxins that bind to nAChRs is 91.46% with a 91.18% sensitivity when predicting conotoxins that do not bind to nAChRs.

59 BASIC BIOLOGICAL SCIENCES↗

NASA NTRS2016

Rolling Bearing Life Prediction, Theory, and Application

A tutorial is presented outlining the evolution, theory, and application of rolling-element bearing life prediction from that of A. Palmgren, 1924; W. Weibull, 1939; G. Lundberg and A. Palmgren, 1947 and 1952; E. Ioannides and T. Harris, 1985; and E. Zaretsky, 1987. Comparisons are made between these life models. The Ioannides-Harris model without a fatigue limit is identical to the Lundberg-Palmgren model. The Weibull model is similar to that of Zaretsky if the exponents are chosen to be identical. Both the load-life and Hertz stress-life relations of Weibull, Lundberg and Palmgren, and Ioannides and Harris reflect a strong dependence on the Weibull slope. The Zaretsky model decouples the dependence of the critical shear stress-life relation from the Weibull slope. This results in a nominal variation of the Hertz stress-life exponent. For 9th- and 8th-power Hertz stress-life exponents for ball and roller bearings, respectively, the Lundberg-Palmgren model best predicts life. However, for 12th- and 10th-power relations reflected by modern bearing steels, the Zaretsky model based on the Weibull equation is superior. Under the range of stresses examined, the use of a fatigue limit would suggest that (for most operating conditions under which a rolling-element bearing will operate) the bearing will not fail from classical rolling-element fatigue. Realistically, this is not the case. The use of a fatigue limit will significantly overpredict life over a range of normal operating Hertz stresses. (The use of ISO 281:2007 with a fatigue limit in these calculations would result in a bearing life approaching infinity.) Since the predicted lives of rolling-element bearings are high, the problem can become one of undersizing a bearing for a particular application. Rules had been developed to distinguish and compare predicted lives with those actually obtained. Based upon field and test results of 51 ball and roller bearing sets, 98 percent of these bearing sets had acceptable life results using the Lundberg- Palmgren equations with life adjustment factors to predict bearing life. That is, they had lives equal to or greater than that predicted. The Lundberg-Palmgren model was used to predict the life of a commercial turboprop gearbox. The life prediction was compared with the field lives of 64 gearboxes. From these results, the roller bearing lives exhibited a load-life exponent of 5.2, which correlated with the Zaretsky model. The use of the ANSI/ABMA and ISO standards load-life exponent of 10/3 to predict roller bearing life is not reflective of modern roller bearings and will underpredict bearing lives.

Life prediction↗

NASA NTRS2016

Predicting Real-Time Safety of the National Airspace System

To gain the situational awareness necessary for informed decision making regarding avoidance of airspace hazards, each operator must consolidate operations-relevant information from disparate sources and apply extensive domain knowledge to correctly interpret not just the current state of the NAS but forecast its (combined) evolution over the duration of the operation. This time- and workload-intensive process is periodically repeated throughout the operation so that changes can be managed in a timely manner.The imprecision, inaccuracies, inconsistency, and incompleteness of the incoming data further challenges the process. To facilitate informed decision making, this paper presents a model-based framework for the textitautomated real-time monitoring and prediction of possible effects of airspace hazards on the safety of the National Airspace System (NAS). First, hazards to flight are identified and transformed into sms, that is, quantities of interest that could be evaluated based on available data and are predictive of an unsafe event. The sms and associated thresholds that specify when an event transitions from emphsafe to emphunsafe are combined with models of airspace operations and aircraft dynamics. The framework can include any hazard to flight that can be modeled quantitatively. Models can be detailed and complex, or they can be considerably simplifed, as appropriate to the application. Real-time NAS safety monitoring and prediction begins with an estimate of the state of the NAS using the dynamic models. Given the state estimate and a probability distribution of future inputs to the NAS, we can then predict the evolution of the NAS - the future state - and the occurrence of hazards and unsafe events. The entire probability distribution of airspace sms is computed, not just point estimates, without significant assumptions regarding the distribution type andor parameters. We demonstrate our overall approach through a simulated scenario in which we predict the occurrence of some unsafe events and show how these predictions evolve in time as flight operations progress. Predictions accounting for common sources of uncertainty are included and it is shown how the predictions improve in time, become more confident, and change dynamically as new information is made available to the prediction algorithm.

safety↗

DOE OSTI2026

Multi-trait multi-environment genomic prediction strategies for Miscanthus sacchariflorus

