SEARCH · Engineering Papers

Results for “Kernel 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.

SourceResearch connectionFrom yearThrough year

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

At least 271 records · Page 15

All sources NASA NTRS DOE OSTI

DOE OSTI2020

Multi-objective optimization of peel and shear strengths in ultrasonic metal welding using machine learning-based response surface methodology

Ultrasonic metal welding (UMW) is a solid-state joining technique with varied industrial applications. Despite of its numerous advantages, UMW has a relative narrow operating window and is sensitive to variations in process conditions. As such, it is imperative to quantitatively characterize the influence of welding parameters on the resulting joint quality. The quantification model can be subsequently used to optimize the parameters. Conventional response surface methodology (RSM) usually employs linear or polynomial models, which may not be able to capture the intricate, nonlinear input-output relationships in UMW. Furthermore, some UMW applications call for simultaneous optimization of multiple quality indices such as peel strength, shear strength, electrical conductivity, and thermal conductivity. To address these challenges, this paper develops a machine learning (ML)-based RSM to model the input-output relationships in UMW and jointly optimize two quality indices, namely, peel and shear strengths. The performance of various ML methods including spline regression, Gaussian process regression (GPR), support vector regression (SVR), and conventional polynomial regression models with different orders is compared. A case study using experimental data shows that GPR with radial basis function (RBF) kernel and SVR with RBF kernel achieve the best prediction accuracy. The obtained response surface models are then used to optimize a compound joint strength indicator that is defined as the average of normalized shear and peel strengths. In addition, the case study reveals different patterns in the response surfaces of shear and peel strengths, which has not been systematically studied in the literature. While developed for the UMW application, the method can be extended to other manufacturing processes.

42 ENGINEERING↗

DOE OSTI2022

Flow and transport in three-dimensional discrete fracture matrix models using mimetic finite difference on a conforming multi-dimensional mesh

Here, we present a comprehensive workflow to simulate single-phase flow and transport in fractured porous media using the discrete fracture matrix approach. The workflow has three primary parts: (1) a method for conforming mesh generation of and around a three-dimensional fracture network, (2) the discretization of the governing equations using a second-order mimetic finite difference method, and (3) implementation of numerical methods for high-performance computing environments. A method to create a conforming Delaunay tetrahedralization of the volume surrounding the fracture network, where the triangular cells of the fracture mesh are faces in the volume mesh, that addresses pathological cases which commonly arise and degrade mesh quality is also provided. Our open-source subsurface simulator uses a hierarchy of process kernels (one kernel per physical process) that allows for both strong and weak coupling of the fracture and matrix domains. We provide verification tests based on analytic solutions for flow and transport, as well as numerical convergence. We also provide multiple expositions of the method in complex fracture networks. In the first example, we demonstrate that the method is robust by considering two scenarios where the fracture network acts as a barrier to flow, as the primary pathway, or offers the same resistance as the surrounding matrix. In the second test, flow and transport through a three-dimensional stochastically generated network containing 257 fractures is presented.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2021

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

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

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2022

Long-time memory effects in a localizable central spin problem

We study the properties of the Nakajima–Zwanzig memory kernel for a qubit immersed in a many-body localized (i.e. disordered and interacting) bath. We argue that the memory kernel decays as a power law in both the localized and ergodic regimes, and show how this can be leveraged to extract t → ∞ populations for the qubit from finite time (Jt ≤ 10 2 ) data in the thermalizing phase. This allows us to quantify how the long-time values of the populations approach the expected thermalized state as the bath approaches the thermodynamic limit. This approach should provide a good complement to state-of-the-art numerical methods, for which the long-time dynamics with large baths are impossible to simulate in this phase. Additionally, our numerics on finite baths reveal the possibility for unbounded exponential growth in the memory kernel, a phenomenon rooted in the appearance of exceptional points in the projected Liouvillian governing the reduced dynamics. In small systems amenable to exact numerics, we find that these pathologies may have some correlation with delocalization.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2021

Seeing through nuclear fuel: Three-dimensional, nondestructive X-ray microscopy and volumetric analyses of neutron-irradiated TRISO-coated fuel kernels

