Advancements on multi-fidelity random Fourier neural networks: application to hurricane modeling for offshore wind energy
Explore the source record for details and available documents.
Davis, Owen Nicholas [Sandia National Laboratories↗
SEARCH · Search NASA
Results for “Neural Network Models”
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.
Evaluating the Trustworthiness of Explainable Artificial Intelligence (XAI) Methods Applied to Regression Predictions of Arctic Sea Ice Motion
Abstract Recent advances in explainable artificial intelligence (XAI) methods show promise for understanding predictions made by machine learning (ML) models. XAI explains how the input features are relevant or important for the model predictions. We train linear regression (LR) and convolutional neural network (CNN) models to make 1-day predictions of sea ice velocity in the Arctic from inputs of present-day wind velocity and previous-day ice velocity and concentration. We apply XAI methods to the CNN and compare explanations to variance explained by LR. We confirm the feasibility of using a novel XAI method [i.e., global layerwise relevance propagation (LRP)] to understand ML model predictions of sea ice motion by comparing it to established techniques. We investigate a suite of linear, perturbation-based, and propagation-based XAI methods in both local and global forms. Outputs from different explainability methods are generally consistent in showing that wind speed is the input feature with the highest contribution to ML predictions of ice motion, and we discuss inconsistencies in the spatial variability of the explanations. Additionally, we show that the CNN relies on both linear and nonlinear relationships between the inputs and uses nonlocal information to make predictions. LRP shows that wind speed over land is highly relevant for predicting ice motion offshore. This provides a framework to show how knowledge of environmental variables (i.e., wind) on land could be useful for predicting other properties (i.e., sea ice velocity) elsewhere. Significance Statement Explainable artificial intelligence (XAI) is useful for understanding predictions made by machine learning models. Our research establishes trustability in a novel implementation of an explainable AI method known as layerwise relevance propagation for Earth science applications. To do this, we provide a comparative evaluation of a suite of explainable AI methods applied to machine learning models that make 1-day predictions of Arctic sea ice velocity. We use explainable AI outputs to understand how the input features are used by the machine learning to predict ice motion. Additionally, we show that a convolutional neural network uses nonlinear and nonlocal information in making its predictions. We take advantage of the nonlocality to investigate the extent to which knowledge of wind on land is useful for predicting sea ice velocity elsewhere.
Hoffman, Lauren [Scripps Institution of Oceanograp↗
DISTEMA: distance map-based estimation of single protein model accuracy with attentive 2D convolutional neural network
Abstract Background Estimation of the accuracy (quality) of protein structural models is important for both prediction and use of protein structural models. Deep learning methods have been used to integrate protein structure features to predict the quality of protein models. Inter-residue distances are key information for predicting protein’s tertiary structures and therefore have good potentials to predict the quality of protein structural models. However, few methods have been developed to fully take advantage of predicted inter-residue distance maps to estimate the accuracy of a single protein structural model. Result We developed an attentive 2D convolutional neural network (CNN) with channel-wise attention to take only a raw difference map between the inter-residue distance map calculated from a single protein model and the distance map predicted from the protein sequence as input to predict the quality of the model. The network comprises multiple convolutional layers, batch normalization layers, dense layers, and Squeeze-and-Excitation blocks with attention to automatically extract features relevant to protein model quality from the raw input without using any expert-curated features. We evaluated DISTEMA’s capability of selecting the best models for CASP13 targets in terms of ranking loss of GDT-TS score. The ranking loss of DISTEMA is 0.079, lower than several state-of-the-art single-model quality assessment methods. Conclusion This work demonstrates that using raw inter-residue distance information with deep learning can predict the quality of protein structural models reasonably well. DISTEMA is freely at https://github.com/jianlin-cheng/DISTEMA
59 BASIC BIOLOGICAL SCIENCES↗
Learning generative neural networks with physics knowledge
Deep generative neural networks have enabled modeling complex distributions, but incorporating physics knowledge into the neural networks is still challenging and is at the core of current physics-based machine learning research. To this end, we propose a physics generative neural network (PhysGNN), a new class of generative neural networks for learning unknown distributions in a physical system described by partial differential equations (PDE). PhysGNN couples PDE systems with generative neural networks. It is a fully differentiable model that allows back-propagation of gradients through both numerical PDE solvers and generative neural networks, and is trained by minimizing the discrete Wasserstein distance between generated and observed probability distributions of the PDE outputs using the stochastic gradient descent method. Moreover, PhysGNN does not require adversarial training like standard generative neural networks, which offers better stability than adversarial training. We show that PhysGNN can learn complex distributions in stochastic inverse problems, where conventional methods such as maximum likelihood estimation and momentum matching methods may be inapplicable when little knowledge is known about the form of unknown distributions or the physical model is too complex. Furthermore, our method allows physics-based generative neural network training for learning complex distributions in the context of differential equations.
