Search NASA⌕ Search

SEARCH · Search NASA

Results for “clustering methods”

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 253 records · Page 14

Preliminary Cyber-Informed Dynamic Branch Conditions for Analysis with the Dynamic Simplified Cyber MELCOR Model

Instrumentation and control of nuclear power is transforming from analog to modern digital assets. These control systems perform key safety and security functions. This transformation is occurring in new plant designs as well as in the existing fleet of plants as the operation of those plants is extended to 60 years. This transformation introduces new and unknown exploits involving both digital asset induced safety issues and security exploits. Traditional the open literature of nuclear power risk assessment tools and cyber security assessment methods have not been modified or developed to address the unique nature of cyber failure modes and of cyber security threats. Starting in 2015, Sandia National Laboratories (SNL) embarked on a two year dynamic cyber-risk informed project to facilitate the analysis of unique cyber failure modes and the time sequencing of cyber faults, both malicious and non-malicious, and impose those cyber exploits and cyber faults onto a nuclear power plant accident sequence simulator code to assess how cyber exploits and cyber faults could interact with a plant's digital instrumentation and control (DI and C) system in the absence of digital controls. This will be achieved by coupling SNL's existing dynamic scheduler code (ADAPT), nuclear accident simulator code (MELCOR), and cyber Emulytics code (SCEPTRE) to demonstrate real-time simulation of cyber exploits and their impact on automatic DI and C responses. Studying such potential time-sequenced cyber-attacks and their risks (i.e., the associated impact and the associated degree of difficulty to achieve the attack vector) as they pertain to accident management will provide a risk informed basis for developing effective cyber security controls for nuclear power. This paper will discuss the rationale behind the proposed generic cyber exploits for an abstracted Pressurized Water Reactor that has adopted a unified DI and C System. Other papers related to this project submitted to this conference discuss: - The MELCOR model for this abstracted reactor, called the Dynamic Simplified Cyber (DSC) model, - Modifications to ADAPT to: Extend the framework for multiple simulators, Incorporation of importance measures into dynamic event tree analysis, Sort and reduce the dynamic event tree based upon inputs to or outputs from the tree, and - A feasibility study for using density based clustering methods with dynamic time wrapping to cluster and prune dynamic event trees. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Identifying structural changes with unsupervised machine learning methods

In this work, unsupervised machine learning methods are used to identify structural changes using the melting point transition in classical molecular dynamics simulations as an example application of the approach. Dimensionality reduction and clustering methods are applied to instantaneous radial distributions of atomic configurations from classical molecular dynamics simulations of metallic systems over a large temperature range. Principal component analysis is used to dramatically reduce the dimensionality of the feature space across the samples using an orthogonal linear transformation that preserves the statistical variance of the data under the condition that the new feature space is linearly independent. From there, k-means clustering is used to partition the samples into solid and liquid phases through a criterion motivated by the geometry of the reduced feature space of the samples, allowing for an estimation of the melting point transition. This pattern criterion is conceptually similar to how humans interpret the data but with far greater throughput, as the shapes of the radial distributions are different for each phase and easily distinguishable by humans. The transition temperature estimates derived from this machine learning approach produce comparable results to other methods on similarly small system sizes. These results show that machine learning approaches can be applied to structural changes in physical systems.

42 ENGINEERING↗

Application of Gaussian Mixture Regression for the Correction of Low Cost PM2.5 Monitoring Data in Accra, Ghana

