SEARCH · Search NASA

Results for “AGE ESTIMATION”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,567 records · Page 87

All sources NASA NTRS DOE OSTI

DOE OSTI2013

CANDIDATE PLANETS IN THE HABITABLE ZONES OF KEPLER STARS

A key goal of the Kepler mission is the discovery of Earth-size transiting planets in ''habitable zones'' where stellar irradiance maintains a temperate climate on an Earth-like planet. Robust estimates of planet radius and irradiance require accurate stellar parameters, but most Kepler systems are faint, making spectroscopy difficult and prioritization of targets desirable. The parameters of 2035 host stars were estimated by Bayesian analysis and the probabilities p{sub HZ} that 2738 candidate or confirmed planets orbit in the habitable zone were calculated. Dartmouth Stellar Evolution Program models were compared to photometry from the Kepler Input Catalog, priors for stellar mass, age, metallicity and distance, and planet transit duration. The analysis yielded probability density functions for calculating confidence intervals of planet radius and stellar irradiance, as well as p{sub HZ}. Sixty-two planets have p{sub HZ} > 0.5 and a most probable stellar irradiance within habitable zone limits. Fourteen of these have radii less than twice the Earth; the objects most resembling Earth in terms of radius and irradiance are KOIs 2626.01 and 3010.01, which orbit late K/M-type dwarf stars. The fraction of Kepler dwarf stars with Earth-size planets in the habitable zone ({eta}{sub Circled-Plus }) is 0.46, with a 95% confidence interval of 0.31-0.64. Parallaxes from the Gaia mission will reduce uncertainties by more than a factor of five and permit definitive assignments of transiting planets to the habitable zones of Kepler stars.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2026

Myriad World Baseline: Global Geodemographic Estimates

The LandScan Myriad World Baseline (MWB) method produces global, residential (nighttime/home-location) gridded geodemographic estimates based on 5-year age/gender cohorts—at 30-arcsecond (≈1 km) resolution. MWB is designed to fill gaps where detailed, georeferenced survey data (e.g., Demographic and Health Surveys (DHS)) are missing or outdated, and to provide a baseline that can support human security analysis, including consequence assessment, “patterns of life” modeling, and scenario-based population futures. MWB’s workflow spatializes household-level age/gender characteristics from the GLOPOP-S dataset by conflating household and gridded expected relative wealth adapted from Global Gridded Relative Deprivation Index (GRDI), then adjusts them to a target year of interest. Age/gender estimates are then applied to harmonize lowest-administrative-level statistics with LandScan residential counts, yielding final geodemographic estimates. Two validation case studies are presented: Ghana (2021) and Tokyo/Kanagawa, Japan (2020), illustrating spatial variability in demographic cohorts and comparing MWB outputs to official gridded statistics. Results show close overall alignment relative to validation criteria including population pyramids and age-dependency ratios.

Tuccillo, Joe [ORNL] (ORCID:0000000259300943)↗

DOE OSTI2016

ON THE INFERENCE OF THE COSMIC-RAY IONIZATION RATE ζ FROM THE HCO{sup +}-to-DCO{sup +} ABUNDANCE RATIO: THE EFFECT OF NUCLEAR SPIN

The chemistry of dense interstellar regions was analyzed using a time-dependent gas–grain astrochemical simulation and a new chemical network that incorporates deuterated chemistry, taking into account nuclear spin states for the hydrogen chemistry and its deuterated isotopologues. With this new network, the utility of the [HCO{sup +}]/[DCO{sup +}] abundance ratio as a probe of the cosmic-ray ionization rate has been re-examined, with special attention paid to the effect of the initial value of the ortho-to-para ratio (OPR) of molecular hydrogen. After discussing the use of the probe for cold cores, we compare our results with previous theoretical and observational results for a molecular cloud close to the supernova remnant W51C, which is thought to have an enhanced cosmic-ray ionization rate ζ caused by the nearby γ -ray source. In addition, we attempt to use our approach to estimate the cosmic-ray ionization rate for L1174, a dense core with an embedded star. Beyond the previously known sensitivity of [HCO{sup +}]/[DCO{sup +}] to ζ , we demonstrate its additional dependence on the initial OPR and, secondarily, on the age of the source, its temperature, and its density. We conclude that the usefulness of the [HCO{sup +}]/[DCO{sup +}] abundance ratio in constraining the cosmic-ray ionization rate in dense regions increases with the age of the source and the ionization rate as the ratio becomes far less sensitive to the initial value of the OPR.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2017

