Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “data clustering”

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

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

At least 1,477 records · Page 82
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Estimating cluster masses from SDSS multiband images with transfer learning

ABSTRACT The total masses of galaxy clusters characterize many aspects of astrophysics and the underlying cosmology. It is crucial to obtain reliable and accurate mass estimates for numerous galaxy clusters over a wide range of redshifts and mass scales. We present a transfer-learning approach to estimate cluster masses using the ugriz-band images in the SDSS Data Release 12. The target masses are derived from X-ray or SZ measurements that are only available for a small subset of the clusters. We designed a semisupervised deep learning model consisting of two convolutional neural networks. In the first network, a feature extractor is trained to classify the SDSS photometric bands. The second network takes the previously trained features as inputs to estimate their total masses. The training and testing processes in this work depend purely on real observational data. Our algorithm reaches a mean absolute error (MAE) of 0.232 dex on average and 0.214 dex for the best fold. The performance is comparable to that given by redMaPPer, 0.192 dex. We have further applied a joint integrated gradient and class activation mapping method to interpret such a two-step neural network. The performance of our algorithm is likely to improve as the size of training data set increases. This proof-of-concept experiment demonstrates the potential of deep learning in maximizing the scientific return of the current and future large cluster surveys.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2025

Reward Driven Workflows for Unsupervised Explainable Analysis of Phases and Ferroic Variants From Atomically Resolved Imaging Data

Rapid progress in aberration corrected electron microscopy necessitates development of robust methods for the identification of phases, ferroic variants, and other pertinent aspects of materials structure from imaging data. While unsupervised methods for clustering and classification are widely used for these tasks, their performance can be sensitive to hyperparameter selection in the analysis workflow. In this study, the effects of descriptors and hyperparameters are explored on the capability of unsupervised ML methods to distill local structural information, exemplified by the discovery of polarization and lattice distortion in Sm − dopped BiFeO 3 (BFO) thin films. It is demonstrated that a reward-driven approach can be used to optimize these key hyperparameters across the full workflow, where rewards are designed to reflect domain wall continuity and straightness, ensuring that the analysis aligns with the material's physical behavior. This approach allows the discovery of local descriptors that are best aligned with the specific physical behavior, providing insight into the fundamental physics of materials. The reward driven workflow is further extended to disentangle structural factors of variation via an optimized variational autoencoder (VAE). Lastly, the importance of well-defined rewards is explored as a quantifiable measure of the success of the workflow.

Barakati, Kamyar [University of Tennessee, Knoxvil↗
DOE OSTI2021

Renewable hydrogen and ammonia for combined heat and power systems in remote locations: Optimal design and scheduling

Abstract Using hydrogen (H ) and ammonia (NH ) for renewable energy storage has the potential to enable economical power and heat supply with high renewable penetrations, especially in remote locations which are characterized by high energy costs. In this work we assess the economic competitiveness of renewable combined heat and power (CHP) systems in Mahaka HI, Nantucket MA, and Northwest Arctic Borough (NWAB) AK by optimally designing these systems for scenarios in which power and heat can be purchased over a range of historical energy prices as well as when 100% renewable supply is required. We use a combined optimal design and scheduling model which minimizes annualized net present cost by determining optimal technology selection and size simultaneously with optimal schedules for each period of a system operating horizon aggregated from full year hourly resolution data via a consecutive temporal clustering algorithm. We find that renewable generation meets at least 85% of power demands and 75% of heat demands under the lowest energy prices investigated. Higher conventional energy prices lead to increased renewable penetration which is facilitated by renewable NH as a seasonal energy storage medium, as are 100% renewable CHP systems. NH is used for power generation with heat cogeneration in all three locations, as well as directly for heating in NWAB. On an annual cost basis, NH ‐enabled 100% renewable CHP is only 3% more expensive in Mahaka and NWAB than systems which can purchase energy at the lowest prices, while it is 15% more expensive in Nantucket.

Palys, Matthew J.↗
DOE OSTI2023

Learning and Predicting Photonic Responses of Plasmonic Nanoparticle Assemblies via Dual Variational Autoencoders

In this work, the application of machine learning is demonstrated for rapid and accurate extraction of plasmonic particles cluster geometries from hyperspectral image data via a dual variational autoencoder (dual-VAE). In this approach, the information is shared between the latent spaces of two VAEs acting on the particle shape data and spectral data, respectively, but enforcing a common encoding on the shape-spectra pairs. It is shown that this approach can establish the relationship between the geometric characteristics of nanoparticles and their far-field photonic responses, demonstrating that hyperspectral darkfield microscopy can be used to accurately predict the geometry (number of particles, arrangement) of a multiparticle assemblies below the diffraction limit in an automated fashion with high fidelity (for monomers (0.96), dimers (0.86), and trimers (0.58). This approach of building structure-property relationships via shared encoding is universal and should have applications to a broader range of materials science and physics problems in imaging of both molecular and nanomaterial systems.