97 MATHEMATICS AND COMPUTING↗
Neural network denoising of HED x-ray images, with an introduction to neural networks [Slides]
We created and trained a new neural network denoiser model using estimates of noise in the data. We showed that our denoiser significantly reduces the combined blur and noise on the training and testing set. We demonstrated the efficacy of neural network denoisers for reducing noise on HED x-ray images. Changes to network architecture and improvements to the noise model in training could improve the model.
47 OTHER INSTRUMENTATION↗
Estimating Watershed Subsurface Permeability From Stream Discharge Data using Deep Neural Networks, Frontiers in Earth Science: Dataset
This data package contains watershed modeling inputs and outputs as well as deep neural networks training and testing results used in "Estimating Watershed Subsurface Permeability From Stream Discharge Data using Deep Neural Networks" (Cromwell et al., 2021). We train various deep neural network models with different architectures to predict subsurface permeability from stream discharge hydrograph at the watershed outlet. The training data are obtained from ensemble simulations of hydrographs corresponding to an permeability ensemble using a fully-distributed, integrated surface-subsurface hydrologic model. The trained model is then applied to estimate the permeability of the real watershed using its observed hydrograph at the outlet. Our study demonstrates that the permeabilities of the soil and geologic facies that make significant contributions to the outlet discharge can be more accurately estimated from the discharge data. Their estimations are also more robust with observation errors. Compared to the traditional ensemble smoother method, DNNs show stronger performance in capturing the nonlinear relationship between permeability and stream hydrograph to accurately estimate permeability. Our study sheds new light on the value of the emerging deep learning methods in assisting integrated watershed modeling by improving parameter estimation, which will eventually reduce the uncertainty in predictive watershed models. For detailed information regarding watershed model description and DNNs setup, please refer to Cromwell et al., 2021. The deep_learning.zip file contains the inputs and outputs for the different deep neural networks. The watershed_model.zip file contains the inputs and outputs for watershed simulation using ATS. The figures.zip file contains raw figures and their corresponding scripts. For a more detailed description, please see the README.md file within each compressed file.
54 ENVIRONMENTAL SCIENCES↗
Machine learning for precise hit position reconstruction in Resistive Silicon Detectors
RSDs are LGAD silicon sensors with 100% fill factor, based on the principle of AC-coupled resistive read-out. Signal sharing and internal charge multiplication are the RSD key features to achieve picosecond-level time resolution and micron-level spatial resolution, thus making these sensors promising candidates as 4D-trackers for future experiments. This paper describes the use of a neural network to reconstruct the hit position of ionizing particles, an approach that can boost the performance of the RSD with respect to analytical models. The neural network has been trained in the laboratory and then validated on test beam data. The device-under-test in this work is a 450 μm-pitch matrix from the FBK RSD2 production, which achieved a resolution of about 65 μm at the DESY Test Beam Facility, a 50% improvement compared to a simple analytical reconstruction method, and a factor two better than the resolution of a standard pixel sensor of equal pitch size with binary read-out. The test beam result is compatible with the laboratory ones obtained during the neural network training, confirming the ability of the machine learning model to provide accurate predictions even in environments very different from the training one. Prospects for future improvements are also discussed.
