Search NASA⌕ Search

SEARCH · Search NASA

Results for “Constrained Gaussian process”

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 73 records · Page 4

Data-Driven Machine Learning for Wind Plant Flow Modeling

In this study, we introduce a data-driven machine learning framework for improving the accuracy of wind plant flow models by learning turbulence model corrections based on data from higher-fidelity simulations. First, a high-dimensional PDE-constrained optimization problem is solved using gradient-based optimization with adjoints to determine optimal eddy viscosity fields that improve the agreement of a medium-fidelity Reynolds-Averaged Navier Stokes (RANS) model with large eddy simulations (LES). A supervised learning problem is then constructed to find general, predictive representations of the optimal turbulence closure. A machine learning technique using Gaussian process regression is trained to predict the eddy viscosity field based on local RANS flow field information like velocities, pressures, and their gradients. The Gaussian process is trained on LES simulations of a single turbine and implemented in a wind plant simulation with 36 turbines. We show improvement over the baseline RANS model with the machine learning correction, and demonstrate the ability to provide accurate confidence levels for the corrections that enable future uncertainty quantification studies.

17 WIND ENERGY↗

Emulator-based Bayesian calibration of a subglacial drainage model

Subglacial drainage models, often motivated by the relationship between hydrology and ice flow, sensitively depend on numerous unconstrained parameters. We explore using borehole water-pressure time series to calibrate the uncertain parameters of a popular subglacial drainage model, taking a Bayesian perspective to quantify the uncertainty in parameter estimates and in the calibrated model predictions. To reduce the computation time associated with Markov Chain Monte Carlo sampling, we construct a fast Gaussian process emulator to stand in for the subglacial drainage model. We first carry out a calibration experiment using synthetic observations consisting of model simulations with hidden parameter values as a demonstration of the method. Using real borehole water pressures measured in western Greenland, we find meaningful constraints on four of the eight model parameters and a factor-of-three reduction in uncertainty of the calibrated model predictions. These experiments illustrate Gaussian process-based Bayesian inference as a useful tool for calibration and uncertainty quantification of complex glaciological models using field data. However, significant differences between the calibrated model and the borehole data suggest that structural limitations of the model, rather than poorly constrained parameters or computational cost, remain the most important constraint on subglacial drainage modelling.

58 GEOSCIENCES↗

The cool-star spectral catalog: A uniform collection of IUE SWP-LOs

Over the past decade and a half of its operations, the International Ultraviolet Explorer has recorded low-dispersion spectrograms in the 1150-2000 A interval of more than 800 stars of late spectral type (F-M). The sub-2000 A region contains a number of emission lines that are key diagnostics of physical conditions in the high-excitation chromospheres and subcoronal 'transition zones' of such stars. Many of the sources have been observed a number of times, and the available collection of SWP-LO exposures in the IUE Archives exceeds 4,000. With support from the Astrophysics Data Program, we have assembled the archival material into a catalog of IUE far-UV fluxes of late-type stars. In order to ensure uniform processing of the spectra, we: (1) photometrically corrected the raw vidicon images with a custom version of the 1985 SWP ITF; (2) identified and eliminated, sharp cosmic-ray 'hits' by means of a spatial filter; (3) extracted the spectral traces with the 'optimal' (weighted-slit) strategy; and (4) calibrated them against a well-characterized reference source, the DA white dwarf G191-B2B. Our approach is similar to that adopted by the IUE Project for its 'Final Archive', but our implementation is specialized to the case of chromospheric emission-line sources. We measured the resulting SWP-LO spectra using a semi-autonomous algorithm that establishes a smooth continuum by numerical filtering, and then fits the significant emissions (or absorptions) by means of a constrained Bevington-type multiple-Gaussian procedure. The algorithm assigns errors to the fitted fluxes - or upper limits in the absence of a significant detection - according to a model based on careful measurements of the noise properties of the IUE's intensified SEC cameras. Here, we describe the 'visualization' strategies we adopted to ensure human-review of the semi-autonomous processing and measuring algorithms; the derivation of the noise model and the assignment of errors; and the structure of the final catalog as delivered to the Astrophysics Data System.