The three-dimensional (3D) characterization of nuclear fuel with X-ray microscopy has historically proven difficult, due to uranium’s high attenuation of easily accessible X-rays, both in a laboratory setting and at a synchrotron user facility. However, this imaging modality provides nondestructive information that can be used to investigate morphological changes arising from external stimuli (e.g., neutron irradiation, high-temperature testing). Using an appropriate X-ray energy spectrum and an adequate X-ray filter, suitable transmissions through properly sized nuclear fuel specimens can be achieved. Here, in this study, we present the methods and results of using a commercially available, laboratory-based X-ray microscope (XRM) to examine the extent of 3D morphological changes of tristructural isotropic (TRISO)-coated fuel particles, specifically uranium oxide/uranium carbide fuel kernels, after high-temperature neutron irradiation.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

DOE OSTI2022

Early arrival waveform inversion using data uncertainties and matching filters with application to near-surface seismic refraction data

We develop an early arrival waveform inversion (EAWI) technique for high-resolution near-surface velocity estimation by iteratively updating the P-wave velocity model to minimize the difference between the observed and calculated seismic refraction data. Traditional EAWI uses a least-squares penalty function and an acoustic forward-modeling engine. Conventional least-squares error is sensitive to data with low signal-to-noise ratio (S/N) and iterations of EAWI stop at a local-minimum data misfit or at the preassigned maximum number of iterations. These stopping criteria can result in overfitting the data. In addition, fitting the elastic field data with an acoustic modeling engine can introduce artifacts in velocity estimation, especially in land data with significant elastic effects. To overcome these challenges, we develop a robust EAWI (REAWI) method by (1) incorporating the data uncertainties into the penalty function and (2) mitigating the elastic effects using a matching filter workflow. The data uncertainties are estimated from waveform reciprocal errors. When full-waveform reciprocity is not available, trace interpolation is applied. The proposed method prevents closely fitting data with low S/N, avoids overall overfitting by stopping the iterations when a normalized chi-square ([Formula: see text]) waveform misfit of one is achieved, and is less affected by elastic effects. Numerical examples and application to near-surface refraction data at a groundwater contamination site suggest that the final REAWI models are more accurate than the corresponding EAWI models, at the same level of misfit. This is the first known application of a matching filter workflow to real land data. The final REAWI models satisfy an appropriate misfit between the real data and predicted elastic P-wave data, making this approach in this respect equivalent to elastic waveform inversion. We also develop a method to analyze model constraint by examining the energy of the wavefield Fréchet derivative thereby avoiding the influence of the data residuals in traditional Fréchet kernels.

