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 · Engineering Papers
Results for “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.
The suitability of differentiable, physics-informed machine learning hydrologic models for ungauged regions and climate change impact assessment
As a genre of physics-informed machine learning, differentiable process-based hydrologic models (abbreviated as δ or delta models) with regionalized deep-network-based parameterization pipelines were recently shown to provide daily streamflow prediction performance closely approaching that of state-of-the-art long short-term memory (LSTM) deep networks. Meanwhile, δ models provide a full suite of diagnostic physical variables and guaranteed mass conservation. Here, we ran experiments to test (1) their ability to extrapolate to regions far from streamflow gauges and (2) their ability to make credible predictions of long-term (decadal-scale) change trends. We evaluated the models based on daily hydrograph metrics (Nash–Sutcliffe model efficiency coefficient, etc.) and predicted decadal streamflow trends. For prediction in ungauged basins (PUB; randomly sampled ungauged basins representing spatial interpolation), δ models either approached or surpassed the performance of LSTM in daily hydrograph metrics, depending on the meteorological forcing data used. They presented a comparable trend performance to LSTM for annual mean flow and high flow but worse trends for low flow. For prediction in ungauged regions (PUR; regional holdout test representing spatial extrapolation in a highly data-sparse scenario), δ models surpassed LSTM in daily hydrograph metrics, and their advantages in mean and high flow trends became prominent. In addition, an untrained variable, evapotranspiration, retained good seasonality even for extrapolated cases. The δ models' deep-network-based parameterization pipeline produced parameter fields that maintain remarkably stable spatial patterns even in highly data-scarce scenarios, which explains their robustness. Combined with their interpretability and ability to assimilate multi-source observations, the δ models are strong candidates for regional and global-scale hydrologic simulations and climate change impact assessment.
54 ENVIRONMENTAL SCIENCES↗
Social vulnerability and power loss mitigation: A case study of Puerto Rico
The increasing occurrence of extreme weather events urges us to reevaluate the resiliency and vulnerability aspects of our most critical infrastructures — such as power grids — as their failures result in both economic loss and severe human hardship. Seen through the lens of alleviating human suffering, it is crucial to be able to identify critical system components of the infrastructure for targeted hardening given resource constraints. This effort is of particular importance in islanded areas such as Puerto Rico where hurricanes are frequent and resources are limited, and where the spatially diverse effects of power loss on human suffering are all the more severe. Recent studies on evaluating infrastructure networks during extreme weather events have taken a simulation based approach that incorporates a variety of component models, such as weather realizations, topological network models, fragility models, and power flow models to estimate expected loss of service. Here, in this work, we expand such a Component Based Event Simulation (CBES) methodology proposed in the literature and integrate it with a social vulnerability modeling component. This paradigm-advancing approach of synthesizing the cutting edge capability of power network modeling and the social impacts of the power transmission network failure is demonstrated for the island of Puerto Rico. Our work exemplifies the efficacy of this integrated modeling framework in developing a decision metric for targeted transmission line hardening.
24 POWER TRANSMISSION AND DISTRIBUTION↗
GridOPTICS/GridPACK
GridPACK is a software framework consisting of a set of modules designed to simplify the development of programs that model the power grid and run on parallel, high performance computing platforms. It also contains several fully developed applications, including powerflow, dynamic simulation, state estimation, Kalman filter analysis (dynamic state estimation), contingency analysis and real time path rating. These applications can be used either standalone or as components in more complicated workflows that combine several different types of application together. The framework modules are available as a combination of libraries and software templates and consist of components for setting up and distributing power grid networks, support for modeling the behavior of individual buses and branches in the network, converting the network models to the corresponding algebraic equations, and parallel routines for manipulating and solving large algebraic systems. The framework also contains a module for distributing tasks evenly amongst computing resources, even if individual tasks vary widely in their execution times. Additional modules support input and output, basic statistical analysis of contingency based calculations, distributed data structures, as well as basic profiling and error management.