Ayres, T.↗

Dynamical Mass Estimates of the β Pictoris Planetary System through Gaussian Process Stellar Activity Modeling

Nearly 15 yr of radial velocity (RV) monitoring and direct imaging enabled the detection of two giant planets orbiting the young, nearby star β Pictoris. The δ Scuti pulsations of the star, which overwhelm planetary signals, need to be carefully suppressed. In this work, we independently revisit the analysis of the RV data following a different approach than available in the literature to model the activity of the star. We show that a Gaussian process (GP) with a stochastically driven damped harmonic oscillator kernel can model the δ Scuti pulsations. It provides similar results to parametric models but with a simpler framework, using only three hyperparameters. It also enables us to model poorly sampled RV data that were excluded from previous analyses, hence extending the RV baseline by nearly five years. Altogether, the orbit and mass of both planets can be constrained from RV only, which was not possible with the parametric modeling. To characterize the system more accurately, we also perform a joint fit of all available relative astrometry and RV data. Our orbital solutions for β Pic b favor a low eccentricity of 0.029$_{−0.024}^{+0.061}$ and a relatively short period of 21.1$_{−0.8}^{+2.0}$ yr. The orbit of β Pic c is eccentric with 0.206$_{−0.063}^{+0.074}$ with a period of 3.36 ± 0.03 yr. We find model-independent masses of 11.7 ± 1.4 and 8.5 ± 0.5 M Jup for β Pic b and c, respectively, assuming coplanarity. The mass of β Pic b is consistent with the hottest start evolutionary models, at an age of 25 ± 3 Myr. A direct detection of β Pic c would provide a second calibration measurement in a coeval system.

79 ASTRONOMY AND ASTROPHYSICS↗

Polymers for Extreme Conditions Designed Using Syntax-Directed Variational Autoencoders

We report the design/discovery of new materials is highly nontrivial owing to the near-infinite possibilities of material candidates and multiple required property/performance objectives. Thus, machine learning tools are now commonly employed to virtually screen material candidates with desired properties by learning a theoretical mapping from material-to-property space, referred to as the forward problem. However, this approach is inefficient and severely constrained by the candidates that the human imagination can conceive. Thus, in this work on polymers, we tackle the materials discovery challenge by solving the inverse problem: directly generating candidates that satisfy desired property/performance objectives. We utilize syntax-directed variational autoencoders (VAE) in tandem with Gaussian process regression (GPR) models to discover polymers expected to be robust under three extreme conditions: (1) high temperatures, (2) high electric field, and (3) high temperature and high electric field, useful for critical structural, electrical, and energy storage applications. This approach to learn from and augment) human ingenuity is general and can be extended to discover polymers with other targeted properties and performance measures.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

A Novel Approach for Determining Source–Receptor Relationships in Model Simulations: A Case Study of Black Carbon Transport in Northern Hemisphere Winter

A Gaussian process (GP) emulator is applied to quantify the contribution of local and remote emissions of black carbon (BC) on the BC concentrations in different regions using a Latin Hypercube sampling strategy for emission perturbations in the offline version of the Community Atmosphere Model Version 5.1 (CAM5) simulations. The source-receptor relationships are computed based on simulations constrained by a standard free-running CAM5 simulation and the ERA-Interim reanalysis product. The analysis demonstrates that the emulator is capable of retrieving the source-receptor relationships based on a small number of CAM5 simulations. Most regions are found susceptible to their local emissions. The emulator also finds that the source-receptor relationships retrieved from the model-driven and the reanalysis-driven simulations are very similar, suggesting that the simulated circulation in CAM5 resembles the assimilated meteorology in ERA-Interim. The robustness of the results provides confidence for applying the emulator to detect dose-response signals in the climate system.

black↗

Non-Gaussian halo mass function and non-spherical halo collapse: theory vs. simulations

