Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Feature importance”

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 289 records · Page 16
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

HyMARC Seedling: Optimized Hydrogen Adsorbents via Machine Learning and Crystal Engineering

This final report is presented in two parts: In Part 1, the H 2 capacities of a diverse set of 918,734 metal-organic frameworks (MOFs) sourced from 19 databases is predicted via machine learning (ML). Using only 7 structural features as input, ML identifies 8,282 MOFs with the potential to exceed the capacities of state-of-the-art materials. The identified MOFs are predominantly hypothetical compounds having low densities (<0.31 g/cm3) in combination with high surface areas (>5,300 m2/g), void fractions ($0.90), and pore volumes (>3.3 cm3/g). The relative importance of the input features are characterized, and dependencies on the ML algorithm and training set size are quantified. The most important features for predicting H 2 uptake are pore volume (for gravimetric capacity) and void fraction (for volumetric capacity). The ML models are available on the web, allowing for rapid and accurate predictions of the hydrogen capacities of MOFs from limited structural data; the simplest models require only a single crystallographic feature. In part 2, ways to improve the poor powder packing density of MOFs is discussed. More specifically, a strategy that improves packing efficiency and volumetric hydrogen gas storage density dramatically through engineered morphologies and controlled-crystal size distributions is presented that holds promise for maximizing storage capacity for a given MOF. The packing density improvement, demonstrated for the benchmark sorbent MOF-5, leads to a significant enhancement of volumetric hydrogen storage performance relative to commercial MOF-5. System model projections demonstrate that engineering of crystal morphology/size or use of a bimodal distribution of cubic crystal sizes in tandem with system optimization can surpass the 25 g/L volumetric capacity of a typical 700 bar compressed storage system and exceed the DOE targets 2020 volumetric capacity (30 g/L). Finally, a critical link between improved powder packing density and reduced damage upon compaction is revealed leading to sorbents with both high surface area and high density.

08 HYDROGEN↗
DOE OSTI2021

Nanoparticle cellular internalization is not required for RNA delivery to mature plant leaves

Rapidly growing interest in the nanoparticle-mediated delivery of DNA and RNA to plants requires a better understanding of how nanoparticles and their cargoes translocate in plant tissues and into plant cells. However, little is known about how the size and shape of nanoparticles influence transport in plants and the delivery efficiency of their cargoes, limiting the development of nanotechnology in plant systems. In this study we employed non-biolistically delivered DNA-modified gold nanoparticles (AuNPs) of various sizes (5-20 nm) and shapes (spheres and rods) to systematically investigate their transport following infiltration into Nicotiana benthamiana leaves. Generally, smaller AuNPs demonstrated more rapid, higher and longer-lasting levels of association with plant cell walls compared with larger AuNPs. We observed internalization of rod-shaped but not spherical AuNPs into plant cells, yet, surprisingly, 10 nm spherical AuNPs functionalized with small-interfering RNA (siRNA) were the most efficient at siRNA delivery and inducing gene silencing in mature plant leaves. These results indicate the importance of nanoparticle size in efficient biomolecule delivery and, counterintuitively, demonstrate that efficient cargo delivery is possible and potentially optimal in the absence of nanoparticle cellular internalization. Additionally, our results highlight nanoparticle features of importance for transport within plant tissues, providing a mechanistic overview of how nanoparticles can be designed to achieve efficacious biocargo delivery for future developments in plant nanobiotechnology.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Detailed simulations of the first deuterium-tritium-filled double shell implosions on the National Ignition Facility

