Search NASA⌕ Search

SEARCH · Search NASA

Results for “ensemble optimization”

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 91 records · Page 5

Estimating Uncertainty in Simulated ENSO Statistics

Abstract Large ensembles of model simulations are frequently used to reduce the impact of internal variability when evaluating climate models and assessing climate change induced trends. However, the optimal number of ensemble members required to distinguish model biases and climate change signals from internal variability varies across models and metrics. Here we analyze the mean, variance and skewness of precipitation and sea surface temperature in the eastern equatorial Pacific region often used to describe the El Niño–Southern Oscillation (ENSO), obtained from large ensembles of Coupled model intercomparison project phase 6 climate simulations. Leveraging established statistical theory, we develop and assess equations to estimate, a priori, the ensemble size or simulation length required to limit sampling‐based uncertainties in ENSO statistics to within a desired tolerance. Our results confirm that the uncertainty of these statistics decreases with the square root of the time series length and/or ensemble size. Moreover, we demonstrate that uncertainties of these statistics are generally comparable when computed using either pre‐industrial control or historical runs. This suggests that pre‐industrial runs can sometimes be used to estimate the expected uncertainty of statistics computed from an existing historical member or ensemble, and the number of simulation years (run duration and/or ensemble size) required to adequately characterize the statistic. This advance allows us to use existing simulations (e.g., control runs that are performed during model development) to design ensembles that can sufficiently limit diagnostic uncertainties arising from simulated internal variability. These results may well be applicable to variables and regions beyond ENSO.

54 ENVIRONMENTAL SCIENCES↗

Scalable Comparative Visualization of Ensembles of Call Graphs

Optimizing the performance of large-scale parallel codes is critical for efficient utilization of computing resources. Code developers often explore various execution parameters, such as hardware configurations, system software choices, and application parameters, and are interested in detecting and understanding bottlenecks in different executions. They often collect hierarchical performance profiles represented as call graphs, which combine performance metrics with their execution contexts. The crucial task of exploring multiple call graphs together is tedious and challenging because of the many structural differences in the execution contexts and significant variability in the collected performance metrics (e.g., execution runtime). In this paper, we present Ensemble CallFlow to support the exploration of ensembles of call graphs using new types of visualizations, analysis, graph operations, and features. We introduce ensemble-Sankey , a new visual design that combines the strengths of resource-flow (Sankey) and box-plot visualization techniques. Whereas the resource-flow visualization can easily and intuitively describe the graphical nature of the call graph, the box plots overlaid on the nodes of Sankey convey the performance variability within the ensemble. Our interactive visual interface provides linked views to help explore ensembles of call graphs, e.g., by facilitating the analysis of structural differences, and identifying similar or distinct call graphs. Finally, we demonstrate the effectiveness and usefulness of our design through case studies on large-scale parallel codes.

97 MATHEMATICS AND COMPUTING↗

Inference-Optimized AI and High Performance Computing for Gravitational Wave Detection at Scale

We introduce an ensemble of artificial intelligence models for gravitational wave detection that we trained in the Summit supercomputer using 32 nodes, equivalent to 192 NVIDIA V100 GPUs, within 2 h. Once fully trained, we optimized these models for accelerated inference using NVIDIA TensorRT. We deployed our inference-optimized AI ensemble in the ThetaGPU supercomputer at Argonne Leadership Computer Facility to conduct distributed inference. Using the entire ThetaGPU supercomputer, consisting of 20 nodes each of which has 8 NVIDIA A100 Tensor Core GPUs and 2 AMD Rome CPUs, our NVIDIA TensorRT-optimized AI ensemble processed an entire month of advanced LIGO data (including Hanford and Livingston data streams) within 50 s. Our inference-optimized AI ensemble retains the same sensitivity of traditional AI models, namely, it identifies all known binary black hole mergers previously identified in this advanced LIGO dataset and reports no misclassifications, while also providing a 3X inference speedup compared to traditional artificial intelligence models. We used time slides to quantify the performance of our AI ensemble to process up to 5 years worth of advanced LIGO data. In this synthetically enhanced dataset, our AI ensemble reports an average of one misclassification for every month of searched advanced LIGO data. We also present the receiver operating characteristic curve of our AI ensemble using this 5 year long advanced LIGO dataset. This approach provides the required tools to conduct accelerated, AI-driven gravitational wave detection at scale.