Geochemistry & Geophysics↗

DOE OSTI2021

AGR-5/6/7 Fuel Fabrication Report

The U.S. Department of Energy Office of Nuclear Energy (DOE NE) and the Idaho National Laboratory (INL) Advanced Reactor Technologies (ART) Advanced Gas Reactor (AGR) Fuel Development and Qualification program (referred to as AGR Fuel program hereafter) are pursuing qualification of tristructural isotropic (TRISO) coated particle fuel for use in high temperature gas cooled reactors (HTGRs). The AGR Fuel program was established to provide a fuel qualification data set in support of the licensing and operation of an HTGR. BWX Technologies Nuclear Operations Group (BWXT-NOG) was subcontracted to fabricate the fuel for the AGR program. Several investments and innovations were realized in preparation to fabricate fuel for the AGR-5/6/7 irradiation experiments that brought fuel fabrication fully out of the laboratory and into engineering-scale operations. These included: • Increased the kernel fabrication line capacity and uniformity • Upgraded ancillary support equipment and processes for the tristructural isotropic (TRISO) coating furnace • Demonstrated efficient production of the matrix precursor powder by dry jet milling of co mingled components • Demonstrated an engineering-scale method for quick and efficient overcoating TRISO particles with the matrix precursor • Demonstrated an automated, multi cavity compacting system with a volumetric feed system • Demonstrated a combined-cycle thermal treatment furnace These changes increased production rates of some of these processes by an order of magnitude or more while eliminating the use of flammable solvents, multiple grinding and sorting operations, and the weighing out of individual die charges. Three fuel kernel lots were fabricated for production of the fuel for AGR-5/6/7. The initial lot (J52R-16-39316) was certified to fuel specifications but was not used because the kernels had a high fraction of internal fissures that caused an unacceptable fraction of the kernels to fragment when charged to the coating furnace where the TRISO coating would be deposited. Fragmented kernels increased the dispersed uranium in the particles and produced a worrisome fraction of dimpled particles with an elevated probability of in-pile failure. After some efforts to identify the cause of the fissure formation, two additional lots were produced with much lower fissure fractions, J52R-16-69317 and 69318. The latter kernel lot was a backup to the first and was not needed. Multiple kernel batches were composited to form each of the lots so as to simulate a commercial-scale operation where kernel batches would also be composited. Multiple TRISO coating runs were performed and the product characterized so that several could be composited into a TRISO lot. TRISO lot J52R-16-98005 conformed to all fuel specifications except the mean outer pyrocarbon (OPyC) layer thickness was thinner than specified. Furthermore, it was determined that the TRISO lot had a dispersed uranium fraction (DUF) that might result in the compacts not meeting the DUF specification. A review of the role of the OPyC layer and consequences of the DUF by the Technical Coordination Team and INL concluded that the fuel was acceptable for use in the AGR-5/6/7 irradiation experiment. The TRISO particles were overcoated with the matrix precursor that had been produced in a jet-milling operation. The overcoating was performed in equipment originally designed to coat pharmaceuticals. The overcoating process performed well; producing highly spherical and uniform overcoats requiring little upgrading and no recycle or rework. TRISO particles were overcoated with the matrix precursor to achieve nominal volumetric packing fractions (PFs) of TRISO particles of 25% and 40% for the irradiation experiments. The 40% PF compacts occupy the first and fifth test capsule in the test train while the inner three capsules are loaded with 25% PF compacts. The resinated graphite matrix precursor powder was a derivative of the German A3-27 matrix formulation, which differs from previous AGR irradiation campaigns that used an A3-3 formulation. Jet milling of the matrix powder precursor produced a finer mean graphite particle size than the milling operations used for the A3-3 matrix powder precursor. Changes made in the matrix formula and equipment yielded compacts with significantly higher matrix density than was attained in previous AGR irradiation campaigns. The changes in the matrix formulation and the means of milling the powders also complicated resolution of the three fuel compact defect fractions, DUF, exposed kernel fraction (EKF), and the silicon carbide defect fraction (SDF). Characterization data from BWXT-NOG had some anomalous results, so samples of the fuel compacts and overcoated TRISO particles were also analyzed by Oak Ridge National Laboratory (ORNL) to ensure that the defect fractions were accurately characterized.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

DOE OSTI2020

TuckerMPI: A Parallel C++/MPI Software Package for Large-scale Data Compression via the Tucker Tensor Decomposition

With this study, our goal is compression of massive-scale grid-structured data, such as the multi-terabyte output of a high-fidelity computational simulation. For such data sets, we have developed a new software package called TuckerMPI, a parallel C++/MPI software package for compressing distributed data. The approach is based on treating the data as a tensor, i.e., a multidimensional array, and computing its truncated Tucker decomposition, a higher-order analogue to the truncated singular value decomposition of a matrix. The result is a low-rank approximation of the original tensor-structured data. Compression efficiency is achieved by detecting latent global structure within the data, which we contrast to most compression methods that are focused on local structure. In this work, we describe TuckerMPI, our implementation of the truncated Tucker decomposition, including details of the data distribution and in-memory layouts, the parallel and serial implementations of the key kernels, and analysis of the storage, communication, and computational costs. We test the software on 4.5 and 6.7 terabyte data sets distributed across 100 s of nodes (1,000 s of MPI processes), achieving compression ratios between 100 and 200,000×, which equates to 99--99.999% compression (depending on the desired accuracy) in substantially less time than it would take to even read the same dataset from a parallel file system. Moreover, we show that our method also allows for reconstruction of partial or down-sampled data on a single node, without a parallel computer so long as the reconstructed portion is small enough to fit on a single machine, e.g., in the instance of reconstructing/visualizing a single down-sampled time step or computing summary statistics. The code is available at https://gitlab.com/tensors/TuckerMPI.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2024