The mass distribution of dark matter halos is a sensitive probe of primordial non-Gaussianity (NG). We derive an analytical formula of the halo mass function by perturbatively computing excursion set path-integrals for a non-Gaussian density field with non-vanishing skewness, f{sub NL}. We assume a stochastic barrier model which captures the main features of the ellipsoidal collapse of halos. Contrary to previous results based on extensions of the Press-Schechter formalism to NG initial conditions, we find that the non-spherical collapse of halos directly alter the signature of primordial NG. This points toward a potential degeneracy between the effect of primordial non-Gaussianity and that of non-linear halo collapse. The inferred mass function is found to be in remarkable agreement with N-body simulations of NG local type. Deviations are well within numerical uncertainties for all values of f{sub NL}{sup loc} in the range of validity of the perturbative calculation (|f{sub nl}{sup loc}|∼<200). Moreover, the comparison with simulation results suggests that for |f{sub NL}|∼>30 the non-linear collapse of halos, as described by our barrier model, strongly deviates from that of Gaussian initial conditions. This is not surprising since the effect of non-linear gravitational processes may be altered by initially large NG. Hence, in the lack of prior theoretical knowledge, halo collapse model parameters should be included in statistical halo mass function data analysis which aim to constrain the signature of primordial NG.

79 ASTRONOMY AND ASTROPHYSICS↗

Radar-Based Bayesian Estimation of Ice Crystal Growth Parameters within a Microphysical Model

The potential for polarimetric Doppler radar measurements to improve predictions of ice microphysical processes within an idealized model–observational framework is examined. In an effort to more rigorously constrain ice growth processes (e.g., vapor deposition) with observations of natural clouds, a novel framework is developed to compare simulated and observed radar measurements, coupling a bulk adaptive-habit model of vapor growth to a polarimetric radar forward model. Bayesian inference on key microphysical model parameters is then used, via a Markov chain Monte Carlo sampler, to estimate the probability distribution of the model parameters. The statistical formalism of this method allows for robust estimates of the optimal parameter values, along with (non-Gaussian) estimates of their uncertainty. To demonstrate this framework, observations from Department of Energy radars in the Arctic during a case of pristine ice precipitation are used to constrain vapor deposition parameters in the adaptive habit model. The resulting parameter probability distributions provide physically plausible changes in ice particle density and aspect ratio during growth. A lack of direct constraint on the number concentration produces a range of possible mean particle sizes, with the mean size inversely correlated to number concentration. Consistency is found between the estimated inherent growth ratio and independent laboratory measurements, increasing confidence in the parameter PDFs and demonstrating the effectiveness of the radar measurements in constraining the parameters. Furthermore, the combined Doppler and polarimetric observations produce the highest-confidence estimates of the parameter PDFs, with the Doppler measurements providing a stronger constraint for this case.

54 ENVIRONMENTAL SCIENCES↗

Radar-Based Bayesian Estimation of Ice Crystal Growth Parameters within a Microphysical Model

The potential for polarimetric Doppler radar measurements to improve predictions of ice microphysical processes within an idealized model–observational framework is examined. In an effort to more rigorously constrain ice growth processes (e.g., vapor deposition) with observations of natural clouds, a novel framework is developed to compare simulated and observed radar measurements, coupling a bulk adaptive-habit model of vapor growth to a polarimetric radar forward model. Bayesian inference on key microphysical model parameters is then used, via a Markov chain Monte Carlo sampler, to estimate the probability distribution of the model parameters. The statistical formalism of this method allows for robust estimates of the optimal parameter values, along with (non-Gaussian) estimates of their uncertainty. To demonstrate this framework, observations from Department of Energy radars in the Arctic during a case of pristine ice precipitation are used to constrain vapor deposition parameters in the adaptive habit model. The resulting parameter probability distributions provide physically plausible changes in ice particle density and aspect ratio during growth. A lack of direct constraint on the number concentration produces a range of possible mean particle sizes, with the mean size inversely correlated to number concentration. Consistency is found between the estimated inherent growth ratio and independent laboratory measurements, increasing confidence in the parameter PDFs and demonstrating the effectiveness of the radar measurements in constraining the parameters. The combined Doppler and polarimetric observations produce the highest-confidence estimates of the parameter PDFs, with the Doppler measurements providing a stronger constraint for this case.