variational autoencoder↗
DOE OSTI2020

Metabolic Pathway Engineering

Modern microbial and enzyme engineering and their advancement are increasingly dependent on the marriage of a wide range of sophisticated technologies. For students entering the field of biotechnology, the outlook is indeed daunting. Expertise at levels beyond that of simple familiarity will be needed to conduct competitive research. It goes without saying that to be competitive, all new researchers in this field will need basic preparation in molecular biology, biochemistry, and genetics. Further, experience working with concepts and experimental tools in enzyme biochemistry and kinetics, gene editing, computational metabolic pathway modeling, experimental pathway flux analysis, and computational clustering tools to process complex data sets will be vital for success. We speculate that most biotechnology researchers in early career at this time will build teams of collaborators to address these disparate science fields rather than attempt to become experts in one lab.

BASIC BIOLOGICAL SCIENCES,BIOMASS FUELS↗
DOE OSTI2020

Using hydrogen and ammonia for renewable energy storage: A geographically comprehensive techno-economic study

Hydrogen and, more recently, ammonia have received worldwide attention as energy storage media. In this work we investigate the economics of using each of these chemicals as well as the two in combination for islanded renewable energy supply systems in 15 American cities representing different climate regions throughout the country. We use an optimal combined capacity planning and scheduling model which minimizes the levelized cost of energy (LCOE) by determining optimal unit selection and size along with unit commitments, production rates, and storage inventories for each period of system operation. These periods are aggregated from full year hourly resolution data via a consecutive temporal clustering algorithm. Ammonia is generally more economical than hydrogen as a single method of energy storage. Additionally, systems which use both hydrogen and ammonia outperform those which use only one storage option and have LCOE between $\$ 0.17$/kWh and $\$ 0.28$/kWh, including full investment in renewable generation infrastructure.

25 ENERGY STORAGE↗
DOE OSTI2014

Hybrid phosphazene anodes for energy storage applications

The use of hybrid cyclic phosphazene polymer/graphite anodes, where the phosphazene serves as distributed loci for Li deposition, has been investigated. Capacity within the hybrid system was found to occur reversibly in distinct regions. At the most positive voltages, above 0.06 V vs Li/Li+, the capacity was associated mostly with Li+ intercalation into graphite. In the most negative region, deposition of Li within the polymer was the predominate mechanism. A transitional region is inferred by the data whereby bulk aggregation or clustering of Li atoms occurs in proximity to the phosphazene sites that then serve as a template for more widespread population of Li within the anode at higher voltages, akin to a nucleation process. In full cells with a mixed oxide cathode, controlling the extent of Li deposition by limiting the charging voltage to 4.45 V enabled repeated cycling with no loss in capacity. Capacities as high as 183 mAh g-1 have been achieved for systems containing as little as 10% graphite while retaining coulombic efficiencies of 98% over 50 cycles. This level of cycling equates to the deposition of 7.4 Li per cyclic phosphazene.

25 ENERGY STORAGE↗
DOE OSTI2015

A cesium copper vanadyl-diphosphate: Synthesis, crystal structure and physical properties

A non-centrosymmetric orthorhombic diphosphate, Cs{sub 2}Cu{sub 1+x}(VO){sub 2−x}(P{sub 2}O{sub 7}){sub 2} (x=0.1) with a=13.7364(2) Å, b=9.2666(2) Å, c=11.5678(2) Å, Z=4, has been isolated. Its 3D framework is built from Cu atoms in square pyramidal and square planar coordination, VO{sub 5} tetragonal pyramids and P{sub 2}O{sub 7} diphosphate groups, sharing vertices. Large channels are fulfilled by cesium atoms. The ESR study reveals a similarity in behaviour of two paramagnetic (Cu and V) subsystems. The temperature dependences of the ESR linewidth and static magnetic susceptibility data present evidences for a cluster type magnetic ordering in the title compound at T⁎=22 K. The weakness of the relevant anomalies reflects presumably obvious Cu{sup 2+} ions and (VO){sup 2+} units disorder in the system. It is supposed that the charge and geometry of the framework are controlled by the Cu{sup 2+}/(VO){sup 2+} ratio; its variation may lead to a design of new materials. - Graphical abstract: A microporous 3D anionic framework of the first copper vanadium-diphosphate Cs{sub 2}Cu{sub 1.1}(VO){sub 1.9}(P{sub 2}O{sub 7}){sub 2}. The similarity in behaviour of Cu and V paramagnetic subsystems revealed by ESR study. - Highlights: • The first copper vanadium-diphosphate Cs{sub 2}Cu{sub 1.1}(VO){sub 1.9}(P{sub 2}O{sub 7}){sub 2} is reported. • A 3D anionic framework is characterized by disorder in distribution of Cu and V atoms. • Structural relations with topologically similar compounds are discussed. • The similarity in behaviour of Cu and V paramagnetic subsystems has been revealed.