Low-cost sensors (LCSs) for air quality monitoring have enormous potential to improve air quality data coverage in resource-limited parts of the world such as sub-Saharan Africa. LCSs, however, are affected by environment and source conditions. To establish high-quality data, LCSs must be collocated and calibrated with reference grade PM2.5 monitors. From March 2020, a low-cost PurpleAir PM2.5 monitor was collocated with a Met One Beta Attenuation Monitor 1020 in Accra, Ghana. While previous studies have shown that multiple linear regression (MLR) and random forest regression (RF) can improve accuracy and correlation between PurpleAir and reference data, MLR and RF yielded suboptimal improvement in the Accra collocation (R2 = 0.81 and R2 = 0.81, respectively). We present the first application of Gaussian mixture regression (GMR) to air quality data calibration and demonstrate improvement over traditional methods by increasing the collocated PM2.5 correlation and accuracy to R2 = 0.88 and MAE = 2.2 μg/cu. m. Gaussian mixture models (GMMs) are a probability density estimator and clustering method from which nonlinear regressions that tolerate missing inputs can be derived. We find that even when given missing inputs, GMR provides better correlation than MLR and RF performed with complete data. GMR also allows us to estimate calibration certainty. When evaluated, 95% confidence intervals agreed with reference PM2.5 data 96% of the time, suggesting that the model accurately assesses its own confidence. Additionally, clustering within the GMM is consistent with climate characteristics, providing confidence that the calibration approach can learn underlying relationships in data.

Sensors↗

Hybrid Quantum Mechanics/Molecular Mechanics Solvation Scheme for Computing Free Energies of Reactions at Metal–Water Interfaces

We report the development of a quantum mechanics/molecular mechanics free energy perturbation (QM/MM-FEP) method for modeling chemical reactions at metal-water interfaces. This novel solvation scheme combines planewave density function theory (DFT), periodic electrostatic embedded cluster method (PEECM) calculations using Gaussian-type orbitals, and classical molecular dynamics (MD) simulations to obtain a free energy description of a complex metal-water system. We derive a potential of mean force (PMF) of the reaction system within the QM/MM framework. A fixed-size, finite ensemble of MM conformations is used to permit precise evaluation of the PMF of QM coordinates and its gradient defined within this ensemble. Local conformations of adsorbed reaction moieties are optimized using sequential MD-sampling and QM-optimization steps. An approximate reaction coordinate is constructed using a number of interpolated states and the free energy difference between adjacent states is calculated using the QM/MM-FEP method. By avoiding on-the-fly QM calculations and by circumventing the challenges associated with statistical averaging during MD sampling, a computational speedup of multiple orders of magnitude is realized. The method is systematically validated against the results of ab initio QM calculations and demonstrated for C–C cleavage in double-dehydrogenated ethylene glycol on a Pt (111) model surface.

Chemical calculations↗

Clustering with Missing Values: No Imputation Required

Clustering algorithms can identify groups in large data sets, such as star catalogs and hyperspectral images. In general, clustering methods cannot analyze items that have missing data values. Common solutions either fill in the missing values (imputation) or ignore the missing data (marginalization). Imputed values are treated as just as reliable as the truly observed data, but they are only as good as the assumptions used to create them. In contrast, we present a method for encoding partially observed features as a set of supplemental soft constraints and introduce the KSC algorithm, which incorporates constraints into the clustering process. In experiments on artificial data and data from the Sloan Digital Sky Survey, we show that soft constraints are an effective way to enable clustering with missing values.

constraints↗

Local Pair Natural Orbital-Based Coupled-Cluster Theory through Full Quadruples (DLPNO–CCSDTQ)

In this work, we implement a local pair natural orbitalbased coupled-cluster method through the full treatment of quadruple excitations (CCSDTQ). The domain-based local pair natural orbital (DLPNO) approach, which has successfully been applied to lower levels of coupled-cluster theory, is utilized in our algorithm, and thus our algorithm is called DLPNO-CCSDTQ. For simplicity in the working equations and in the implementation, we t 1 -dress the twoelectron integrals as well as Fock matrix elements. Our method can recover CCSDTQ-CCSDT and CCSDTQ-CCSDT(Q) energy differences on the order of 0.01−0.05 kcal mol −1 , even at a loose quadruples natural orbital (QNO) occupation number cutoff of 3.33 × 10 −6 . To highlight the capabilities of our code and its potential future applications, we showcase computations that would be intractable with canonical CCSDTQ, such as the benzene dimer, (H 2 O) 17 , and adamantane. With sufficient computing resources, computations up to 15 heavy atoms (40 atoms overall) may be feasible for fully bonded 3D systems.