The first indirectly driven, liquid DT-filled double shell inertial confinement fusion (ICF) implosions have recently been successfully performed on the National Ignition Facility (NIF). Double shells are a class of alternative designs that use a low-Z outer shell to compress a foam cushion that accelerates a high-Z inner shell to efficiently compress a liquid DT core. Double shells are challenging to fabricate, field, and model. Important engineering features enabling double shell fabrication include a fill-tube penetrating all shells and a carefully designed and very narrow (few μm) step-joint in the ablator. Due to the higher density materials involved, high Atwood number instabilities are also important at many material interfaces. In this paper, numerical simulations of double shell implosions using the Los Alamos National Laboratory multi-physics radiation-hydrodynamics code xRAGE will be discussed. An extensive effort has been under way for several years to develop the code capabilities for ICF simulations in a common modeling framework to allow ease of simulation setup and standardization of the computational methodology. This paper will present a wide range of simulation results capturing, quantifying, and comparing the impact of all these degradation mechanisms on implosion performance. Brief comparisons with recent experimental results and suggestions for future improvements will also be discussed. Our results suggest that capsule surface roughness and the step-joint gap have the largest impact on implosion performance. Initial experimental data may suggest that the sensitivity to the step-joint gap could provide the dominant explanation for DT-filled double shell experiments that have been fielded on NIF thus far.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Modeling the distribution of the endangered Jemez Mountains salamander (Plethodon neomexicanus) in relation to geology, topography, and climate

The Jemez Mountains salamander (Plethodon neomexicanus; hereafter JMS) is an endangered salamander restricted to the Jemez Mountains in north-central New Mexico, United States. This strictly terrestrial and lungless species requires moist surface conditions for activities such as mating and foraging. Threats to its current habitat include fire suppression and ensuing severe fires, changes in forest composition, habitat fragmentation, and climate change. Forest composition changes resulting from reduced fire frequency and increased tree density suggest that its current aboveground habitat does not mirror its historically successful habitat regime. However, because of its limited habitat area and underground behavior, we hypothesized that geology and topography might play a significant role in the current distribution of the salamander. We modeled the distribution of the JMS using a machine learning algorithm to assess how geology, topography, and climate variables influence its distribution. The best habitat suitability model indicates that geology type and maximum winter temperature (November to March) were most important in predicting the distribution of the salamander (23.5% and 50.3% permutation importance, respectively). Minimum winter temperature was also an important variable (21.4%), suggesting this also plays a role in salamander habitat. Our habitat suitability map reveals low uncertainty in model predictions, and we found slight discrepancies between the designated critical habitat and the most suitable areas for the JMS. Because geological features are important to its distribution, we recommend that geological and topographical data are considered, both during survey design and in the description of localities of JMS records once detected.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

The Fall and Rise of the Global Climate Model

Abstract Global models are an essential tool for climate projections, but conventional coarse‐resolution atmospheric general circulation models suffer from errors both in their parameterized cloud physics and in their representation of climatically important circulation features. A notable recent study by Terai et al. (2020, https://doi.org/10.1029/2020ms002274) documents a global model capable of reproducing the regime‐based effect of aerosols on cloud liquid water path expected from observational evidence. This may represent a significant advance in cloud process fidelity in global models. Such models can be expected to give a better estimate of the effective radiative forcing of the climate. If this advance in cloud process representation can be matched by advances in the representation of circulation features such as monsoons, then such models may also be able to navigate the complex tangle between spatially heterogeneous aerosol–cloud interactions and regional circulation patterns. This tight link between aerosol and circulation results in anthropogenic perturbations of climate variables of societal importance, such as regional rainfall distributions. Upcoming global models with km‐scale resolution may improve the regional circulation and be able to take advantage of the Terai et al. (2020, https://doi.org/10.1029/2020ms002274) improvement in cloud physics. If so, an era of significantly improved regional climate projection capabilities may soon dawn. If not, then the improvement in cloud physics might spur intensified efforts on problems in model dynamics. Either way, based on the rapid changes in aerosol emissions in the near future, learning to make reliable projections based on biased models is a skill that will not go out of style.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Analysis of electrically conductive adhesives in shingled solar modules by X-ray imaging techniques