Effective Radii of Young, Massive Star Clusters in Two LEGUS Galaxies

We present a study of the effective (half-light) radii and other structural properties of a systematically selected sample of young, massive star clusters (≥5 × 10{sup 3} M{sub ⊙} and ≤200 Myr) in two nearby spiral galaxies, NGC 628 and NGC 1313. We use Hubble Space Telescope (HST) WFC3/UVIS and archival ACS/WFC data obtained by the Legacy Extragalactic UV Survey (LEGUS), an HST Treasury Program. We measure effective radii with GALFIT, a two-dimensional image-fitting package, and with a new technique to estimate effective radii from the concentration index of observed clusters. The distribution of effective radii from both techniques spans ∼0.5–10 pc and peaks at 2–3 pc for both galaxies. We find slight positive correlations between effective radius and cluster age in both galaxies, but no significant relationship between effective radius and galactocentric distance. Clusters in NGC 1313 display a mild increase in effective radius with cluster mass, but the trend disappears when the sample is divided into age bins. We show that the vast majority of the clusters in both galaxies are much older than their dynamical times, suggesting they are gravitationally bound objects. We find that about half of the clusters in NGC 628 are underfilling their Roche lobes, based on their Jacobi radii. Our results suggest that the young, massive clusters in NGC 628 and NGC 1313 are expanding, due to stellar mass loss or two-body relaxation, and are not significantly influenced by the tidal fields of their host galaxies.

79 ASTRONOMY AND ASTROPHYSICS↗

NASA NTRS2012

Simulated Wake Characteristics Data for Closely Spaced Parallel Runway Operations Analysis

A simulation experiment was performed to generate and compile wake characteristics data relevant to the evaluation and feasibility analysis of closely spaced parallel runway (CSPR) operational concepts. While the experiment in this work is not tailored to any particular operational concept, the generated data applies to the broader class of CSPR concepts, where a trailing aircraft on a CSPR approach is required to stay ahead of the wake vortices generated by a lead aircraft on an adjacent CSPR. Data for wake age, circulation strength, and wake altitude change, at various lateral offset distances from the wake-generating lead aircraft approach path were compiled for a set of nine aircraft spanning the full range of FAA and ICAO wake classifications. A total of 54 scenarios were simulated to generate data related to key parameters that determine wake behavior. Of particular interest are wake age characteristics that can be used to evaluate both time- and distance- based in-trail separation concepts for all aircraft wake-class combinations. A simple first-order difference model was developed to enable the computation of wake parameter estimates for aircraft models having weight, wingspan and speed characteristics similar to those of the nine aircraft modeled in this work.

Guerreiro, Nelson M.↗

NASA NTRS2017

Carotid Intima Media Thickness in the Astronaut Corps: Association to Spacecraft

Background: Carotid Intima Media Thickness (CIMT) has been demonstrated to be predictive of future cardiovascular events. Within various populations, radiation exposure, stress, and physical confinement have all been linked to an increased CIMT. Recent research discovered CIMT was significantly increased in ten long duration astronauts from pre-flight to four days post flight. The relationship between spaceflight and CIMT is not understood and trends in CIMT within the larger astronaut population are unknown. Methods: In 2010, CIMT was offered as part of the astronaut annual exam at the JSC Flight Medicine Clinic using a standardized CIMT screening protocol and professional sonographers. Between 2010 and 2016, CIMT measurements were collected on 213 NASA astronauts and payload specialists. The values used in this retrospective chart review are the mean of the CIMT from the right and left. Spaceflight exposure was categorized based on the total number of days spent in space at the time of the ground-based ultrasound (0, 1-29, 30-100, 101-200, ≥200). Linear regression with generalized estimating equations were used to estimate the association between spaceflight exposures and CIMT. Results: 530 studies were completed among 213 astronauts with a mean of 2.5 studies (range 1-6) per astronaut over the six year period. As in other populations, CIMT was significantly associated with age; however, gender was not. While there was no significant direct correlation between total spaceflight exposure and CIMT found, astronauts with 30-100 spaceflight days and astronauts with greater than 100 spaceflight days had significantly increased CIMT over astronauts who had never flown (p=0.002 and p=<0.0001 respectively) after adjustment for age. Conclusion: Further work is needed to fully understand CIMT and its association to spaceflight. Current occupational surveillance activities are under way to study CIMT values in conjunction with other cardiovascular risk factors among astronauts as compared to the general population.