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

Density Functionals for Hydrogen Storage: Defining the H2Bind275 Test Set with Ab Initio Benchmarks and Assessment of 55 Functionals

Efficient and high-capacity storage materials are indispensable for a hydrogen-based economy. In silico tools can accelerate the process of discovery of new adsorbent materials with optimal hydrogen adsorption enthalpies. Density functional theory is well-poised to become a very useful tool for enabling high-throughput screening of potential materials. In this work, we have identified density functional approximations that provide good performance for hydrogen binding applications following a two-pronged approach. First, we have compiled a data set (H2Bind275) that comprehensively represents the hydrogen binding problem capturing the chemical and mechanistic diversity in the binding sites encountered in hydrogen storage materials. We have also computed reference interaction energies for this data set using coupled-cluster theory. Second, we have assessed the performance of 55 density functional approximations for predicting H 2 interaction energies and have identified two hybrid density functionals (ωB97X-V and ωB97M-V), two double hybrid density functionals (DSD-PBEPBE-D3(BJ) and PBE0-DH), and one semilocal density functional (B97M-V) as the best performing ones. We have recommended the addition of empirical dispersion corrections to systematically underbinding density functionals such as revPBE, BLYP, and B3LYP for improvements in performance at negligible additional cost. We have also recommended the usage of the def2-TZVPP basis set as it represents a good compromise between accuracy and cost, limiting the finite basis set errors to less than 1 kJ/mol.

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

Data-Driven Many-Body Models for Molecular Fluids: CO 2 /H 2 O Mixtures as a Case Study

In this study, we extend the scope of the many-body TTM-nrg and MB-nrg potential energy functions (PEFs), originally introduced for halide ion–water and alkali-metal ion–water interactions, to the modeling of carbon dioxide (CO2) and water (H2O) mixtures as prototypical examples of molecular fluids. Both TTM-nrg and MB-nrg PEFs are derived entirely from electronic structure data obtained at the coupled cluster level of theory and are, by construction, compatible with MB-pol, a many-body PEF that has been shown to accurately reproduce the properties of water. Although both TTM-nrg and MB-nrg PEFs adopt the same functional forms for describing permanent electrostatics, polarization, and dispersion, they differ in the representation of short-range contributions, with the TTM-nrg PEFs relying on conventional Born–Mayer expressions and the MB-nrg PEFs employing multidimensional permutationally invariant polynomials. By providing a physically correct description of many-body effects at both short and long ranges, the MB-nrg PEFs are shown to quantitatively represent the global potential energy surfaces of the CO 2 –CO 2 and CO 2 –H 2 O dimers and the energetics of small clusters, as well as to correctly reproduce various properties in both gas and liquid phases. Furthermore, building upon previous studies of aqueous systems, our analysis provides further evidence for the accuracy and efficiency of the MB-nrg framework in representing molecular interactions in fluid mixtures at different temperature and pressure conditions.

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

The Rise of Neural Networks for Materials and Chemical Dynamics

Machine learning (ML) is quickly becoming a premier tool for modeling chemical processes and materials. ML-based force fields, trained on large data sets of high-quality electron structure calculations, are particularly attractive due their unique combination of computational efficiency and physical accuracy. This Perspective summarizes some recent advances in the development of neural network-based interatomic potentials. Designing high-quality training data sets is crucial to overall model accuracy. One strategy is active learning, in which new data are automatically collected for atomic configurations that produce large ML uncertainties. Another strategy is to use the highest levels of quantum theory possible. Transfer learning allows training to a data set of mixed fidelity. A model initially trained to a large data set of density functional theory calculations can be significantly improved by retraining to a relatively small data set of expensive coupled cluster theory calculations. These advances are exemplified by applications to molecules and materials.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2019