97 MATHEMATICS AND COMPUTING↗

NCAR Contribution to A U.S. National Multi-Model Ensemble (NMME) ISI Prediction System

NCAR brought the latest version of the Community Earth System Model (version 1, CESM1) into the mix of models in the NMME effort. This new version uses our newest atmospheric model CAM5 and produces a coupled climate and ENSO that are generally as good or better than those of the Community Climate System Model version 4 (CCSM4). Compared to CCSM4, the new coupled model has a superior climate response with respect to low clouds in both the subtropical stratus regimes and the Arctic. However, CESM1 has been run to date using a prognostic aerosol model that more than doubles its computational cost. We are currently evaluating a version of the new model using prescribed aerosols and expect it will be ready for integrations in summer 2012. Because of this NCAR has not been able to complete the hindcast integrations using the NCAR loosely-coupled ensemble Kalman filter assimilation method nor has it contributed to the current (Stage I) NMME operational utilization. The expectation is that this model will be included in the NMME in late 2012 or early 2013. The initialization method will utilize the Ensemble Kalman Filter Assimilation methods developed at NCAR using the Data Assimilation Research Testbed (DART) in conjunction with Jeff Anderson’s team in CISL. This methodology has been used in our decadal prediction contributions to CMIP5. During the course of this project, NCAR has setup and performed all the needed hindcast and forecast simulations and provide the requested fields to our collaborators. In addition, NCAR researchers have participated fully in research themes (i) and (ii). Specifically, i) we have begun to evaluate and optimize our system in hindcast mode, focusing on the optimal number of ensemble members, methodologies to recalibrate individual dynamical models, and accessing our forecasts across multiple time scales, i.e., beyond two weeks, and ii) we have begun investigation of the role of different ocean initial conditions in seasonal forecasts. The completion of the calibration hindcasts for Seasonal to Interannual (SI) predictions and the maintenance of the data archive associated with the NCAR portion of this effort has been the responsibility of the Project Scientist I (Alicia Karspeck) that was partially supported on this project.

54 ENVIRONMENTAL SCIENCES↗

Design of photovoltaics for modules with 50% efficiency

Abstract We describe a spectrum splitting solar module design approach using ensembles of 2–20 subcells with bandgaps optimized for the AM 1.5D spectrum. Device physics calculations and experimental data determine radiative efficiency parameters for III ‐V compound semiconductor subcells and enable modification of conventional detailed balance calculations to predict module efficiency while retaining computational speed for a wide search of the design space. Accounting for nonideal absorption and recombination rates due to realistic material imperfections allows us to identify the minimum subcell quantity, quality, electrical connection configuration, and concentration required for 50% module efficiency with realistic optical losses and modeled contact resistance losses. We predict a module efficiency of 50% or greater will be possible with 7–10 electrically independent subcells in a spectral splitting optic at 300–500 suns concentration, assuming a 90% optical efficiency and 98% electrical efficiency, provided the subcells can achieve an average external radiative efficiency of 3–5% and a short circuit current that is at least 90% of the ideal. In examining spectrum splitting solar cells with both series‐connected and electrically independent subcells, we identify a new design trade‐off independent of the challenges of fabricating optimal bandgap combinations. Series‐connected ensembles, having a single set of electrical contacts, are less sensitive to lumped series resistance losses than ensembles where each subcells are contacted independently. By contrast, ensembles with electrically independent subcells can achieve lower radiative losses when the subcells are designed for good optical confinement. Distributing electrically independent subcells in a concentrating receiver module allows flexibility in subcell selection and fabrication, and can achieve ultra‐high efficiency with conventional III ‐V cell technology.