The failure mechanisms of electrically conductive adhesives (ECAs) in solar modules are difficult to study since the ECA layer is not easily accessible within the module package. In this work, we present two complementary imaging modalities—X-ray radiography and X-ray microcomputed tomography (XCT)—that reveal important morphological features of the ECA within a shingled module. X-ray radiography uses single X-ray projections to provide fast and non-destructive imaging of the shingled interconnection, illuminating the alignment of the ECA relative to the busbars, the size and shape of the ECA, and the presence of voids within it. Through X-ray radiography, we observed, for example, that the average void coverage area of ECA segments reduced from 36.7% to 4.4% when an ECA was cured for 60 s prior to module lamination. XCT is a three-dimensional imaging technique that can identify regions in which the ECA makes electrical contact to busbars on cells and regions in which the ECA has cracked, among other features. XCT can also be used to image the individual metal particles within ECA, from which the metal volume fraction of an ECA was found here to be 70.4%. This is a quantity that is not often reported by ECA manufacturers but is important to ensure isotropic conduction. As X-ray projections can be performed non-destructively on full modules, the technique may be used to pinpoint ECA failures in accelerated degradation testing. XCT is complementary and is suited to forensic analysis of failing modules.

14 SOLAR ENERGY↗
DOE OSTI2022

Formation of the N≡N Triple Bond from Reductive Coupling of a Paramagnetic Diruthenium Nitrido Compound

Construction of nitrogen–nitrogen triple bonds via homocoupling of metal nitrides is an important fundamental reaction relevant to a potential Nitrogen Economy. Here, we report that room temperature photolysis of Ru 2 (chp) 4 N 3 (chp– = 2-chloro-6-hydroxypyridinate) in CH 2 Cl 2 produces N 2 via reductive coupling of Ru2(chp)4N nitrido species. Computational analysis reveals that the nitride coupling transition state (TS) features an out-of-plane “zigzag” geometry instead of the anticipated planar zigzag TS. However, with intentional exclusion of dispersion correction, the planar zigzag TS geometry can also be found. Both the out-of-plane and planar zigzag TS geometries feature two important types of orbital interactions: (1) donor–acceptor interactions involving intermolecular donation of a nitride lone pair into an empty Ru–N π* orbital and (2) Ru–N π to Ru–N π* interactions derived from coupling of nitridyl radicals. The relative importance of these two interactions is quantified both at and after the TS. Our analysis shows that both interactions are important for the formation of the N–N σ bond, while radical coupling interactions dominate the formation of N–N π bonds. Comparison is made to isoelectronic Ru 2 -oxo compounds. Formation of an O–O bond via bimolecular oxo coupling is not observed experimentally and is calculated to have a much higher TS energy. As a result, the major difference between the nitrido and oxo systems stems from an extremely large driving force, ~–500 kJ/mol, for N–N coupling vs a more modest driving force for O–O coupling, –40 to –140 kJ/mol.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Navigating Transition-Metal Chemical Space: Artificial Intelligence for First-Principles Design

Conspectus The variability of chemical bonding in open-shell transition-metal complexes not only motivates their study as functional materials and catalysts but also challenges conventional computational modeling tools. Here, tailoring ligand chemistry can alter preferred spin or oxidation states as well as electronic structure properties and reactivity, creating vast regions of chemical space to explore when designing new materials atom by atom. Although first-principles density functional theory (DFT) remains the workhorse of computational chemistry in mechanism deduction and property prediction, it is of limited use here. DFT is both far too computationally costly for widespread exploration of transition-metal chemical space and also prone to inaccuracies that limit its predictive performance for localized d electrons in transition-metal complexes. These challenges starkly contrast with the well-trodden regions of small-organic-molecule chemical space, where the analytical forms of molecular mechanics force fields and semiempirical theories have for decades accelerated the discovery of new molecules, accurate DFT functional performance has been demonstrated, and gold-standard methods from correlated wavefunction theory can predict experimental results to chemical accuracy. The combined promise of transition-metal chemical space exploration and lack of established tools has mandated a distinct approach. In this Account, we outline the path we charted in exploration of transition-metal chemical space starting from the first machine learning (ML) models (i.e., artificial neural network and kernel ridge regression) and representations for the prediction of open-shell transition-metal complex properties. The distinct importance of the immediate coordination environment of the metal center as well as the lack of low-level methods to accurately predict structural properties in this coordination environment first motivated and then benefited from these ML models and representations. Once developed, the recipe for prediction of geometric, spin state, and redox potential properties was straightforwardly extended to a diverse range of other properties, including in catalysis, computational “feasibility”, and the gas separation properties of periodic metal–organic frameworks. Interpretation of selected features most important for model prediction revealed new ways to encapsulate design rules and confirmed that models were robustly mapping essential structure–property relationships. Encountering the special challenge of ensuring that good model performance could generalize to new discovery targets motivated investigation of how to best carry out model uncertainty quantification. Distance-based approaches, whether in model latent space or in carefully engineered feature space, provided intuitive measures of the domain of applicability. With all of these pieces together, ML can be harnessed as an engine to tackle the large-scale exploration of transition-metal chemical space needed to satisfy multiple objectives using efficient global optimization methods. In practical terms, bringing these artificial intelligence tools to bear on the problems of transition-metal chemical space exploration has resulted in ML-model assessments of large, multimillion compound spaces in minutes and validated new design leads in weeks instead of decades.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