Cluster chemistry↗

Comparison of Clustering Techniques for Residential Energy Behavior using Smart Meter Data

Current practice in whole time series clustering of residential meter data focuses on aggregated or subsampled load data at the customer level, which ignores day-to-day differences within customers. This information is critical to determine each customer’s suitability to various demand side management strategies that support intelligent power grids and smart energy management. Clustering daily load shapes provides fine-grained information on customer attributes and sources of variation for subsequent models and customer segmentation. In this paper, we apply 11 clustering methods to daily residential meter data. We evaluate their parameter settings and suitability based on 6 generic performance metrics and post-checking of resulting clusters. Finally, we recommend suitable techniques and parameters based on the goal of discovering diverse daily load patterns among residential customers. To the authors’ knowledge, this paper is the first robust comparative review of clustering techniques applied to daily residential load shape time series in the power systems’ literature.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Geometries, Binding Energies, Ionization Potentials, and Electron Affinities of Metal Clusters: Mg n 0, ± 1 , n = 1–7

Equilibrium geometries, binding energies, adiabatic ionization potentials, and adiabatic electron affinities for neutral and singly charged magnesium clusters, Mg n 0, ± 1 , n = 1–7, have been computed using 39 exchange-correlation (XC) functionals in Kohn–Sham density functional theory and several coupled-cluster methods with single, double, and triple excitations, including CCSD(T) for all species, CCSD(2)T and CR-CC(2,3) for species with n = 1–3, and CCSDt, CC(t;3), and CCSDT for species with n = 1 and 2. We have used augmented polarized–valence and polarized–core–valence correlation-consistent basis sets. We have found that the geometry and binding energy of the weakly bound Mg 2 dimer requires a robust treatment of connected triple excitations, represented in this work by the CR-CC(2,3), CC(t;3), and full CCSDT methods, which are more accurate than the popular quasi-perturbative CCSD(T) approximation, but CCSD(T) is sufficiently accurate to be applied to other Mg clusters. We have also demonstrated that for all Mg clusters examined in this study, hybrid XC functionals generally have higher accuracy than local ones, with PW6B95, SOGGA11-X, M11, and PWB6K being the most accurate, both for the geometries and for the binding energies, ionization potentials, and electron-detachment energies.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Synthesis Analysis of Mulit-Dimensional Ozone Lidar Measurements in Coastal Environments Toward Improving Ozone Simulation

Large surface ozone variations are frequently observed in coastal regions which makes the modeling of air quality in these regions challenging. The ability to properly model coastal regions relies heavily on the proper detailed observations. There have been three air quality studies (over Chesapeake Bay and the Long Island Sound) that focused on thorough measurements in these complex coastal regions; OWLETS 1 & 2 and LISTOS. Over the course of the campaigns, there were surface ozone observations taken from LIDAR (LIght Detection And Ranging) instruments for a total of 89 days at a high temporal rate. We developed a clustering method that characterizes the multiple timestamps of measurements into a range of ozone events that are specific to these coastal regions. We simulated coastal ozone for these cluster groups using WRF-GC and WRF-Chem to identify gaps within the model. Using the characterized ozone clusters, we were able to identify multiple coastal mechanisms that led to incorrect model simulations.

Claudia Bernier↗

Data–Driven Velocity Model Evaluation Using K–Means Clustering

In this work, we develop a data-driven clustering method to evaluate a velocity model using surface wave velocity dispersion. This is done by first computing theoretical dispersion curves for 1-D velocity profiles of all the grid locations and then splitting the resulting dispersion curves into a certain number of groups via the K-means clustering. The observed dispersion curves are also clustered following the same procedure and the velocity model is assessed by comparing the spatial patterns obtained for the observed and synthetic data sets. The method is applied to evaluate two community velocity models in southern California, CVM-S4.26 and CVM-H15.1, using phase velocity maps derived for 3–16 s Rayleigh waves. We found a good correlation in the spatial distribution of clusters between the result of CVM-S4.26 and that of the observed data, suggesting that the CVM-S4.26 fits the observed dispersion maps better than the CVM-H15.1 in terms of features extracted from the clustering analysis.