Warmann, Emily C.↗

Ion-Stockmayer clusters: Minima, classical thermodynamics, and variational ground state estimates of Li{sup +}(CH{sub 3}NO{sub 2}){sub n} (n = 1–20)

Structural optimizations, classical NVT ensemble, and variational Monte Carlo simulations of ion Stockmayer clusters parameterized to approximate the Li{sup +}(CH{sub 3}NO{sub 2}){sub n} (n = 1–20) systems are performed. The Metropolis algorithm enhanced by the parallel tempering strategy is used to measure internal energies and heat capacities, and a parallel version of the genetic algorithm is employed to obtain the most important minima. The first solvation sheath is octahedral and this feature remains the dominant theme in the structure of clusters with n ≥ 6. The first “magic number” is identified using the adiabatic solvent dissociation energy, and it marks the completion of the second solvation layer for the lithium ion-nitromethane clusters. It corresponds to the n = 18 system, a solvated ion with the first sheath having octahedral symmetry, weakly bound to an eight-membered and a four-membered ring crowning a vertex of the octahedron. Variational Monte Carlo estimates of the adiabatic solvent dissociation energy reveal that quantum effects further enhance the stability of the n = 18 system relative to its neighbors.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Final Report: Grant DE-FG02-05ER15682. Simulation of Complex Microphase Formation in Pure and Nanoparticle-filled Diblock Copolymers

The goal of this project was to use molecular simulation to quantify the impact of additives on the onset and structure of bicontinuous phases in linear diblock copolymers (DBC). The focus was on understanding how additives with selective affinity for a given block will distribute and perturb the structure of complex bicontinuous phases (like gyroid, double diamond, and plumbers nightmare whose minority component block forms two interweaving 3D networks) in DBCs; it was hypothesized that a suitable choice of additive type, size, affinity, and concentration may suppress or stabilize a particular bicontinuous phase. The ultimate goal in this line of investigation is to elucidate the rational design of the optimal additive for which the composition range of stability of a particular bicontinuous phase is maximized. Ours are the first published simulation studies to report on the formation of the gyroid phase in DBC melts and of other bicontinuous phases in DBC-modified by homopolymer. The following tasks were carried out: (i) simulation of bicontinuous phases of pure DBCs via both on-lattice Monte Carlo simulations and continuum-space Monte Carlo and molecular dynamics simulations, (ii) determination of the effect of selective additives (homopolymer) of different sizes on such bicontinuous phases, and (iii) development of novel Monte Carlo methods to map out reliable phase diagrams and improve ergodic sampling; in particular, optimized expanded-ensemble techniques for measuring free-energies and for chemical potential equilibration.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Optimizing a dynamical decoupling protocol for solid-state electronic spin ensembles in diamond

In this study, we demonstrate significant improvements of the spin coherence time of a dense ensemble of nitrogen-vacancy (NV) centers in diamond through optimized dynamical decoupling (DD). Cooling the sample down to 77 K suppresses longitudinal spin relaxation T 1 effects and DD microwave pulses are used to increase the transverse coherence time T 2 from ~0.7ms up to ~30ms. Furthermore, we extend previous work of single-axis (Carr-Purcell-Meiboom-Gill) DD towards the preservation of arbitrary spin states. Following a theoretical and experimental characterization of pulse and detuning errors, we compare the performance of various DD protocols. We also identify that the optimal control scheme for preserving an arbitrary spin state is a recursive protocol, the concatenated version of the XY8 pulse sequence. The improved spin coherence might have an immediate impact on improvements of the sensitivities of ac magnetometry. Moreover, the protocol can be used on denser diamond samples to increase coherence times up to NV-NV interaction time scales, a major step towards the creation of quantum collective NV spin states.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Ensemble-Based Support Vector Regression with Gravitational Search Algorithm Optimization for Estimating Magnetic Relative Cooling Power of Manganite Refrigerant in Magnetic Refrigeration Application