46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
Simultaneous control of the electron temperature and safety factor profiles in DIII-D using model-based optimal control techniques
Future tokamak power plants will likely operate using a single, well-defined plasma scenario, either in steady state or for very long pulse lengths. In order to enhance the robustness of the scenario, feedback controllers for a variety of plasma properties will be necessary to counteract any disturbances and ensure safe operation. However, only a limited set of actuators will be available to control many different quantities. Because of this, it is necessary to develop controllers that are able to regulate multiple plasma properties using a limited set of actuators. To this end, a controller has been developed for the simultaneous regulation of both the electron temperature and safety factor profiles in DIII-D. This algorithm uses a linear quadratic integral control synthesis approach based on a linearized model of the dynamics of the two profiles. Two neural network surrogate models, NubeamNet and MMMnet, are included to improve the fidelity of the model. Furthermore, the controller has been tested in simulation using COTSIM, and has demonstrated the ability to simultaneously track changes in both the electron temperature and safety factor targets, including changes in both the magnitude and the shape of the profiles.
70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
Tencoder: tensor-product encoder-decoder architecture for predicting solutions of PDEs with variable boundary data
It is widely hoped that artificial intelligence will boost data-driven surrogate models in science and engineering. However, fundamental spatial aspects of AI surrogate models remain under-studied. We investigate the ability of neural-network surrogate models to predict solutions to PDEs under variable boundary values. We do not wish to retrain the model when the boundary values change but to make them inputs to the model and infer the solution of the PDE under those boundary conditions. Such a capability is essential to making AI-based surrogate models practically useful. While simple feedforward networks are used for one-dimensional (1D) Poisson equation, an encoder-decoder architecture with a tensor-product layer is developed for the two-dimensional Poisson equation posed on a rectangular domain. We show that it is indeed possible to infer solutions to PDEs from variable boundary data using neural networks in this relatively simple setting, and point to future directions.
Kashi, Aditya↗
Uncertainty quantification of a deep learning fuel property prediction model
Deep learning models are being widely used in the field of combustion. Given the black-box nature of typical neural network based models, uncertainty quantification (UQ) is critical to ensure the reliability of predictions as well as the training datasets, and for a principled quantification of noise and its various sources. Deep learning surrogate models for predicting properties of chemical compounds and mixtures have been recently shown to be promising for enabling data-driven fuel design and optimization, with the ultimate goal of improving efficiency and lowering emissions from combustion engines. In this study, UQ is performed for a multi-task deep learning model that simultaneously predicts the research octane number (RON), Motor Octane Number (MON), and Yield Sooting Index (YSI) of pure components and multicomponent blends. The deep learning model is comprised of three smaller networks: Extractor 1, Extractor 2, and Predictor, and a mixing operator. The molecular fingerprints of individual components are encoded via Extractor 1 and Extractor 2, the mixing operator generates fingerprints for mixtures/blends based on linear mixing operation, and the predictor maps the fingerprint to the target properties. Two different classes of UQ methods, Monte Carlo ensemble methods and Bayesian neural networks (BNNs), are employed for quantifying the epistemic uncertainty. Combinations of Bernoulli and Gaussian distributions with DropConnect and DropOut techniques are explored as ensemble methods. All the DropConnect, DropOut and Bayesian layers are applied to the predictor network. Aleatoric uncertainty is modeled by assuming that each data point has an independent uncertainty associated with it. The results of the UQ study are further analyzed to compare the performance of BNN and ensemble methods. Although this study is confined to UQ of fuel property prediction, the methodologies are applicable to other deep learning frameworks that are being widely used in the combustion community.