3D-equivariant graph neural networks for protein model quality assessment

Quality assessment (QA) of predicted protein tertiary structure models plays an important role in ranking and using them. With the recent development of deep learning end-to-end protein structure prediction techniques for generating highly confident tertiary structures for most proteins, it is important to explore corresponding QA strategies to evaluate and select the structural models predicted by them since these models have better quality and different properties than the models predicted by traditional tertiary structure prediction methods. We develop EnQA, a novel graph-based 3D-equivariant neural network method that is equivariant to rotation and translation of 3D objects to estimate the accuracy of protein structural models by leveraging the structural features acquired from the state-of-the-art tertiary structure prediction method—AlphaFold2. We train and test the method on both traditional model datasets (e.g. the datasets of the Critical Assessment of Techniques for Protein Structure Prediction) and a new dataset of high-quality structural models predicted only by AlphaFold2 for the proteins whose experimental structures were released recently. Our approach achieves state-of-the-art performance on protein structural models predicted by both traditional protein structure prediction methods and the latest end-to-end deep learning method—AlphaFold2. It performs even better than the model QA scores provided by AlphaFold2 itself. The results illustrate that the 3D-equivariant graph neural network is a promising approach to the evaluation of protein structural models. Integrating AlphaFold2 features with other complementary sequence and structural features is important for improving protein model QA.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Ensemble methods for quantification of potassium oxide in ChemCam Mars and laboratory spectra

In this paper we test new approaches for predicting the amount of element oxides in rock samples from the ChemCam instrument suite onboard the NASA Curiosity rover by focusing on K 2 O. Using the expanded dataset compiled by Gasda et al. (2021) with and without the Earth to Mars (E2M and NoE2M) transformation discussed in Clegg et al. (2017) we trained blended submodels using the “double blending” technique and compared these to ensemble methods (Random Forest, ExtraTrees, and Gradient Boosting Regression). We found that ensemble methods performed similar to blended submodels when looking at RMSE-P on the laboratory spectra and provided significant advantages when looking at spectra coming from Mars. For the full model, blended submodels achieved an RMSE-P of 0.62 and 0.60 (E2M and NoE2M respectively) while Gradient Boosting Regression resulted in a slightly improved RMSE-P of 0.59 and 0.60. More importantly, by employing a local RMSE-P estimation technique where model performance is evaluated based on nearby test samples we found that using ensemble methods can lower the quantification limit for K 2 O from the current value of ≈0.6 wt% to ≈0.08 wt% using Extra Trees and Random Forest. This would allow for a much larger range of K 2 O values to be quantified on Mars with greater certainty given that most targets seen on Mars tend to have <1 wt% K2O. Finally, we used both Mean Decrease in Impurity (MDI) and permutation importance techniques to investigate the wavelengths used by the ensemble methods and found that they correspond to known potassium emission lines. This suggests that ensemble methods can provide an easier to train and improved alternative to blended submodels for predicting potassium compositions from Laser Induced Breakdown Spectroscopy (LIBS) data.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Machine learning prediction of incidence of Alzheimer’s disease using large-scale administrative health data