Magnetic refrigeration technology (MRT) is considered an energy-efficient and environmental-friendly system of refrigeration that has a considerable potential of replacing the classical gas-compression expansion method of refrigeration. Inclusion of manganite-based material (MBM) in MRT as a magnetic refrigerant has attracted significant attention recently due to cost effectiveness of the refrigerant as well as better resistance to oxidation and corrosion as compared to the commonly used metal gadolinium refrigerant. Relative cooling power (RCP) is one of the most important parameters to be considered while assessing the usefulness of MBM. Its value can be altered through doping with external materials and accurate estimation of the dopant influence is required to achieve the right amount of RCP. This present research work proposes support vector regression (SVR) ensemble models with gravitational search algorithm (GSA) hyper-parameters optimization, for estimating RCP of MBM and to determine the influence of dopants on RCP using ionic radii and dopant concentrations as descriptors. GSA-SVR ensemble model (GSE) is developed by employing the outputs of five different SVR models as descriptors while GSA-SVR ensemble model with averaging (GSEA) uses the average of the five different SVR models as its descriptor. The novel ensemble models outperform other SVR models, specifically; GSE performs better than GSA-SVR model and the conventional SVR model with performance improvement of 269.14% and 283.61%, respectively on the basis of root mean square error (RMSE). Furthermore, GSEA outperforms GSE, GSA-SVR model and conventional SVR with performance improvement of 27.51%, 370.70%, and 389.14%, respectively on the basis of RMSE. The developed GSE and GSEA also perform better than the existing RCP model in the literature with performance improvement of 11.53% and 42.21%, respectively. The results of this research work will not only serve to circumvent the experimental challenges of RCP measurement without loss of experimental precision but also further promotes environment-friendly system of refrigeration.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Stereoselective isomerization of an ensemble of adsorbed molecules with multiple orientations: Stochastic laser pulse optimization for selective switching between achiral and chiral atropisomers

We present quantum dynamical simulations for the laser driven isomerization of an ensemble of surface mounted stereoisomers with multiple orientations. The model system 1-(2-cis-fluoroethenyl)-2-fluorobenzene supports two chiral and one achiral atropisomers upon torsion around the C-C single bond connecting phenyl ring and ethylene group. An infrared picosecond pulse is used to excite the internal rotation around the chiral axis, thereby controlling the chirality of the molecule. In order to selectively switch the molecules--independent of their orientation on a surface--from their achiral to either their left- or right-handed form, a stochastic pulse optimization algorithm is applied. The stochastic pulse optimization is performed for different sets of defined orientations of adsorbates corresponding to the rotational symmetry of the surface. The obtained nonlinearly polarized laser pulses are highly enantioselective for each orientation.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Detection of small atom numbers through image processing

We demonstrate improved detection of small trapped atomic ensembles through advanced postprocessing and optimal analysis of absorption images. A fringe-removal algorithm reduces imaging noise to the fundamental photon-shot-noise level and proves beneficial even in the absence of fringes. A maximum-likelihood estimator is then derived for optimal atom-number estimation in well-localized ensembles and is applied to real experimental data to measure the population differences and intrinsic atom shot noise between spatially separated ensembles each comprising between 10 and 2000 atoms. The combined techniques improve our signal-to-noise ratio by a factor of 3, to a minimum resolvable population difference of 17 atoms, close to our ultimate detection limit.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Inverse design of photonic surfaces via multi fidelity ensemble framework and femtosecond laser processing

We demonstrate a multi-fidelity (MF) machine learning ensemble framework for the inverse design of photonic surfaces, trained on a dataset of 11,759 samples that we fabricate using high throughput femtosecond laser processing. The MF ensemble combines an initial low fidelity model for generating design solutions, with a high fidelity model that refines these solutions through local optimization. The combined MF ensemble can generate multiple disparate sets of laser-processing parameters that can each produce the same target input spectral emissivity with high accuracy (root mean squared errors < 2%). SHapley Additive exPlanations analysis shows transparent model interpretability of the complex relationship between laser parameters and spectral emissivity. Finally, the MF ensemble is experimentally validated by fabricating and evaluating photonic surface designs that it generates for improved efficiency energy harvesting devices. Our approach provides a powerful tool for advancing the inverse design of photonic surfaces in energy harvesting applications.