Suffredini, John↗

DOE OSTI2019

Coupling RAVEN to SAPHIRE for Performing Time Dependent Probabilistic Risk Assessment

Cost Risk Analysis Framework Tool (CRAFT) was recently developed by Idaho National Laboratory (INL) to perform economical risk analysis by combining physics, risk and cost information of components in a Nuclear Power Plant (NPP) that can be quickly deployable through the nuclear industry to evaluate the risk associated to capital projects and asset management [1]. In this framework, RAVEN is coupled with SAPHIRE to perform the time dependent risk/safety assessment of NPPs. Traditionally, the risk associated to the NPPs can be determined by classical Probabilistic Risk Assessment (PRA) tools, such as SAPHIRE, under the assumption that all failure rates are constant over time. In some circumstances, however, the assumption is rendered invalid since that the failure rate of a component may gradually change in a systematic way, for example, because of aging and degradation. In addition, recent data analysis in [2] has suggested decreasing values of failure rates for several important components of NPPs. A PRA that includes these components should employ the time-dependent failure rates in order to accurately estimate the risk at any given time. This time-dependent PRA is achieved by the coupling between RAVEN and SAPHIRE.

42 ENGINEERING↗

DOE OSTI2019

Coupling RAVEN to SAPHIRE for Performing Time Dependent Probabilistic Risk Assessment

42 ENGINEERING↗

DOE OSTI2026

A Systematic Review and Integrated Approach to Modeling of Aging Utility Scale PV Systems

The growing deployment of utility-scale photovoltaic (PV) systems has increased the importance of techno-economic modeling operational photovoltaic (PV) systems for predicting energy yield, optimizing asset management, and informing financial decisions. Through a systematic review of literature and current industry practices, we review the different common modeling practices of a system's configuration and age, performance and degradation, operation and maintenance (O&M), while also focusing on specific considerations for repowering, revamping, and decommissioning. Building on the synthesis, we develop a structured framework for techno-economic modeling of operating PV systems that integrate performance and degradation analysis, a decommissioning and repowering cost model that estimates the system's end-of-life costs to reduce uncertainty quantifications and improve consistency across the sector. This research contributes to improved modeling methodologies and potentially to reduced financial performance requirements by providing practitioners with input resources and practical approaches to estimate performance, degradation, and costs associated with continued operation, revamping, repowering, or decommissioning decisions.

14 SOLAR ENERGY↗

DOE OSTI2017

An Incipient Debris Disk in the Chamaeleon I Cloud

The point at which a protoplanetary disk becomes a debris disk is difficult to identify. To better understand this, here we study the ∼40 au separation binary T 54 in the Chamaeleon I cloud. We derive a K5 spectral type for T 54 A (which dominates the emission of the system) and an age of ∼2 Myr. However, the dust disk properties of T 54 are consistent with those of debris disks seen around older- and earlier-type stars. At the same time, T 54 has evidence of gas remaining in the disk, as indicated by [Ne ii], [Ne iii], and [O i] line detections. We model the spectral energy distribution of T 54 and estimate that ∼3×10{sup −3} M{sub ⊕} of small dust grains (<0.25 μm) are present in an optically thin circumbinary disk along with at least ∼3×10{sup −7} M{sub ⊕} of larger (>10 μm) grains within a circumprimary disk. Assuming a solar-like mixture, we use Ne line luminosities to place a minimum limit on the gas mass of the disk (∼3×10{sup −4} M{sub ⊕}) and derive a gas-to-dust mass ratio of ∼0.1. We do not detect substantial accretion, but we do see Hα in emission in one epoch, which is suggestive that there may be intermittent dumping of small amounts of matter onto the star. Considering the low dust mass, the presence of gas, and young age of T 54, we conclude that this system is on the bridge between the protoplanetary and debris disk stages.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2018

Using tracer-derived groundwater transit times to assess storage within a high-elevation watershed of the upper Colorado River Basin, USA