Nationwide population-based cohort provides a new opportunity to build an automated risk prediction model based on individuals’ history of health and healthcare beyond existing risk prediction models. We tested the possibility of machine learning models to predict future incidence of Alzheimer’s disease (AD) using large-scale administrative health data. From the Korean National Health Insurance Service database between 2002 and 2010, we obtained de-identified health data in elders above 65 years (N = 40,736) containing 4,894 unique clinical features including ICD-10 codes, medication codes, laboratory values, history of personal and family illness and socio-demographics. To define incident AD we considered two operational definitions: “definite AD” with diagnostic codes and dementia medication (n = 614) and “probable AD” with only diagnosis (n = 2026). We trained and validated random forest, support vector machine and logistic regression to predict incident AD in 1, 2, 3, and 4 subsequent years. For predicting future incidence of AD in balanced samples (bootstrapping), the machine learning models showed reasonable performance in 1-year prediction with AUC of 0.775 and 0.759, based on “definite AD” and “probable AD” outcomes, respectively; in 2-year, 0.730 and 0.693; in 3-year, 0.677 and 0.644; in 4-year, 0.725 and 0.683. The results were similar when the entire (unbalanced) samples were used. Important clinical features selected in logistic regression included hemoglobin level, age and urine protein level. This study may shed a light on the utility of the data-driven machine learning model based on large-scale administrative health data in AD risk prediction, which may enable better selection of individuals at risk for AD in clinical trials or early detection in clinical settings.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Whole plant traits of coastal dune vegetation and implications for interactions with dune dynamics

Abstract Coastal dunes are important protective features against sea level rise and coastal storms. Interactions between dune plant aboveground structures and sediment trapping that allow for dune building and maintenance are well established. More recently, studies documenting belowground biomass for promoting erosion resistance in dominant dune species have been conducted, yet a knowledge gap remains regarding species‐specific characterization of whole plants, specifically with respect to roots, rhizomes, and belowground stems. Our objective was to quantify above‐ and belowground traits of four dominant dune grasses to document the potential for species‐specific effects on dune growth, maintenance, and erosion resistance. We examined above‐ and belowground traits among four prominent dune grasses of the Atlantic and Gulf Coasts of North America: Ammophila breviligulata , Panicum amarum , Spartina patens , and Uniola paniculata . Whole plant samples of each species were collected from the foredune at the US Army Engineer Research and Development Center's Field Research Facility in Duck, North Carolina, USA, and quantified for several above‐ and belowground traits (e.g., stem height, rhizome number and length, root surface area by diameter class, root tensile strength, and mycorrhizal percent infection). Belowground factors known to impact important dune processes, such as rhizome length, mycorrhizal percent infection, and root traits, differed substantially among species. When visualized in multivariate space, all species significantly differed in suites of above‐ and belowground traits. When considering belowground only, Ammophila and Spartina were similar, despite differences in biomass allocation. Species separated along axes related to mycorrhizal association, biomass allocation, and root construction. The four co‐occurring dune grass species were dissimilar in suites of plant traits. Belowground trait differences were driven by those describing root construction, biomass allocation, and mycorrhizal infection. Dissimilarity in above‐ and belowground suites of traits may demonstrate different approaches for surviving the dune environment. Incorporating belowground traits into modeling will enhance predictions of dune response to climate change through interactions between vegetation and dune dynamics that facilitate coastal resistance and resilience.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Characterizing Structure-Dependent TiS 2 /Water Interfaces Using Deep-Neural-Network-Assisted Molecular Dynamics

As a promising layered electrode material, TiS 2 -based capacitive deionization (CDI) devices for water desalination have attracted significant attention. However, TiS 2 /H 2 O interfacial features, potentially important for device optimization, remain unidentified. Using Deep Potential Molecular Dynamics (DPMD), we characterized distinct aqueous interfaces introduced by four TiS 2 terminations expected to be present as water intercalates into TiS 2 , namely, Armchair, Zigzag, Zigzag-L, and Zigzag-R. First, we assessed important representative physical properties of the system to validate the deep potentials (DPs). DPMD simulations agree well with experiments and first-principles simulations, suggesting the DPs are accurate and reliable. Subsequent simulations of these TiS 2 /water interfaces revealed how TiS 2 surface termination influences the structure of interfacial water. This effect is most evident in the first and second water layers close to the TiS 2 surface, and more pronounced when spontaneous dissociative adsorption of water occurs. The extent of water dissociation on each surface was evaluated using enhanced sampling. Zigzag-L is the only interface where proton transfer from adsorbed water to TiS 2 surface S atoms is thermodynamically and kinetically favored. The coexistence of surface four-fold-coordinated Ti (Ti 4c ) and one-fold-coordinated S (S 1c ) is found to be essential to making proton transfer feasible on the Zigzag-L surface. Furthermore, remaining unprotonated S 1c atoms can act as good proton acceptors after water dissociation. Thus, TiS 2 with Zigzag-L termination may be a surface to avoid in CDI device construction, given that pH fluctuations adversely affect performance. Furthermore, this work provides new understanding of TiS 2 /H 2 O interfacial features that could aid future design and optimization of TiS 2 -based CDI devices for water desalination.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