97 MATHEMATICS AND COMPUTING↗

International challenge to predict the impact of radioxenon releases from medical isotope production on a comprehensive nuclear test ban treaty sampling station

Abstract The International Monitoring System (IMS) is part of the verification regime for the Comprehensive Nuclear-Test-Ban-Treaty Organization (CTBTO). At entry-into-force, half of the 80 radionuclide stations will be able to measure concentrations of several radioactive xenon isotopes produced in nuclear explosions, and then the full network may be populated with xenon monitoring afterward (Bowyer et al., 2013). Fission-based production of 99Mo for medical purposes also releases radioxenon isotopes to the atmosphere (Saey, 2009). One of the ways to mitigate the effect of emissions from medical isotope production is the use of stack monitoring data, if it were available, so that the effect of radioactive xenon emissions could be subtracted from the effect from a presumed nuclear explosion, when detected at an IMS station location. To date, no studies have addressed the impacts the time resolution or data accuracy of stack monitoring data have on predicted concentrations at an IMS station location. Recently, participants from seven nations used atmospheric transport modeling to predict the time-history of 133Xe concentration measurements at an IMS station in Germany using stack monitoring data from a medical isotope production facility in Belgium. Participants received only stack monitoring data and used the atmospheric transport model and meteorological data of their choice. Some of the models predicted the highest measured concentrations quite well (a high composite statistical model comparison rank or a small mean square error with the measured values). The results suggest release data on a 15 min time spacing is best. The model comparison rank and ensemble analysis suggests that combining multiple models may provide more accurate predicted concentrations than any single model. Further research is needed to identify optimal methods for selecting ensemble members and those methods may depend on the specific transport problem. None of the submissions based only on the stack monitoring data predicted the small measured concentrations very well. The one submission that best predicted small concentrations also included releases from nuclear power plants. Modeling of sources by other nuclear facilities with smaller releases than medical isotope production facilities may be important in discriminating those releases from releases from a nuclear explosion.

133Xe↗

The Quantum Approximation Optimization Algorithm for MaxCut: A Fermionic View

Farhi et al. recently proposed a class of quantum algorithms, the Quantum Approximate Optimization Algorithm (QAOA), for approximately solving combinatorial optimization problems. A level-p QAOA circuit consists of steps in which a classical Hamiltonian, derived from the cost function, is applied followed by a mixing Hamiltonian. The 2p times for which these two Hamiltonians are applied are the parameters of the algorithm. As p increases, however, the parameter search space grows quickly. The success of the QAOA approach will depend, in part, on finding effective parameter-setting strategies. Here, we analytically and numerically study parameter setting for QAOA applied to MAXCUT. For level-1 QAOA, we derive an analytical expression for a general graph. In principle, expressions for higher p could be derived, but the number of terms quickly becomes prohibitive. For a special case of MAXCUT, the Ring of Disagrees, or the 1D antiferromagnetic ring, we provide an analysis for arbitrarily high level. Using a Fermionic representation, the evolution of the system under QAOA translates into quantum optimal control of an ensemble of independent spins. This treatment enables us to obtain analytical expressions for the performance of QAOA for any p. It also greatly simplifies numerical search for the optimal values of the parameters. By exploring symmetries, we identify a lower-dimensional sub-manifold of interest; the search effort can be accordingly reduced. This analysis also explains an observed symmetry in the optimal parameter values. Further, we numerically investigate the parameter landscape and show that it is a simple one in the sense of having no local optima.

quantum algorithm↗

Quantum mechanics/molecular mechanics minimum free-energy path for accurate reaction energetics in solution and enzymes: Sequential sampling and optimization on the potential of mean force surface