Extreme Wet-Bulb Temperatures in China: The Significant Role of Moisture

This paper investigates the extreme wet bulb temperatures (TWs) in China, with a focus on understanding the relative contributions of temperature and moisture to the extremes. Analysis of station observations shows that daily extreme temperatures (T), specific humidity (q) and TWs generally co-occur in southeastern China while extreme TW and T rarely overlap in the arid and semi-arid North and Northwest China. Overall, q contributes more than T to extreme TWs, especially in North and Northwest China. Based on the relative contributions of q and T, regional extreme TW events are classified as q-dominated and T-dominated, respectively, to study their large-scale environment. Cluster analysis of global reanalysis data shows that extreme TWs are generally accompanied by increased surface air temperature and humidity, concomitant with anomalous high-pressure and notable water vapor flux convergence. However, important differences are also seen in the large-scale environment during q-dominated and T-dominated extreme TWs. During q-dominated extreme TWs, the large-scale environment favors convection, as indicated by ascending motions, decreased downward solar radiation (DSR) as well as increased precipitable water and near surface relative humidity. In contrast, during T-dominated extreme TWs, convection is inhibited by the large-scale environment that features descending motion and increased DSR, decreased precipitable water as well as decreased relative humidity. Consistent with the contrasting environments, we demonstrate that T-dominated extreme TWs tend to last longer than q-dominated extremes. Given the significant role of moisture in extreme TWs, more research is needed to understand its impacts on heat stress now and in the future.

extreme wet bulb temperatures, Moisture, synoptic ↗
DOE OSTI2016

Toward a modular multi-material nanoparticle synthesis and assembly strategy via bionanocombinatorics: bifunctional peptides for linking Au and Ag nanomaterials

Materials-binding peptides represent a unique avenue towards controlling the shape and size of nanoparticles (NPs) grown under aqueous conditions. Here, employing a bionanocombinatorics approach, two such materials-binding peptides were linked at either end of a photoswitchable spacer, forming a multi-domain materials-binding molecule to control the in situ synthesis and organization of Ag and Au NPs under ambient conditions. These multi-domain molecules retained the peptides’ ability to nucleate, grow, and stabilize Ag and Au NPs in aqueous media. Disordered co-assemblies of the two nanomaterials were observed by TEM imaging of dried samples after sequential growth of the two metals, and showed a clustering behavior that was not observed without both metals and the linker molecules. While TEM evidence indicated the formation of AuNP/AgNP assemblies upon drying, SAXS analysis indicated that no extended assemblies existed in solution, suggesting that sample drying plays an important role in facilitating NP clustering. Molecular simulations and experimental data revealed tunable materials-binding based upon the isomerization state of the photoswitchable unit and metal employed. This work is a first step in generating externally actuated biomolecules with specific material-binding properties that could be used as the building blocks to achieve multi-material switchable NP assemblies.

36 MATERIALS SCIENCE↗
DOE OSTI2013

Elaboration of m-cresol polyamide12/ polyaniline composite films for antistatic applications

The present work deals with the preparation of transparent antistatic films from an extreme dilution of an intrinsically conducting polymer (ICP) with not coloured polymers. Our approach is based on the chemical polymerization of a very thin layer of Polyaniline (PANI) around particles of an insulating polymer (PA12). Films were obtained by dissolving the synthesized core-shell particles in m-Cresol. The electric property and structure relationships were investigated by using dielectric relaxation spectroscopy, X-ray diffraction and micro-Raman spectroscopy. Composite films exhibited a well established dc conductivity over all the frequency range for 10 wt. % of PANI concentration related to the conductive properties of the PANI clusters. X-ray diffraction data show broader and lower intensity of PA12 peaks when increasing PANI content, probably due to the additional doping effect of m- cresol. The doping of PA12/PANI films with Dodecyl benzene sulfonic acid (DBSA) was unequivocally verified by Raman spectroscopy.

36 MATERIALS SCIENCE↗
DOE OSTI2015

Ground state analytical ab initio intermolecular potential for the Cl{sub 2}-water system

The chlorine/water interface is of crucial importance in the context of atmospheric chemistry. Modeling the structure and dynamics at this interface requires an accurate description of the interaction potential energy surfaces. We propose here an analytical intermolecular potential that reproduces the interaction between the Cl{sub 2} molecule and a water molecule. Our functional form is fitted to a set of high level ab initio data using the coupled-cluster single double (triple)/aug-cc-p-VTZ level of electronic structure theory for the Cl{sub 2} − H{sub 2}O complex. The potential fitted to reproduce the three minima structures of 1:1 complex is validated by the comparison of ab initio results of Cl{sub 2} interacting with an increasing number of water molecules. Finally, the model potential is used to study the physisorption of Cl{sub 2} on a perfectly ordered hexagonal ice slab. The calculated adsorption energy, in the range 0.27 eV, shows a good agreement with previous experimental results.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2013