58 GEOSCIENCES↗

A validation of a simple model for the calculation of the ionization energies in X-ray laser-cluster interactions

The outer-ionization of an electron from a cluster is an unambiguous quantity, while the inner-ionization threshold is not, resulting in different microscopic quantum-classical hybrid models used in laser-cluster interactions. A simple local ionization threshold model for the change in the ionization energy is proposed and examined, for atoms and ions, at distances in between the initial configuration of the cluster to well into the cluster's disintegration. This model is compared with a full Hartree-Fock energy calculation which accounts for the electron correlation effects using the coupled cluster method with single and double excitations with perturbative triples (CCSD(T)). Good agreement is found between the two lending a strong theoretical support to works which rely on such models for the final and transient properties of the laser-cluster interaction.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Three-dimensional potential energy surfaces of ArNO ($\tilde{X}$ 2 Π)

Until now, the potential energy surfaces (PESs) of the ArNO complex found in the literature were two-dimensional, with the NO interatomic distance being fixed. In this work, we present the first accurate three-dimensional ground state X 2 Π PESs (both A ' and A ") of ArNO computed at the CCSD(T)/CBS level of theory. The equilibrium geometries and the well depths ( D e ) are compared to several other electronic structure methods. We found that using the multireference method, MRCI-F12 makes the surfaces much shallower (by 25%) and the depth of the surfaces does not agree with experimental data. The explicitly correlated coupled-cluster method underestimates the well depth as well. Analytic representations for both A ' and A " surfaces were fit to 4380 ab initio points to within 2.71 cm -1 .A three-dimensional Numerov propagator method in Delves coordinates is used to compute the bound state spectrum up to J tot = 6.5. The recommended dissociation energies are D 0 = 97.2 cm -1 for the adiabatic ground state and D e = 133.7 (128.1) cm -1 for A ' ( A ")

36 MATERIALS SCIENCE↗

An Orbital Meteoroid Stream Survey Using the Southern Argentina Agile Meteor Radar (SAAMER) Based on a Wavelet Approach

Over a million individually measured meteoroid orbits were collected with the Southern Argentina Agile MEteor Radar (SAAMER) between 2012-2015. This provides a robust statistical database to perform an initial orbital survey of meteor showers in the Southern Hemisphere via the application of a 3D wavelet transform. The method results in a composite year from all 4 years of data, enabling us to obtain an undisturbed year of meteor activity with more than one thousand meteors per day. Our automated meteor shower search methodology identified 58 showers. Of these showers, 24 were associated with previously reported showers from the IAU catalogue while 34 showers are new and not listed in the catalogue. Our searching method combined with our large data sample provides unprecedented accuracy in measuring meteor shower activity and description of shower characteristics in the Southern Hemisphere. Using simple modeling and clustering methods we also propose potential parent bodies for the newly discovered showers.

Meteor showers↗

A Network-Based Algorithm for Clustering Multivariate Repeated Measures Data

The National Aeronautics and Space Administration (NASA) Astronaut Corps is a unique occupational cohort for which vast amounts of measures data have been collected repeatedly in research or operational studies pre-, in-, and post-flight, as well as during multiple clinical care visits. In exploratory analyses aimed at generating hypotheses regarding physiological changes associated with spaceflight exposure, such as impaired vision, it is of interest to identify anomalies and trends across these expansive datasets. Multivariate clustering algorithms for repeated measures data may help parse the data to identify homogeneous groups of astronauts that have higher risks for a particular physiological change. However, available clustering methods may not be able to accommodate the complex data structures found in NASA data, since the methods often rely on strict model assumptions, require equally-spaced and balanced assessment times, cannot accommodate missing data or differing time scales across variables, and cannot process continuous and discrete data simultaneously. To fill this gap, we propose a network-based, multivariate clustering algorithm for repeated measures data that can be tailored to fit various research settings. Using simulated data, we demonstrate how our method can be used to identify patterns in complex data structures found in practice.