Robert S. Schrom↗

Learned adaptive properties for mitigation of weight perturbations in embedded spiking networks

Recent years have seen an increased importance of neural network inference in edge-based scenarios, which impose size and power constraints requiring novel computing devices. These same edge scenarios may require operating over long periods of time, or exposure to extreme environments, resulting in a drift of neural network weights that cause degraded performance. In searching for ways to develop neural network approaches that perform robustly under these conditions, we propose a biologically-inspired mechanism for the dynamic adaptation of within-neuron parameters that is guided by a global context signal carrying information about perturbations and variability in incoming stimuli. Specifically, we demonstrate that adaptive voltage thresholds or neuronal time constants, when informed by a global context signal, can enable network-level mechanisms to recover from perturbed synaptic weights. Consistent with prior literature, the context-modulated approach is effective for recurrent, but not feedforward networks, by modulating network level dynamics. We demonstrate this approach successfully recovers performance in image classification tasks and spatiotemporal tracking tasks under idealized and Gaussian noise as well as for realistic perturbations from a memristive device when exposed to ionizing radiation. Finally, we discuss how this approach enables the design of robust and energy-efficient neuromorphic systems that perform well, even in resource-constrained scenarios with extreme environments such as edge processing.

context modulation↗

Designing Monte Carlo Simulation and an Optimal Machine Learning to Optimize and Model Space Missions

This paper investigates applying artificial intelligence (AI) algorithms to attitude control system of satellites to optimally tune the controller using high performance computing. This methodology is applied to the Virtual Telescope for X-ray Observation mission, which is a precise formation of two separate spacecraft observing multiple objects in the space in the X-ray domain. The mission is divided into phases based on the instrumentation and the mission goal. To reach an stable precise formation robust to stochastic slew and slew rate (i.e., Euler angles and angular velocities) in a minimal constrained time T , consumed energy of the attitude control system, denoted as E, and root-mean-square state error of attitude control system, denoted as e, are minimized. Monte-Carlo simulation is used for the sensitivity analysis of optimization and designing a controller. Deep neural networks (DNN), Gaussian processes (GP), and support vector regression (SVR) learn this optimization as a surrogate model, while their hyperparameters are optimized in a novel approach. THETA supercomputer at Argonne Leadership Computing Facility (ALCF) is used for optimizing the hyperparameters of DNN. The surrogate model meets the requirements of the mission, and it shows a better performance over the optimization and Monte-Carlo. The optimal DNN can satisfy the mission requirements e and T while reducing E for 90% compared to the other given methods.

42 ENGINEERING↗

Wavelength Dependence of Activity-induced Photometric Variations for Young Cool Stars in Hyades

We investigate photometric variations due to stellar activity that induce systematic radial-velocity errors (so-called “jitter”) for the four targets in the Hyades open cluster observed by the K2 mission (EPIC 210721261, EPIC 210923016, EPIC 247122957, and EPIC 247783757). Applying Gaussian process regressions to the K2 light curves and the near-infrared (NIR) light curves observed with the IRSF 1.4 m telescope, we derive the wavelength dependences of the photometric signals due to stellar activity. To estimate the temporal variations in the photometric variability amplitudes between the two observation periods of K2 and IRSF, separated by more than 2 yr, we analyze a number of K2 targets in Hyades that have also been observed in Campaigns 4 and 13 and find a representative variation rate over 2 yr of 38% ± 71%. Taking this temporal variation into account, we constrain projected sizes and temperature contrast properties of the starspots in the stellar photosphere to be approximately 10% and 0.95%, respectively. These starspot properties can induce relatively large differences in the variability amplitude over different observational passbands, and we find that radial-velocity jitter may be more suppressed in the NIR than previously expected. Our result supports profits of ongoing exoplanet search projects that are attempting to detect or confirm young planets in open clusters via radial-velocity measurements in the NIR.