AN X-RAY SURVEY OF THE YOUNG STELLAR POPULATION OF THE LYNDS 1641 AND IOTA ORIONIS REGIONS

We present an XMM-Newton survey of the part of the Orion A cloud south of the Orion Nebula. This survey includes the Lynds 1641 (L1641) dark cloud, a region of the Orion A cloud with very few massive stars and hence a relatively low ambient UV flux, and the region around the O9 III star {iota} Orionis. In addition to proprietary data, we used archival XMM data of the Orion Nebula Cluster (ONC) to extend our analysis to a major fraction of the Orion A cloud. We have detected 1060 X-ray sources in L1641 and the {iota} Ori region. About 94% of the sources have Two Micron All Sky Survey and Spitzer counterparts, 204 and 23 being Class II and Class I or protostar objects, respectively. In addition, we have identified 489 X-ray sources as counterparts to Class III candidates, given they are bright in X-rays and appear as normal photospheres at mid-IR wavelengths. The remaining 205 X-ray sources are likely distant active galactic nuclei or other galactic sources not related to Orion A. We find that Class III candidates appear more concentrated in two main clusters in L1641. The first cluster of Class III stars is found toward the northern part of L1641, concentrated around {iota} Ori. The stars in this cluster are more evolved than those in the Orion Nebula. We estimate a distance of 300-320 pc for this cluster showing that it is in the foreground of the Orion A cloud. Another cluster rich in Class III stars is located in L1641 South and appears to be a slightly older cluster embedded in the Orion A cloud. Furthermore, other evolved Class III stars are found north of the ONC toward NGC 1977.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2015

The sizes of z ∼ 6–8 lensed galaxies from the Hubble Frontier Fields Abell 2744 data

We investigate the sizes of z ∼ 6–8 dropout galaxies using the complete data of the Abell 2744 cluster and parallel fields in the Hubble Frontier Fields program. By directly fitting light profiles of observed galaxies with lensing-distorted Sérsic profiles on the image plane with the glafic software, we accurately measure intrinsic sizes of 31 z ∼ 6–7 and 8 z∼8 galaxies, including those as faint as M{sub UV}≃−16.6. We find that half-light radii r{sub e} positively correlates with UV luminosity at each redshift, although the correlation is not very tight. The largest (r{sub e}>0.8 kpc) galaxies are mostly red in UV color while the smallest (r{sub e}<0.08 kpc) ones tend to be blue. We also find that galaxies with multiple cores tend to be brighter. Combined with previous results at 2.5≲z≲12, our result confirms that the average r{sub e} of bright ((0.3–1)L{sub z=3}{sup ∗}) galaxies scales as r{sub e}∝(1+z){sup −m} with m=1.24±0.1. We find that the ratio of r{sub e} to virial radius is virtually constant at 3.3 ± 0.1% over a wide redshift range, where the virial radii of hosting dark matter halos are derived based on the abundance matching. This constant ratio is consistent with the disk formation model by Mo et al. with j{sub d}∼m{sub d}, where j{sub d} and m{sub d} are the fractions of the angular momentum and mass within halos confined in the disks. A comparison with various types of local galaxies indicates that our galaxies are most similar to circumnuclear star-forming regions of barred galaxies in the sense that a sizable amount of stars are forming in a very small area.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Neural network acceleration of large-scale structure theory calculations

Here, we make use of neural networks to accelerate the calculation of power spectra required for the analysis of galaxy clustering and weak gravitational lensing data. For modern perturbation theory codes, evaluation time for a single cosmology and redshift can take on the order of two seconds. In combination with the comparable time required to compute linear predictions using a Boltzmann solver, these calculations are the bottleneck for many contemporary large-scale structure analyses. Here, we construct neural network-based surrogate models for Lagrangian perturbation theory (LPT) predictions of matter power spectra, real and redshift space galaxy power spectra, and galaxy-matter cross power spectra that attain ~ 0.1% (at one sigma) accuracy over a broad range of scales in a ωCDM parameter space. The neural network surrogates can be evaluated in approximately one millisecond, a factor of 1000 times faster than the full Boltzmann code and LPT computations. In a simulated full-shape redshift space galaxy power spectrum analysis, we demonstrate that the posteriors obtained using our surrogates are accurate compared to those obtained using the full LPT model. We make our surrogate models public at https://github.com/sfschen/EmulateLSS, so that others may take advantage of the speed gains they provide to enable rapid iteration on analysis settings, something that is essential in complex contemporary large-scale structure analyses.

79 ASTRONOMY AND ASTROPHYSICS↗