Genomic selection holds the potential to serve as a strategic tool to enhance the genetic gain of complex traits in Miscanthus breeding programs. The development of improved cultivars requires their assessment for various traits across diverse environments to ensure suitable overall performance. Hence, the multi-trait multi-environment (MTME) genomic prediction (GP) models offer an opportunity to improve selection accuracy. This study aims to evaluate the potential of five GP models: (1) three MTME models including genotype-by-trait-by-environment interaction (G×E×T) and (2) two single-trait multi-environment (STME) models (with and without G×E interaction). A Miscanthus sacchariflorus population comprising 336 genotypes evaluated in three environments and scored for four traits (biomass yield YDY, total culm number TCM, average internode length AIL, and culm node number CNN) was analyzed. The predictive ability of the models was evaluated considering three cross-validation schemes resembling realistic scenarios (CV1: predicting new genotypes, CVP: predicting missing traits in a given environment, and CV2: predicting partially observed genotypes). On average, in all cross-validation schemes compared to the STME the predictive ability of the MTME models was 10% to 70% higher for TCM and AIL. On the other hand, for YDY and CNN, both STME models performed similarly or slightly better (between 5 to 64%) than the MTME models in most environments. While the MTME models were not successful for all traits when compared to their STME counterparts, MTME models improved the prediction of the performance of genotypes that were untested across environments or lacked trait information in a specific environment. Overall, our study suggests that MTME GP models can be implemented in Miscanthus breeding programs to improve the predictive ability of the complex traits, shorten breeding cycles, and accelerate selection decisions.

genomic prediction (GP)↗

DOE OSTI2026

Evaluation of methods and improvement of predictions for specification properties of petroleum-based and alternative aviation fuels

To support our research and process modeling for liquid fuels, including blends, from petroleum and synthetic sources such as from biomass intermediates, we evaluated composition-based prediction methods and improved predictions for five key specification properties of petroleum-based and alternative aviation fuels, namely distillation temperatures (10 % distilled, t 10 , and final boiling point, t FBP ), density, flash point, net heat of combustion, and freezing point. The types of fuels included were petroleum-based jet fuels, jet-fuel surrogate mixtures, synthetic blending components obtained from different sources, and blends of Jet A with many synthetic blending components. Expanded datasets to update associated parameters allowed significant improvements for one of the prediction methods used in earlier work, namely the Modified Weighted Average method published initially by Shi et al. By considering the importance of lighter compounds for flash points and heavier compounds for freezing points, the revised Modified Weighted Average method was further improved. For liquid density, the revised Modified Weighted Average method gave the best overall results. The revised Modified Weighted Average method, the American Society for Testing and Materials D7215 method, and the D7215 method modified by another group gave comparable results for flash point, while the revised Modified Weighted Average and D3338 methods gave the best results for net heat of combustion. Freezing point was well predicted using the revised Modified Weighted Average method and showed the most significant improvements over current predictions. Distillation temperature t 10 was not well predicted, while t FBP was predicted with a mean absolute error comparable to experimental reproducibility.

09 BIOMASS FUELS↗

DOE OSTI2024

Quantifying sources of subseasonal prediction skill in CESM2

Abstract Subseasonal prediction fills the gap between weather forecasts and seasonal outlooks. There is evidence that predictability on subseasonal timescales comes from a combination of atmosphere, land, and ocean initial conditions. Predictability from the land is often attributed to slowly varying changes in soil moisture and snowpack, while predictability from the ocean is attributed to sources such as the El Niño Southern Oscillation. Here we use a set of subseasonal reforecast experiments with CESM2 to quantify the respective roles of atmosphere, land, and ocean initial conditions on subseasonal prediction skill over land. These reveal that the majority of prediction skill for global surface temperature in weeks 3–4 comes from the atmosphere, while ocean initial conditions become important after week 4, especially in the Tropics. In the CESM2 subseasonal prediction system, the land initial state does not contribute to surface temperature prediction skill in weeks 3–6 and climatological land conditions lead to higher skill, disagreeing with our current understanding. However, land-atmosphere coupling is important in week 1. Subseasonal precipitation prediction skill also comes primarily from the atmospheric initial condition, except for the Tropics, where after week 4 the ocean state is more important.