Previous watershed assessments have relied on annual baseflow to evaluate the groundwater contribution to streams. To quantify the volume of groundwater in storage, additional information such as groundwater mean transit time (MTT) is needed. This study determined the groundwater MTT in the West Fork Duchesne watershed in Utah (USA) with lumped-parameter modeling of environmental tracers (SF{sub 6}, CFCs, and {sup 3}H/{sup 3}He) from 21 springs. Approximately 30% of the springs exhibited an exponential transit time distribution (TTD); the remaining ~70% were best characterized by a piston-flow TTD. The flow-weighted groundwater MTT for the West Fork watershed is about 40 years with approximately 20 years in the unsaturated zone. A cumulative distribution of these ages revealed that most of the groundwater is between 30 and 50 years old, suggesting that declining recharge associated with 5–10-year droughts is less likely to have a profound effect on this watershed compared with systems with shorter MTTs. The estimated annual baseflow of West Fork stream flow based on chemical hydrograph separation is ~1.7 × 10{sup 7} m{sup 3}/year, a proxy for groundwater discharge. Using both MTT and groundwater discharge, the volume of mobile groundwater stored in the watershed was calculated to be ~6.5 × 10{sup 8} m{sup 3}, or ~20 m thickness of active groundwater storage and recharge of ~0.09 m/year (assuming porosity = 15%). Future watershed-scale assessments should evaluate groundwater MTT, in addition to annual baseflow, to quantify groundwater storage and more accurately assess watershed susceptibility to drought, groundwater extraction, and land-use change.

58 GEOSCIENCES↗

DOE OSTI2015

A high-resolution Hubble Space Telescope study of apparent Lyman continuum leakers at z ∼ 3

We present U{sub 336}V{sub 606}J{sub 125}H{sub 160} follow-up Hubble Space Telescope (HST) observations of 16 z ∼ 3 candidate Lyman continuum (LyC) emitters in the HS1549+1919 field. With these data, we obtain high spatial-resolution photometric redshifts of all sub-arcsecond components of the LyC candidates in order to eliminate foreground contamination and identify robust candidates for leaking LyC emission. Of the 16 candidates, we find one object with a robust LyC detection that is not due to foreground contamination. This object (MD5) resolves into two components; we refer to the LyC-emitting component as MD5b. MD5b has an observed 1500 Å to 900 Å flux-density ratio of (F{sub UV}/F{sub LyC}){sub obs}=4.0±2.0, compatible with predictions from stellar population synthesis models. Assuming minimal IGM absorption, this ratio corresponds to a relative (absolute) escape fraction of f{sub esc,rel}{sup MD5b} = 75%–100% (f{sub esc,abs}{sup MD5b} = 14%–19%). The stellar population fit to MD5b indicates an age of ≲50 Myr, which is in the youngest 10% of the HST sample and the youngest third of typical z ∼ 3 Lyman break galaxies, and may be a contributing factor to its LyC detection. We obtain a revised, contamination-free estimate for the comoving specific ionizing emissivity at z = 2.85, indicating (with large uncertainties) that star-forming galaxies provide roughly the same contribution as QSOs to the ionizing background at this redshift. Our results show that foreground contamination prevents ground-based LyC studies from obtaining a full understanding of LyC emission from z ∼ 3 star-forming galaxies. Future progress in direct LyC searches is contingent upon the elimination of foreground contaminants through high spatial-resolution observations, and upon acquisition of sufficiently deep LyC imaging to probe ionizing radiation in high-redshift galaxies.

79 ASTRONOMY AND ASTROPHYSICS↗

NASA NTRS1995

Ocean observations with EOS/MODIS: Algorithm development and post launch studies

Several significant accomplishments were made during the present reporting period. (1) Initial simulations to understand the applicability of the MODerate Resolution Imaging Spectrometer (MODIS) 1380 nm band for removing the effects of stratospheric aerosols and thin cirrus clouds were completed using a model for an aged volcanic aerosol. The results suggest that very simple procedures requiring no a priori knowledge of the optical properties of the stratospheric aerosol may be as effective as complex procedures requiring full knowledge of the aerosol properties, except the concentration which is estimated from the reflectance at 1380 nm. The limitations of this conclusion will be examined in the next reporting period; (2) The lookup tables employed in the implementation of the atmospheric correction algorithm have been modified in several ways intended to improve the accuracy and/or speed of processing. These have been delivered to R. Evans for implementation into the MODIS prototype processing algorithm for testing; (3) A method was developed for removal of the effects of the O2 'A' absorption band from SeaWiFS band 7 (745-785 nm). This is important in that SeaWiFS imagery will be used as a test data set for the MODIS atmospheric correction algorithm over the oceans; and (4) Construction of a radiometer, and associated deployment boom, for studying the spectral reflectance of oceanic whitecaps at sea was completed. The system was successfully tested on a cruise off Hawaii on which whitecaps were plentiful during October-November. This data set is now under analysis.