The Misrepresentation of the Southern African Easterly Jet in Models and Its Implications for Aerosol, Clouds, and Precipitation Distributions

The southern African easterly jet (AEJ-S) is an important midtropospheric feature critical to understanding the tropical convective system over central Africa and the aerosol–cloud interactions over the southeast Atlantic Ocean. However, it remains unclear how well models represent the AEJ-S and its influence on aerosol transport, clouds, and precipitation distribution. Here, in this work, we use ground- and satellite-based observations and reanalysis datasets to assess the representation of AEJ-S in the Coupled Model Intercomparison Project phase 6 (CMIP6) models between September and October during the peak of midtropospheric winds, aerosol transport, clouds, and precipitation. We find that most CMIP6 models have difficulty accurately simulating the strength, position, and spatial distribution of the AEJ-S. Specifically, the AEJ-S is relatively weaker and at a slightly lower altitude in the ensemble of CMIP6 models than represented by observation and reanalysis datasets. To assess the influence of the misrepresented the AEJ-S on CMIP6-simulated aerosol, clouds, and precipitation distributions, we performed composite analyses using models with low and high biases based on the estimates of their midtropospheric easterly wind speed. We find that the misrepresentation of the AEJ-S in CMIP6 models is associated with the overestimation of clouds and precipitation over central Africa, the underestimation of clouds over the southeast Atlantic Ocean, and the limitation of aerosol transport over the continent or the deviation of its spatial distribution from the typical zonal transport over the Atlantic Ocean. Because aerosols, clouds, and precipitation are important components of the regional climate system, we conclude that accurate representation of the AEJ-S is essential over central Africa and the southeast Atlantic Ocean.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Machine Learning in the Context of Laser-Induced Breakdown Spectroscopy

The integration of machine learning (ML) with Laser-Induced Breakdown Spectroscopy (LIBS) has revolutionized the analytical capabilities of LIBS. The combi-nation of both methods enables more accurate and efficient data analysis. While LIBS itself is a powerful technique for elemental analysis, the vast amount of spectral data it generates can be hard to interpret. Machine learning addresses these challenges by leveraging algorithms that can learn from data, identify patterns, and make predictions without explicit programming for the interpretation of each specific task. In LIBS application, ML techniques are used to enhance various analytical processes. For example, ML algorithms can classify materials based on their spectral fingerprints, predict the concentration of elements in a sample, and identify underlying patterns within complex datasets. Here, this application improves the precision of LIBS analyses while significantly reducing the time required for data processing and interpretation. In this chapter, the fundamental concepts of ML will be discussed first. Following this, the process of data splitting and the importance of feature selection will be examined. Several machine learning methods will then be closely examined, exploring how each can benefit LIBS analysis and highlighting their respective advantages and shortcomings. This structured approach will provide a comprehensive understanding of the integration of ML in the context of LIBS analysis.