Energy loss of a heavy fermion in a collisional QED plasma

We compute the energy loss of heavy fermions moving in a plasma, taking into account the modification of the photon collective modes induced by collisions using a Bhatnagar-Gross–Krook collisional kernel. We include contributions from both hard and soft scatterings of the heavy fermion using a collisionally modified hard-thermal-loop resummed propagator. Using this method, one does not need to introduce a separation scale between hard- and soft-momentum exchanges. To place our calculation in context, we review other theoretical approaches to computing the collisional energy loss of fermions and discuss the systematics and results obtained in each approach compared to using a resummed propagator for both hard and soft momentum exchanges. Our final results indicate that self-consistently including the effect of collisions in the self-energies of the resummed propagator results in an increased energy loss compared to using collisionless hard-thermal-loop propagators. The effect becomes larger as the magnitude of the coupling constant and the velocity of the fermion increase. Published by the American Physical Society 2024

Astronomy & Astrophysics↗

DOE OSTI2022

Improved Data Interpretation through Identification of Time Series Periodicity Changes

Analysis and interpretation of time series data is easiest when the data values occur at uniform intervals in time, but actual data may have differing data sampling frequencies, such as monthly and daily readings. Applying data analysis techniques, such as smoothing, to such a data set may not give a representative result between time segments. The ability to automatically distinguish time segments of differing data frequency would provide a means for applying data analysis independently to each segment, though a suitable blending at segment boundaries would be required. A method for detecting frequency changes was developed and applied to Gaussian and median smoothing of hydraulic head data from groundwater wells at the U.S. Department of Energy Hanford Site in southeastern Washington state. The process identifies time segments of high-frequency (daily) or low-frequency (greater than daily) data using adjusted-bandwidth Gaussian kernel density estimation and a threshold value, which are further refined to address small blocks of low-frequency data within larger blocks of high-frequency data. User-selectable levels of smoothing are then applied independently to the time segments prior to combining the segment results for a single smoothed data set. This time segment identification approach provides effective low- and high-frequency data separation, which provides a method to apply data analysis independently to each time segment.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2021

Noise and error analysis and optimization in particle-based kinetic plasma simulations

In this paper we analyze the noise in macro-particle methods used in plasma physics and fluid dynamics, leading to approaches for minimizing the total error, focusing on electrostatic models in one dimension. We begin by describing kernel density estimation for continuous values of the spatial variable x, expressing the kernel in a form in which its shape and width are represented separately. The covariance matrix of the noise in the density is computed, first for uniform true density. The bandwidth of the covariance matrix C(x,y) is related to the width of the kernel. A feature that stands out is the presence of constant negative terms in the elements of the covariance matrix both on and off-diagonal. These negative correlations are related to the fact that the total number of particles is fixed at each time step; they also lead to the property ∫C(x,y)dy = 0. We investigate the effect of these negative correlations on the electric field computed by Gauss's law, finding that the noise in the electric field is related to a process called the Ornstein-Uhlenbeck bridge, leading to a covariance matrix of the electric field with variance significantly reduced relative to that of a Brownian process. For non-constant density, p(x), still with continuous x, we analyze the total error in the density estimation and discuss it in terms of bias-variance optimization (BVO). For some characteristic length l, determined by the density and its second derivative, and kernel width h, having too few particles within h leads to too much variance; for h that is large relative to l, there is too much smoothing of the density. The optimum between these two limits is found by BVO. For kernels of the same width, it is shown that this optimum (minimum) is weakly sensitive to the kernel shape. Next, we repeat the analysis for x discretized on a grid. In this case the charge deposition rule is determined by a particle shape. An important property to be respected in the discrete system is the exact preservation of total charge on the grid; this property is necessary to ensure that the electric field is equal at both ends, consistent with periodic boundary conditions. We find that if the particle shapes satisfy a partition of unity property, the particle charge deposited on the grid is conserved exactly. Further, if the particle shape is expressed as the convolution of a kernel with another kernel that satisfies the partition of unity, then the particle shape obeys the partition of unity. This property holds for kernels of arbitrary width, including widths that are not integer multiples of the grid spacing. Furthermore, we show results relaxing the approximations used to do BVO optimization analytically, by doing numerical computations of the total error as a function of the kernel width, on a grid in x. The comparison between numerical and analytical results shows good agreement over a range of particle shapes. We discuss the practical implications of our results, including the criteria for design and implementation of computationally efficient particle shapes that take advantage of the developed theory.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2021