47 OTHER INSTRUMENTATION↗

CAFE AU LAIT: Compute-Aware Federated Augmented Low-Rank AI Training

Federated finetuning is crucial for unlocking the knowledge embedded in pretrained Large Language Models (LLMs) when data are geographically distributed across clients. Unlike finetuning with data from a single institution, federated finetuning allows collaboration across multiple institutions, enabling the utilization of diverse and decentralized datasets while preserving data privacy. Given the high computing costs of LLM training and the emphasis on energy efficiency in Federated Learning (FL), Low-Rank Adaptation (LoRA) has emerged as a widely adopted algorithm due to its significantly reduced number of trainable parameters. However, this assumes that all data silos have the necessary computing resources to compute local updates of LLMs. Nevertheless, in practice, the computing resources across clients are highly heterogeneous: while some may have access to hundreds of GPUs, others might have limited or no GPU access. Recently, federated finetuning using synthetic data has been proposed, allowing clients to participate in a collaborative training run without training LLMs locally. However, our experimental results reveal a performance gap between models trained using synthetic data and those trained using local updates. Motivated by the observed heterogeneity in computing resources and the performance gap, we propose a novel two-stage algorithm that leverages the storage and computing capabilities of a strong server. In the first stage, under the coordination of the strong server, clients with limited computing resources collaborate to generate synthetic data, which is transferred to and stored on the strong server. In the second stage, the strong server uses this synthetic data on behalf of the resource-constrained clients to perform federated LoRA finetuning alongside clients with sufficient computing resources. This approach ensures that all clients can participate in the finetuning process. Experimental results demonstrate that incorporating local updates from even a small fraction of clients improves performance compared to using synthetic data for all clients. Furthermore, we incorporate the Gaussian mechanism in both stages to guarantee client-level differential privacy.

Wang, Jiayi [ORNL]↗

Parallel derivative-free optimization for simulation-based design of behind-the-meter energy systems

In this work, the integrated design and dispatch of behind-the-meter or distributed resources (e.g. stationary battery storage and solar PV generation) is considered. A simulation-based framework is employed, generating high-fidelity results with closed-loop predictive control at a fine resolution, at the expense of high computational cost (several minutes to a few hours per design point). To address this challenge, parallel derivative-free design methods are considered. Four methods are compared, including state-of-the-art surrogate-based methods (Radial-Basis Functions and Gaussian processes) and sampling strategies, an evolutionary-based method, and a simple sequential grid refinement method. As a case study, two types of design problem with increasing complexity are considered, namely, the design of behind-the-meter resources (three design variables) and the inclusion of grid capacity (four design variables). The second yields a constrained design problem for which violations can only be determined after solving the computationally expensive simulation. For the three-dimensional case, all methods present a good performance, achieving a solution within 1% of the optimum after the first iteration, with the sequential grid refinement exhibiting the fastest convergence and achieving the best final objective value. This indicates that the parallel evaluation of multiple sampling points may be more important than the choice of method for small decision spaces. For the four-dimensional constrained case, the Genetic Algorithm presents the best tradeoff between performance and computational effort, while the rough objective function terrain generated by constraint violation penalties reduces the performance of surrogate-based methods. Contour plots with flat regions indicate flexibility in the optimal design and highlight the importance of characterizing the solution space.