47 OTHER INSTRUMENTATION↗
DOE OSTI2024

Analyzing the impact of design factors on solar module thermomechanical durability using interpretable machine learning techniques

Solar modules in utility-scale systems are expected to maintain decades of lifetime to rival conventional energy sources. However, cyclic thermomechanical loading often degrades their long-term performance, highlighting the importance of effective design to mitigate thermal expansion mismatches between module materials. Given the complex composition of solar modules, isolating the impact of individual components on overall durability remains a challenging task. In this work, we analyze a comprehensive data set that comprises bill-of-materials (BOM) and thermal cycling power loss from 251 distinct module designs to identify the predominant design factors and their impacts on the thermomechanical durability of modules. The methodology of our analysis combines machine learning modeling (random forest) and Shapley additive explanation (SHAP) to correlate design factors with power loss and interpret the model’s decision-making. The interpretation reveals that silicon type (monocrystalline or polycrystalline), encapsulant thickness, busbar numbers, and wafer thickness predominantly influence the degradation. With lower power loss of around 0.6% on average in the SHAP analysis, monocrystalline cells present better durability than polycrystalline cells. This finding is further substantiated by statistical testing on our raw data set. The SHAP analysis also demonstrates that while thicker encapsulants lead to reduced power loss, further increasing their thickness over around 0.6 to 0.7 mm does not yield additional benefits, particularly for the front side one. In addition, other important BOM features such as the number of busbars are analyzed. This study provides a blueprint for utilizing explainable machine learning techniques in a complex material system and can potentially guide future research on optimizing the design of solar modules.

14 SOLAR ENERGY↗
DOE OSTI2021

Photometric redshifts for the S-PLUS Survey: Is machine learning up to the task?

The Southern Photometric Local Universe Survey (S-PLUS) is a novel project that aims to map the Southern Hemisphere using a twelve filter system, comprising five broad-band SDSS-like filters and seven narrow-band filters optimized for important stellar features in the local universe. In this paper we use the photometry and morphological information from the first S-PLUS data release (S-PLUS DR1) cross-matched to unWISE data and spectroscopic redshifts from Sloan Digital Sky Survey DR15. We explore three different machine learning methods (Gaussian Processes with GPz and two Deep Learning models made with TensorFlow) and compare them with the currently used template-fitting method in the S-PLUS DR1 to address whether machine learning methods can take advantage of the twelve filter system for photometric redshift prediction. Using tests for accuracy for both single-point estimates such as the calculation of the scatter, bias, and outlier fraction, and probability distribution functions (PDFs) such as the Probability Integral Transform (PIT), the Continuous Ranked Probability Score (CRPS) and the Odds distribution, we conclude that a deep-learning method using a combination of a Bayesian Neural Network and a Mixture Density Network offers the most accurate photometric redshifts for the current test sample. In conclusion, it achieves single-point photometric redshifts with scatter (σ NMAD ) of 0.023, normalized bias of -0.001, and outlier fraction of 0.64% for galaxies with r_auto magnitudes between 16 and 21.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Evaluation of artificial neural network performance for classification of potato plants infected with potato virus Y using spectral data on multiple varieties and genotypes

Potato virus Y (Potyviridae, PVY) is a plant virus that poses a significant threat to potato producers on a global basis. The pathogen has disrupted seed potato supplies and negatively impacted yield and quality of commercial potato crops. The potato industry currently manages PVY infection levels via insecticide applications, regional seed certification programs that rely on field scouting to visually assess individual plants for infection status, and destructive and costly tissue sampling coupled with laboratory assays. Despite these efforts, PVY continues to confound potato industry stakeholders resulting in economic harm. Remote sensing and machine learning provide for the development of new tools to more accurately detect and spatially quantify PVY-infected plants versus the current state of the art. However, there is a need to understand how the occurrence of many different potato varieties impact the dynamics of developing models to detect potato plants impacted with PVY and their potential effectiveness. This study evaluates classification modelling outcomes using spectral datasets collected in different temporal and spatial environments (greenhouse and a production field) on multiple potato varieties consisting of labelled instances of plants infected with PVY and those not infected with the virus. A modelling framework was developed to support iterative modelling runs using artificial neural network (ANN) architectures configured as binary classifiers to develop sample populations to support statistical analysis on model performance using specific spectral subsets. When using spectral data to detect PVY-infected plants, ANN models achieved the highest mean accuracy of 0.894 on a single variety. Conversely, the same ANN model architecture only achieved a mean accuracy of 0.575 on a spectral data set representing 29 potato breeding lines. Additionally, statistical analysis indicates spectral regions including the red edge, near infrared and shortwave infrared contain more important spectral features for the ANN classifier introduced in this research.

60 APPLIED LIFE SCIENCES↗