Koslovsky, Matthew↗

Explicitly correlated treatment of the Ar-NO{sup +} cation

We present an application of the recently developed explicitly correlated coupled cluster method to the generation of the three-dimensional potential energy surface (PES) of the Ar-NO{sup +} cationic complex. A good overall agreement is found with the standard coupled clusters techniques employing correlation consistent atomic basis sets (aug-cc-pVnZ, n= D, T, Q) of Wright et al. This PES is then used in quantum close-coupling scattering and variational calculations to treat the nuclear motions. The bound states energies of the Ar-NO{sup +} complex obtained by both approaches are in good agreement with the available experimental results. The analysis of the vibrational wavefunctions shows strong anharmonic resonances between the low frequency modes (intermonomer bending and stretching modes) and the wavefunctions exhibit large amplitude motions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Towards complete assignment of the infrared spectrum of the protonated water cluster H + (H 2 O) 21

The spectroscopic features of protonated water species in dilute acid solutions have been long sought after for understanding the microscopic behavior of the proton in water with gas-phase water clusters H + (H 2 O) n extensively studied as bottom-up model systems. We present a new protocol for the calculation of the infrared (IR) spectra of complex systems, which combines the fragment-based Coupled Cluster method and anharmonic vibrational quasi-degenerate perturbation theory, and demonstrate its accuracy towards the complete and accurate assignment of the IR spectrum of the H + (H 2 O) 21 cluster. The site-specific IR spectral signatures reveal two distinct structures for the internal and surface four-coordinated water molecules, which are ice-like and liquid-like, respectively. The effect of inter-molecular interaction between water molecules is addressed, and the vibrational resonance is found between the O-H stretching fundamental and the bending overtone of the nearest neighboring water molecule. The revelation of the spectral signature of the excess proton offers deeper insight into the nature of charge accommodation in the extended hydrogen-bonding network underpinning this aqueous cluster.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Clonal Selection Based Artificial Immune System for Generalized Pattern Recognition

The last two decades has seen a rapid increase in the application of AIS (Artificial Immune Systems) modeled after the human immune system to a wide range of areas including network intrusion detection, job shop scheduling, classification, pattern recognition, and robot control. JPL (Jet Propulsion Laboratory) has developed an integrated pattern recognition/classification system called AISLE (Artificial Immune System for Learning and Exploration) based on biologically inspired models of B-cell dynamics in the immune system. When used for unsupervised or supervised classification, the method scales linearly with the number of dimensions, has performance that is relatively independent of the total size of the dataset, and has been shown to perform as well as traditional clustering methods. When used for pattern recognition, the method efficiently isolates the appropriate matches in the data set. The paper presents the underlying structure of AISLE and the results from a number of experimental studies.

pattern recognition↗

Binding Energy of d¹º Transition Metals to Alkenes By Wave Function Theory and Density Functional Theory

The structures of Pd(PH₃)₂ and Pt(PH₃)₂ complexes with ethene and conjugated CnH n+2 systems (n=4, 6, 8, and 10) were studied. Their binding energies were calculated using both wave function theory (WFT) and density functional theory (DFT). Previously it was reported that the binding energy of the alkene to the transition metal does not depend strongly on the size of the conjugated C n H n+2 ligand, but that DFT methods systematically underestimate the binding energy more and more significantly as the size of the conjugated system is increased. Our results show that recently developed density functionals predict the binding energy for these systems much more accurately. New benchmark calculations carried out by the coupled cluster method based on Brueckner orbitals with double excitations and a quasiperturbative treatment of connected triple excitations (BCCD(T)) with a very large basis set agree even better with the DFT predictions than do the previous best estimates. The mean unsigned error in absolute and relative binding energies of the alkene ligands to Pd(PH₃)₂ is 2.5 kcal/mol for the ωB97 and M06 density functionals and 2.9 kcal/mol for the M06-L functional. Adding molecular mechanical damped dispersion yields even smaller mean unsigned errors: 1.3 kcal/mol for the M06-D functional, 1.5 kcal/mol for M06- L-D, and 1.8 kcal/mol for B97-D and ωB97X-D. The new functionals also lead to improved accuracy for the analogous Pt complexes. These results show that recently developed density functionals may be very useful for studying catalytic systems involving Pd d¹º centers and alkenes.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