Gordon, Howard R.↗

NASA NTRS1994

Farby-Perot observations and new models of the HH 47A and HH 47D bow shocks

We present new models for the HH 47A and HH 47D bow shocks based on line flux and velocity maps obtained with an imaging Fabry-Perot spectrometer. We confirm that HH 47A and HH 47D each show a bow shock/Mach disk morphology, and that velocity variability in the outflow can account for the observed structures. While it was suggested a decade ago that the inner working surface HH 47A appears to be traveling into the wake of HH 47D, we find kinematic evidence that the outer bow shock HH 47D is also not the primary ejection event in the outflow but follows in the wake of previously ejected material. By comparing the observed line ratios and line profiles to those predicted by our bow shock models, we find that both bow shocks have substantially lower shock velocities than their space motions would imply, and that the emission from each bow shock is systematically blueshifted from the rest-frame velocity of the ambient emission, indicating a comoving preshock medium. We derive kinematic ages of approximately 1150 yr for HH 47D and approximately 550 yr for HH 47A, which implies that the stellar driving source may undergo repetitive eruptions similar to FU Orionis-type outbursts every several hundred years. This timescale is similar to estimates made by Reipurth and collaborators for the separation between major outbursts in the HH 34 and HH 111 stellar jets.

Morse, Jon A.↗

NASA NTRS1996

Relating the Hadamard Variance to MCS Kalman Filter Clock Estimation

The Global Positioning System (GPS) Master Control Station (MCS) currently makes significant use of the Allan Variance. This two-sample variance equation has proven excellent as a handy, understandable tool, both for time domain analysis of GPS cesium frequency standards, and for fine tuning the MCS's state estimation of these atomic clocks. The Allan Variance does not explicitly converge for the nose types of alpha less than or equal to minus 3 and can be greatly affected by frequency drift. Because GPS rubidium frequency standards exhibit non-trivial aging and aging noise characteristics, the basic Allan Variance analysis must be augmented in order to (a) compensate for a dynamic frequency drift, and (b) characterize two additional noise types, specifically alpha = minus 3, and alpha = minus 4. As the GPS program progresses, we will utilize a larger percentage of rubidium frequency standards than ever before. Hence, GPS rubidium clock characterization will require more attention than ever before. The three sample variance, commonly referred to as a renormalized Hadamard Variance, is unaffected by linear frequency drift, converges for alpha is greater than minus 5, and thus has utility for modeling noise in GPS rubidium frequency standards. This paper demonstrates the potential of Hadamard Variance analysis in GPS operations, and presents an equation that relates the Hadamard Variance to the MCS's Kalman filter process noises.

Hutsell, Steven T.↗

NASA NTRS1987

Logistic Risk Model for the Unique Effects of Inherent Aerobic Capacity on (+)G(sub z) Tolerance Before and After Simulated Weightlessness

Small sample size (n less than 1O) and inappropriate analysis of multivariate data have hindered previous attempts to describe which physiologic and demographic variables are most important in determining how long humans can tolerate acceleration. Data from previous centrifuge studies conducted at NASA/Ames Research Center, utilizing a 7-14 d bed rest protocol to simulate weightlessness, were included in the current investigation. After review, data on 25 women and 22 men were available for analysis. Study variables included gender, age, weight, height, percent body fat, resting heart rate, mean arterial pressure, Vo(sub 2)max and plasma volume. Since the dependent variable was time to greyout (failure), two contemporary biostatistical modeling procedures (proportional hazard and logistic discriminant function) were used to estimate risk, given a particular subject's profile. After adjusting for pro-bed-rest tolerance time, none of the profile variables remained in the risk equation for post-bed-rest tolerance greyout. However, prior to bed rest, risk of greyout could be predicted with 91% accuracy. All of the profile variables except weight, MAP, and those related to inherent aerobic capacity (Vo(sub 2)max, percent body fat, resting heart rate) entered the risk equation for pro-bed-rest greyout. A cross-validation using 24 new subjects indicated a very stable model for risk prediction, accurate within 5% of the original equation. The result for the inherent fitness variables is significant in that a consensus as to whether an increased aerobic capacity is beneficial or detrimental has not been satisfactorily established. We conclude that tolerance to +Gz acceleration before and after simulated weightlessness is independent of inherent aerobic fitness.