An algorithm to locate the centers of baryon acoustic oscillations

The cosmic structure formed from baryon acoustic oscillations (BAO) in the early universe is imprinted in the galaxy distribution observable in large-scale surveys and is used as a standard ruler in contemporary cosmology. Typically, BAOs are detected as a preferential length scale in two-point statistics, which gives little information about the location of the BAO structures in real space. The aim of the algorithm described in this paper is to find probable centers of BAOs in the cosmic matter distribution. The algorithm convolves the three-dimensional distribution of matter density with a spherical shell kernel of variable radius placed at different locations. The locations that correspond to the highest values of the convolution correspond to the probable centers of BAOs. This method is realized in an open-source, computationally efficient algorithm. Here, we describe the algorithm and present the results of applying it to the SDSS DR9 CMASS survey and associated mock catalogs. A detailed performance study demonstrates the ability of the algorithm to locate BAO centers and in doing so presents a novel detection of the BAO scale in galaxy surveys.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2022

Inferring safety critical events from vehicle kinematics in naturalistic driving environment: Application of deep learning Algorithms

Advances in sensing technology has enabled the collection of countless terabytes of second-by-second kinematics data. Such data provides opportunities for real-time monitoring of driving behavior and identification of safety critical events (SCEs) including crashes and near crashes. The concept of volatility is relevant in this context, which identifies instability and erratic variations in driving behavior prior to involvement in SCEs. This study utilized vehicle kinematics from a large-scale naturalistic driving data to develop a deep learning approach based on 1D convolutional neural networks (CNN) for inferring SCEs. The data are unique in the sense that such accurate pre-crash data at high fidelity are not available in traditional crash repositories. This study contributes to the literature by providing a first attempt at predicting responses to SCEs by developing deep learning-based CNN architectures using novel driving volatility based kinematic thresholds for a sample of 9553 events. The key contribution lies in developing a volatility-based CNN input layout that is acceptable to CNN schemes and represents the motion kinematics such as speed, acceleration and volatility measures. Several 1D-CNN architectures were developed using layers, numbers of convolutions, layer patterns, and kernels. Shallow and deep architectures were tested, revealing higher accuracy of shallow architectures in detecting SCEs. The optimal number of epochs were identified using an early stopping method while the CNN performance was improved by increasing the number of epochs. The ensemble CNN had the highest predictive accuracy of 95.6% for detection of crashes and near crashes, which was 2.5% higher than the optimal CNN using 20% hold out test data. The ensemble CNN also outperformed classical machine learning models and model performance reported in past studies on detection of SCEs. Finally, these results have implications for identification of safety hotspots and providing real-time alerts and warnings in connected and highly automated vehicle environment including society of automotive engineers levels 3–5.

42 ENGINEERING↗

DOE OSTI2025

Radiation Effects on Network on Chips (NoC) Laboratory Directed Research and Development (LDRD) project

This project was motivated by State-of-the-Art (SOTA) technology that incorporates Network on Chips (NOC) for efficient data communication across the various computer kernels. For example, on the AMD Versal Field Programmable Gate Arrays (FPGA), an NoC has been incorporated for fast data communication from the programmable logic and other computer kernels (processing system, adaptable intelligence engines, etc.). The radiation effects on the legacy technology of this FPGA, such as the programmable logic, are well understood, and established methods exist to measure cross-sections when new families/generations are released; however, newly incorporated technologies, such as the NoC, are not fully understood and could introduce new failure points into the mission space.