33 ADVANCED PROPULSION SYSTEMS↗
PyTorch Implementation of Log-Additive Convolutional Neural Networks
This code is a collection of python code that defines, trains, and tests Log-Additive Convolutional Neural Networks. The model components and training routine are based on the PyTorch python library. The code implements the Log-Additive Convolutional Neural Networks as described in Pagendam et al. 2023. In addition to the Log-Additive Convolutional Neural Networks, this library also defines the Log-Normal Density loss function as described in Pagendam et al. 2023. Code from this paper is not publicly available, so the Pytorch implementation of this type of model is unique to this library.
Callis, Skylar↗
A data-driven method for modelling dissipation rates in stratified turbulence
We present a deep probabilistic convolutional neural network (PCNN) model for predicting local values of small-scale mixing properties in stratified turbulent flows, namely the dissipation rates of turbulent kinetic energy and density variance, $\varepsilon$ and $\chi$ . Inputs to the PCNN are vertical columns of velocity and density gradients, motivated by data typically available from microstructure profilers in the ocean. The architecture is designed to enable the model to capture several characteristic features of stratified turbulence, in particular the dependence of small-scale isotropy on the buoyancy Reynolds number $Re_b:=\varepsilon /(\nu N^2)$ , where $\nu$ is the kinematic viscosity and $N$ is the background buoyancy frequency, the correlation between suitably locally averaged density gradients and turbulence intensity and the importance of capturing the tails of the probability distribution functions of values of dissipation. Empirically modified versions of commonly used isotropic models for $\varepsilon$ and $\chi$ that depend only on vertical derivatives of density and velocity are proposed based on the asymptotic regimes $Re_b\ll 1$ and $Re_b\gg 1$ , and serve as an instructive benchmark for comparison with the data-driven approach. When trained and tested on a simulation of stratified decaying turbulence which accesses a range of turbulent regimes (associated with differing values of $Re_b$ ), the PCNN outperforms assumptions of isotropy significantly as $Re_b$ decreases, and additionally demonstrates improvements over the fitted empirical models. A differential sensitivity analysis of the PCNN facilitates a comparison with the theoretical models and provides a physical interpretation of the features enabling it to make improved predictions.
42 ENGINEERING↗
Coupled Lake‐Atmosphere‐Land Physics Uncertainties in a Great Lakes Regional Climate Model
Abstract This study develops a surrogate‐based method to assess the uncertainty within a convective permitting integrated modeling system of the Great Lakes region, arising from interacting physics parameterizations across the lake, atmosphere, and land surface. Perturbed physics ensembles of the model during the 2018 summer are used to train a neural network surrogate model to predict lake surface temperature (LST) and near‐surface air temperature (T2m). Average physics uncertainties are determined to be 1.5C for LST and T2m over land, and 1.9C for T2m over lake, but these have significant spatiotemporal variations. We find that atmospheric physics parameterizations alone are the dominant sources of uncertainty (45%–53%), while lake and land parameterizations account for 33% and 38% of the uncertainty of LST and T2m over land respectively. Interactions of atmosphere physics parameterizations with those of the land and lake contribute to an additional 13%–17% of the total variance. LST and T2m over the lake are more uncertain in the deeper northern lakes, particularly during the rapid warming phase that occurs in late spring/early summer. The LST uncertainty increases with sensitivity to the lake model's surface wind stress scheme. T2m over land is more uncertain over forested areas in the north, where it is most sensitive to the land surface model, than the more agricultural land in the south, where it is most sensitive to the atmospheric planetary boundary and surface layer scheme. Uncertainty also increases in the southwest during multiday temperature declines with higher sensitivity to the land surface model.