Richter, Jadwiga H. (ORCID:0000000170480781)↗

DOE OSTI2024

Data-driven prediction of scaling and ignition of inertial confinement fusion experiments

Recent advances in inertial confinement fusion (ICF) at the National Ignition Facility (NIF), including ignition and energy gain, are enabled by a close coupling between experiments and high-fidelity simulations. Neither simulations nor experiments can fully constrain the behavior of ICF implosions on their own, meaning pre- and postshot simulation studies must incorporate experimental data to be reliable. Linking past data with simulations to make predictions for upcoming designs and quantifying the uncertainty in those predictions has been an ongoing challenge in ICF research. We have developed a data-driven approach to prediction and uncertainty quantification that combines large ensembles of simulations with Bayesian inference and deep learning. The approach builds a predictive model for the statistical distribution of key performance parameters, which is jointly informed by past experiments and physics simulations. The prediction distribution captures the impact of experimental uncertainty, expert priors, design changes, and shot-to-shot variations. We have used this new capability to predict a 10× increase in ignition probability between Hybrid-E shots driven with 2.05 MJ compared to 1.9 MJ, and validated our predictions against subsequent experiments. We describe our new Bayesian postshot and prediction capabilities, discuss their application to NIF ignition and validate the results, and finally investigate the impact of data sparsity on our prediction results.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2025

GRUMDN: A Multi-Task Model for Predicting Human Patterns-of-Life from Stay Transition Data

Understanding human patterns-of-life (PoL) is essential towards ensuring safe and secure indoor facility environment as well as outdoor urban environment. Prediction of human movement in between places of interest is vital in understanding human PoL. Movement between spaces maybe represented and detected in one of the two forms: 1) trajectories: locations measured at regular time intervals by mobile sensors, bluetooth or GPS sensors; or 2) stay transitions: semantic PoI (points of interest) and stay duration data measurable by eventbased sensors that collect data when a check-in or check-out event is detected. Stay transition data provides a more compressed data format compared to trajectories data, especially in situations with longer stay durations, while preserving the information necessary for PoL analysis. Now as introduced briefly in the paper, our deployed end application (Digital Twin of a facility with non-player characters, besides the interactive user in virtual reality) needed a well-performing and validated AI/ML model for simulating high quality stay transitions behavior. In this study we thus primarily present our findings with developing and validating that model, which is a multi-task neural network for stay transition prediction. The neural network consists of two heads, for corresponding two tasks of stay category prediction and stay duration prediction. We evaluated gated recurrent units and multi-layer perceptrons of varying network sizes for stay category prediction; while mixture density networks, noisy generator-only networks, and generative adversarial networks of varying network sizes for stay duration prediction. We have then evaluated four multi-task models, constructed by combining these specialized models, on their ability to predict stay transition data. We tested our models on datasets from two different cases: 1) a simulation-generated dataset of indoor movement within the HFIR (high flux isotope reactor) nuclear reactor facility at Oak Ridge National Laboratory (ORNL); and 2) the GeoLife human mobility dataset of outdoor urban movement available in literature. Our results indicate that GRUMDN, which combines gated recurrent units (GRU) for stay category prediction task, and mixture density networks (MDN) for stay duration prediction task, did overall outperform other multitask models and the current state-of-the-art.

Gunaratne, Chathika [ORNL] (ORCID:0000000225088745↗

DOE OSTI2025

A Digital Twin Framework Utilizing Machine Learning for Robust Predictive Maintenance: Enhancing Tire Health Monitoring

We introduce a novel digital twin (DT) framework for the predictive maintenance of long-term physical systems. Using monitoring tire health as an application, we show how the DT framework can be used to enhance automotive safety and efficiency, and how the technical challenges can be overcome using a three-step approach. First, to manage the data complexity over a long operation span, we employ data reduction techniques to concisely represent physical tires using historical performance and usage data. Relying on these data, for fast real-time prediction, we train a transformer-based model offline on our concise dataset to predict future tire health over time, represented as remaining casing potential (RCP). Based on our architecture, our model quantifies both epistemic and aleatoric uncertainties, providing reliable confidence intervals around predicted RCP. Second, to incorporate real-time data, we update the predictive model in the DT framework, ensuring its accuracy throughout its lifespan with the aid of hybrid modeling and the use of the discrepancy function. Third, to assist decision-making in predictive maintenance, we implement a tire state decision algorithm, which strategically determines the optimal timing for tire replacement based on RCP forecasted by our transformer model. This approach ensures that our DT accurately predicts system health, continually refines its digital representation, and supports predictive maintenance decisions. Furthermore, our framework effectively embodies a physical system, leveraging big data and machine learning (ML) for predictive maintenance, model updates, and decision-making.

advanced computing infrastructure↗