36 MATERIALS SCIENCE↗

DOE OSTI2025

The Transition Density Formalism in the First Compton Computation on $^4$He

The method and results of the first theory description of 4He Compton scattering at nuclear energies is presented, with a focus on figures. It uses the same Compton kernels familiar from proton, deuteron and 3He Compton scattering in Chiral Effective Field Theory with explicit Delta degrees of freedom, applicable between about 50 and 130MeV. The result compares well to data from HIγS, MAXlab and Illinois. The sensitivity of the cross section on the (static) scalar-isoscalar polarisabilities of the nucleon is explored. The project is part of the synergetic international effort of experimentalists and theorists in Compton scattering on one- and few-nucleon systems.

Griesshammer, Harald [The George Washington Univer↗

DOE OSTI2023

Systematics in asteroseismic modelling: application of a correlated noise model for oscillation frequencies

ABSTRACT The detailed modelling of stellar oscillations is a powerful approach to characterizing stars. However, poor treatment of systematics in theoretical models leads to misinterpretations of stars. Here, we propose a more principled statistical treatment for the systematics to be applied to fitting individual mode frequencies with a typical stellar model grid. We introduce a correlated noise model based on a Gaussian process (GP) kernel to describe the systematics given that mode frequency systematics are expected to be highly correlated. We show that tuning the GP kernel can reproduce general features of frequency variations for changing model input physics and fundamental parameters. Fits with the correlated noise model better recover stellar parameters than traditional methods that either ignore the systematics or treat them as uncorrelated noise.

Li, Tanda (ORCID:0000000163962563)↗

DOE OSTI2020

Probabilistic Solar Power Forecasting Using Bayesian Model Averaging

There is rising interest in probabilistic forecasting to mitigate risks from solar power uncertainty, but the numerical weather prediction (NWP) ensembles readily available to system operators are often biased and underdispersed. We propose a Bayesian model averaging (BMA) post-processing method suitable for forecasting power from utility-scale photovoltaic (PV) plants at multiple time horizons up to at least the day-ahead timescale. BMA is a kernel dressing technique for NWP ensembles in which the forecast is a weighted sum of member-specific probability density functions. We tailor BMA for utility-scale PV forecasting by modeling power clipping at the AC inverter rating and advance the theory of BMA with a new beta kernel parameterization that accommodates theoretical constraints not previously addressed. BMA is demonstrated for a case study of 11 utility-scale PV plants in Texas, forecasting at hourly resolution for the complete year 2018. BMA's mixture-model approach mitigates underdispersion of the raw ensemble to significantly improve forecast calibration, while consistently outperforming an ensemble model output statistics (EMOS) parametric approach from the literature. At 4-hour lead time, the BMA post-processing achieves continuous ranked probability skill scores of 2--36% over the raw ensemble, with consistent performance at multiple lead times suitable for power system operations.

14 SOLAR ENERGY↗

DOE OSTI2021

Towards Efficient Uncertainty estimation in deep learning for robust energy prediction in crystal materials

In material science, recent studies have started to explore the potential of using deep learning to improve property prediction from high-fidelity simulations, e.g, density functional theory (DFT). However, the design spaces are sometimes too large and intractable to sample completely. This results in a critical question that is how to evaluate the confidence and robustness of the prediction. In this paper, we propose an efficient approach to estimate uncertainty in deep learning using a single forward pass and then apply it for robust prediction of the total energy in crystal lattice structures. Our approach is built upon the deep kernel learning (DKL) that originally introduces to leverage the expressiveness of deep neural networks as input with a probabilistic prediction of Gaussian processes (GPs) as output. Existing DKL methods have difficulties in the accuracy of predictive uncertainty, training stability, and scaling to large datasets, which lead to significant barriers in real-world applications. We propose to address these challenges by using an inducing point approximate GP in feature space combined with spectral normalization as a regularization. We finally demonstrate our robust performance on an artificial example and a real-world application from materials chemistry.

Bi, Sirui↗