To accurately determine the reaction path and its energetics for enzymatic and solution-phase reactions, we present a sequential sampling and optimization approach that greatly enhances the efficiency of the ab initio quantum mechanics/molecular mechanics minimum free-energy path (QM/MM-MFEP) method. In the QM/MM-MFEP method, the thermodynamics of a complex reaction system is described by the potential of mean force (PMF) surface of the quantum mechanical (QM) subsystem with a small number of degrees of freedom, somewhat like describing a reaction process in the gas phase. The main computational cost of the QM/MM-MFEP method comes from the statistical sampling of conformations of the molecular mechanical (MM) subsystem required for the calculation of the QM PMF and its gradient. In our new sequential sampling and optimization approach, we aim to reduce the amount of MM sampling while still retaining the accuracy of the results by first carrying out MM phase-space sampling and then optimizing the QM subsystem in the fixed-size ensemble of MM conformations. The resulting QM optimized structures are then used to obtain more accurate sampling of the MM subsystem. This process of sequential MM sampling and QM optimization is iterated until convergence. The use of a fixed-size, finite MM conformational ensemble enables the precise evaluation of the QM potential of mean force and its gradient within the ensemble, thus circumventing the challenges associated with statistical averaging and significantly speeding up the convergence of the optimization process. To further improve the accuracy of the QM/MM-MFEP method, the reaction path potential method developed by Lu and Yang [Z. Lu and W. Yang, J. Chem. Phys. 121, 89 (2004)] is employed to describe the QM/MM electrostatic interactions in an approximate yet accurate way with a computational cost that is comparable to classical MM simulations. The new method was successfully applied to two example reaction processes, the classical S{sub N}2 reaction of Cl{sup -}+CH{sub 3}Cl in solution and the second proton transfer step of the reaction catalyzed by the enzyme 4-oxalocrotonate tautomerase. The activation free energies calculated with this new sequential sampling and optimization approach to the QM/MM-MFEP method agree well with results from other simulation approaches such as the umbrella sampling technique with direct QM/MM dynamics sampling, demonstrating the accuracy of the iterative QM/MM-MFEP method.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Selecting a Classification Ensemble and Detecting Process Drift in an Evolving Data Stream

We characterize the commercial behavior of a group of companies in a common line of business using a small ensemble of classifiers on a stream of records containing commercial activity information. This approach is able to effectively find a subset of classifiers that can be used to predict company labels with reasonable accuracy. Performance of the ensemble, its error rate under stable conditions, can be characterized using an exponentially weighted moving average (EWMA) statistic. The behavior of the EWMA statistic can be used to monitor a record stream from the commercial network and determine when significant changes have occurred. Results indicate that larger classification ensembles may not necessarily be optimal, pointing to the need to search the combinatorial classifier space in a systematic way. Results also show that current and past performance of an ensemble can be used to detect when statistically significant changes in the activity of the network have occurred. The dataset used in this work contains tens of thousands of high level commercial activity records with continuous and categorical variables and hundreds of labels, making classification challenging.

Ensemble classifiers↗

Robust Optimized Pulse Schemes for Atomic Fountain Interferometry

The robustness of an atomic fountain interferometer with respect to variations in the initial velocity of the atoms and deviations from the optimal pulse amplitude is examined. We numerically simulate the dynamics of an interferometer in momentum space with a maximum separation of and map out the expected signal contrast depending on the variance of the initial velocity distribution and the value of the laser field amplitude. We show that an excitation scheme based on rapid adiabatic passage significantly enhances the expected signal contrast, compared to the commonly used scheme consisting of a series of and pulses. We demonstrate further substantial increase of the robustness by using optimal control theory to identify splitting and swapping pulses that perform well on an ensemble average of pulse amplitudes and velocities. Our results demonstrate the ability of optimal control to significantly enhance future implementations of atomic fountain interferometry.

74 ATOMIC AND MOLECULAR PHYSICS↗