Ludwig, David A.↗

NASA NTRS1998

The Local ISM and its Interaction with the Winds of Nearby Late-type Stars

We present new Goddard High-Resolution Spectrograph (GHRS) observations of the Ly-alpha and Mg II absorption lines seen toward the nearby stars 61 Cyg A and 40 Eri A. We use these data to measure interstellar properties along these lines of sight and to search for evidence of circumstellar hydrogen walls, which are produced by collisions between the stellar winds and the Local InterStellar Medium (LISM). We were able to model the Ly-alpha lines of both stars without hydrogen-wall absorption components, but for 61 Cyg A the fit required a stellar Ly-alpha, line profile with an improbably deep self-reversal, and for 40 Eri A the fit required a very low deuterium-to-hydrogen ratio that is inconsistent with previous GHRS measurements. Since these problems could be rectified simply by including stellar hydrogen-wall components with reasonable attributes, our preferred fits to the data include these components. We have explored several ways in which the hydrogen-wall properties measured here and in previous work can be used to study stellar winds and the LISM. We argue that the existence of a hydrogen wall around 40 Eri A and a low H I column density along that line of sight imply that either the interstellar density must decrease toward 40 Eri A or the hydrogen ionization fraction (chi) must increase. We find that hydrogen-wall temperatures are larger for stars with faster velocities through the LISM. The observed temperature-velocity relation is consistent with the predictions of hydromagnetic shock jump conditions. More precise comparison of the data and the jump conditions suggests crude upper limits for both chi and the ratio of magnetic to thermal pressure in the LISM (alpha): chi less than 0.6 and alpha less than 2. The latter upper limit corresponds to a limit on the LISM magnetic field of B less than 5 micro G. These results imply that the plasma Mach number of the interstellar wind flowing into the heliosphere is M(sub A) greater than 1.3, which indicates that the collision is supersonic and that there should therefore be a bow shock outside the heliopause in the upwind direction. Finally, we estimate stellar wind pressures (P sub wind) from the measured hydrogen-wall column densities. These estimates represent the first empirical measurements of wind properties for late-type main-sequence stars. The wind pressures appear to be correlated with stellar X-ray surface fluxes, F(x), in a manner consistent with the relation P(wind) varies as F(x)(exp -1/2), a relation that is also consistent with the variations of P(sub wind) and F(sub x) observed during the solar activity cycle. If this relation can in fact be generalized to solar-like stars, as is suggested by our data, then it is possible to estimate stellar wind properties simply by measuring stellar X-rays. One implication of this is that stellar wind pressures and mass-loss rates are then predicted to increase with time, since F(sub x) is known to decrease with stellar age.

Wood, Brian E.↗

DOE OSTI2023

Machine-Learning Assisted Identification of Battery Life Models

Predictive battery life models are commonly utilized to extrapolate degradation trends observed during accelerated aging tests for simulation of degradation in real-world applications. Thus, fitting accelerated aging data as accurately as possible and with low uncertainty is crucial for making believable projections of battery lifetime, but it is challenging to identify algebraic expressions that accurately fit multivariate degradation trends. A review of models published in literature reveal some common expressions for fitting calendar aging data, which is only dependent on temperature and state-of-charge, but no consistency across many models for fitting cycle aging data, indicating the need for a statistically rigorous data driven approach for developing empirical models. This talk will describe a machine-learning assisted method for identification of predictive battery life models utilizing bilevel optimization and symbolic regression. Bilevel optimization with cross-validation is used to statistically determine cell- and stress-dependent model parameters, while symbolic regression identifies both linear and multiplicative candidate expressions to predict stress-dependent degradation rates by selecting low-order subsets of features from a generated feature library. Because model expressions are identified empirically, it is crucial to ensure resulting models behave according to physical expectations, so the stability of models for interpolation or extrapolation is interrogated qualitatively through simulation and quantitatively through cross-validation and uncertainty quantification via bootstrap resampling. This model identification approach substantially improves upon models identified purely using expert judgement in terms of both accuracy and uncertainty. Model simulation and validation is then conducted by deriving a state-equation form of the predictive model, enabling simulation of battery aging under dynamic stresses. This enables validation of the predictive battery model on lab-based tests with varying conditions or on drive-cycle or application-cycle testing protocols. Parameter uncertainty can be carried forward into model simulation, giving lifetime estimates and confidence windows for cell- or system-level lifetime. The financial impact of battery model uncertainty can be estimated by incorporating uncertainty into a technoeconomic model.

battery↗