54 ENVIRONMENTAL SCIENCES↗
A Neural Network-Based Power Mismatch Elimination Strategy for Integrated Solar and ESS AC/DC Systems (MARS)
The multiport autonomous reconfigurable solar power plant (MARS) is a promising concept for the integration of photovoltaic (PV) and energy storage system (ESS) to the transmission ac grid and a high-voltage direct current (HVdc) link. The presence of PV and ESS in each arm of the MARS results in uneven distribution of active power among different submodules (SMs), thereby leading to unbalanced SM capacitor voltages and potentially compromising the system stability. Moreover, in the case of partial shadings, shaded PV SMs will suffer from decreased power injections causing power mismatch in the MARS system. To address this issue, a neural-network-based power mismatch elimination (NNPME) strategy is proposed in this article. The proposed NNPME strategy optimizes ESS usage and leverages both dc and ac circulating currents to facilitate power transfer among the SMs, arms, and phases of the MARS system. Simulation and control hardware-in-the-loop (cHIL) experiments demonstrate the effectiveness of the proposed NNPME strategy. Compared with the traditional approaches, the proposed NNPME strategy can significantly enhance system efficiency and ensure stable and continuous operation, even in the presence of uneven power distribution within the MARS system.
14 SOLAR ENERGY↗
Solving key challenges in collider physics with foundation models
Foundation models are neural networks that are capable of simultaneously solving many problems. Large language foundation models like ChatGPT have revolutionized many aspects of daily life, but their impact for science is not yet clear. In this paper, we use a new foundation model for hadronic jets to solve three key challenges in collider physics. In particular, we show how experiments can (1) save significant computing power when developing reconstruction algorithms, (2) perform a complete uncertainty quantification for high-dimensional measurements, and (3) search for new physics with model agnostic methods using low-level inputs. In each case, there are significant computational or methodological challenges with current methods that limit the science potential of deep learning algorithms. By solving each problem, we take jet foundation models beyond proof-of-principle studies and into the toolkit of practitioners.
72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
Enhancing the Harris and Crighton Particle-in-Cell (PIC) Stress Model with Bayesian Learning
This is ongoing work towards developing and integrating a Bayesian Neural Network (BNN) model for Particle-in-Cell (PIC) stress in MFiX legacy code.
Marayikkottu, Akhil V.↗
Predicting impurity spectral functions using machine learning
The Anderson Impurity Model (AIM) is a canonical model of quantum many-body physics. Here we investigate whether machine learning models, both neural networks (NN) and kernel ridge regression (KRR), can accurately predict the AIM spectral function in all of its regimes, from empty orbital, to mixed valence, to Kondo. To tackle this question, we construct two large spectral databases containing approximately 410 000 and 600 000 spectral functions of the single-channel impurity problem. We show that the NN models can accurately predict the AIM spectral function in all of its regimes, with pointwise mean absolute errors down to 0.003 in normalized units. We find that the trained NN models outperform models based on KRR and enjoy a speedup on the order of 10 5 over traditional AIM solvers. Finally, the required size of the training set of our model can be significantly reduced using farthest point sampling in the AIM parameter space, which is important for generalizing our method to more complicated multichannel impurity problems of relevance to predicting the properties of real materials.
75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
Landsat 8 monitoring of multi-depth suspended sediment concentrations in Lake Erie’s Maumee River using machine learning
Satellite remote sensing has been widely used to map suspended sediment concentration (SSC) in waterbodies. However, due to the complexity of sediment-water interactions, it has been difficult to derive linear and non-linear regression equations to reliably predict SSC, especially when trying to estimate depth of integrated sediment. Herein, this study uses Landsat 8 OLI (Operational Land Imager) sensor to map SSC within the Maumee River in Ohio, USA, at multiple depth intervals (15, 61, 91, and 182 cm). Simple linear least squares regression (LLSR), and three common machine learning models: random forest (RF), support vector regression (SVR), and model averaged neural network (MANN) were used to estimate SSC at the depth intervals. All machine learning models significantly outperformed LLSR while RF performed the best. In both RF and MANN, R2 (coefficient of determination) increases with depth with a maximum R2 of 0.89 and 0.83, respectively, at a depth of 0–182 cm. The results show that machine learning models can implement nonlinear relationships that produce better predictions than traditional linear regression methods in estimating depth integrated SSC, especially when samples are limited.
47 OTHER INSTRUMENTATION↗