24 POWER TRANSMISSION AND DISTRIBUTION↗

UCAC3 PIXEL PROCESSING

The third US Naval Observatory CCD Astrograph Catalog, UCAC3, was released at the IAU General Assembly on 2009 August 10. It is a highly accurate, all-sky astrometric catalog of about 100 million stars in the R = 8-16 mag range. Recent epoch observations are based on over 270,000 CCD exposures, which have been re-processed for the UCAC3 release applying traditional and new techniques. Challenges in the data have been high dark current and asymmetric image profiles due to the poor charge transfer efficiency of the detector. Non-Gaussian image profile functions were explored and correlations are found for profile fit parameters with properties of the CCD frames. These were utilized to constrain the image profile fit models and adequately describe the observed point-spread function of stellar images with a minimum number of free parameters. Using an appropriate model function, blended images of double stars could be fit successfully. UCAC3 positions are derived from two-dimensional image profile fits with a five-parameter, symmetric Lorentz profile model. Internal precisions of about 5 mas per coordinate and single exposure are found, which are degraded by the atmosphere to about 10 mas. However, systematic errors exceeding 100 mas are present in the x, y data which have been corrected in the astrometric reductions following the x, y data reduction step described here.

79 ASTRONOMY AND ASTROPHYSICS↗

Statistics of the geomagnetic secular variation for the past 5Ma

A new statistical model is proposed for the geomagnetic secular variation over the past 5Ma. Unlike previous models, the model makes use of statistical characteristics of the present day geomagnetic field. The spatial power spectrum of the non-dipole field is consistent with a white source near the core-mantle boundary with Gaussian distribution. After a suitable scaling, the spherical harmonic coefficients may be regarded as statistical samples from a single giant Gaussian process; this is the model of the non-dipole field. The model can be combined with an arbitrary statistical description of the dipole and probability density functions and cumulative distribution functions can be computed for declination and inclination that would be observed at any site on Earth's surface. Global paleomagnetic data spanning the past 5Ma are used to constrain the statistics of the dipole part of the field. A simple model is found to be consistent with the available data. An advantage of specifying the model in terms of the spherical harmonic coefficients is that it is a complete statistical description of the geomagnetic field, enabling us to test specific properties for a general description. Both intensity and directional data distributions may be tested to see if they satisfy the expected model distributions.

Constable, C. G.↗

Statistics of the geomagnetic secular variation for the past 5 m.y

A new statistical model is proposed for the geomagnetic secular variation over the past 5Ma. Unlike previous models, the model makes use of statistical characteristics of the present day geomagnetic field. The spatial power spectrum of the non-dipole field is consistent with a white source near the core-mantle boundary with Gaussian distribution. After a suitable scaling, the spherical harmonic coefficients may be regarded as statistical samples from a single giant Gaussian process; this is the model of the non-dipole field. The model can be combined with an arbitrary statistical description of the dipole and probability density functions and cumulative distribution functions can be computed for declination and inclination that would be observed at any site on Earth's surface. Global paleomagnetic data spanning the past 5Ma are used to constrain the statistics of the dipole part of the field. A simple model is found to be consistent with the available data. An advantage of specifying the model in terms of the spherical harmonic coefficients is that it is a complete statistical description of the geomagnetic field, enabling us to test specific properties for a general description. Both intensity and directional data distributions may be tested to see if they satisfy the expected model distributions.

Constable, C. G.↗

Constraints on variations in inflaton decay rate from modulated preheating

Modulated (p)reheating is thought to be an alternative mechanism for producing super-horizon curvature perturbations in CMB. But large non-gaussianity and iso-curvature perturbations produced by this mechanism rule out its acceptability as the sole process responsible for generating CMB perturbations. We explore the situation where CMB perturbations are mostly generated by usual quantum fluctuations of inflaton during inflation, but a modulated coupling constant between inflaton and a secondary scalar affects the preheating process and produces some extra curvature perturbations. If the modulating scalar field is considered to be a dark matter candidate, coupling constant between the fields has to be unnaturally fine tuned in order to keep the local-form non-gaussianity and the amplitude of iso-curvature perturbations within observational limit; otherwise parameters of the models have to be tightly constrained. Those constraints imply that the curvature perturbations generated by modulated preheating should be less than 15% of the total observed CMB perturbations. On the other hand if the modulating scalar field is not a dark matter candidate, parameters of the models could not be constrained, but the constraints on the maximum amount of the curvature perturbations coming from modulated preheating remain valid.

79 ASTRONOMY AND ASTROPHYSICS↗