Palmer, Bruce↗
Neural networks for parameter estimation in intractable models
The goal is to use deep learning models to estimate parameters in statistical models when standard likelihood estimation methods are computationally infeasible. For instance, inference for max-stable processes is exceptionally challenging even with small datasets, but simulation is straightforward. Data from model simulations are used to train deep neural networks and learn statistical parameters from max-stable models. The proposed neural network-based method provides a competitive alternative to current approaches, as demonstrated by considerable accuracy and computational time improvements. Finally, it serves as a proof of concept for deep learning in statistical parameter estimation and can be extended to other estimation problems.
97 MATHEMATICS AND COMPUTING↗
Locatability of faults in combinational networks.
A formal model for the study of reliable combinational networks is introduced and used to determine network properties conducive to the location of faults. The usual concept of fault location is generalized to be an interval on the partially ordered set of subsets of network nodes that classifies nodes into three disjoint sets: a faulty set, a fault-free set, and an indeterminate set. After developing basic properties of locatable faults, necessary and sufficient conditions for a fault to be locatable in an arbitrary network are stated and proven.
Gray, F. G.↗
Modeling the U.S. Western Electric Interconnection to Understand the Consequences of Hydrometeorological Extremes and Options for Risk Mitigation
Electricity grid operators around the world face a dual challenge; withstanding increasingly severe weather and the longer term impacts of climate change, while simultaneously decarbonizing. Extreme weather events such as heat waves and droughts are rising in both severity and frequency, which is threatening the reliability of electricity grids through increased demand, generation capacity losses, and equipment failures. Consequently, incorporating hydrometeorological stressors into computational power systems analysis is becoming an even more critical tool in long term planning and short term operations. However, there is a general lack of open-source customizable grid simulation software capable of exhaustively stress testing the grid under hydrometeorological uncertainty, and/or examining potential risk mitigation pathways. A related, persistent challenge for power system modelers is striking an appropriate balance between model fidelity (e.g. spatial scale and time resolution) and computational tractability (wall clock run-time). In this study, we are proposing a solution to this problem with open-source software that allows users to seamlessly customize the scale and track the accuracy of grid operations models. Our approach allows users to search over numerous model parameters (network topology, mathematical formulation, economic hurdle rates, and transmission line scaling) to identify model instantiations that accommodate experimental design. Further, we use this approach to demonstrate the importance of including extreme weather events in model validation and model selection. Focusing on the occurrence of heatwaves and droughts in the U.S. Western Interconnection, we examine role of extreme events in balancing tradeoffs between model fidelity and run-time at the model design stage.
Economics↗
Modeling design and control problems involving neural network surrogates
Here, we consider nonlinear optimization problems that involve surrogate models represented by neural networks. We demonstrate first how to directly embed neural network evaluation into optimization models, highlight a difficulty with this approach that can prevent convergence, and then characterize stationarity of such models. We then present two alternative formulations of these problems in the specific case of feedforward neural networks with ReLU activation: as a mixed-integer optimization problem and as a mathematical program with complementarity constraints. For the latter formulation we prove that stationarity at a point for this problem corresponds to stationarity of the embedded formulation. Each of these formulations may be solved with state-of-the-art optimization methods, and we show how to obtain good initial feasible solutions for these methods. We compare our formulations on three practical applications arising in the design and control of combustion engines, in the generation of adversarial attacks on classifier networks, and in the determination of optimal flows in an oil well network.
97 MATHEMATICS AND COMPUTING↗
Junction angles in drainage networks
A theoretical model relating the junction angle of a tributary in a drainage network to its own Shreve magnitude and the Shreve magnitude of the stream into which it flows is developed and verified using published data (including Landsat maps) on U.S. networks with different underlying rock types, relief, and basin lengths. Good agreement is found, and a tendency for the slopes of individual network links to evolve toward graded longitudinal profiles as the network approaches an equilibrium (graded') planimetric pattern is noted. The differences between the network patterns predicted by the model and those observed on Mars are indicated, confirming the results of Pieri (1979, 1980).
Pieri, D. C.↗
Pre-trained Microscopy Models for Segmentation and Microstructure Quantification
We introduce pre-trained convolutional neural network (CNN) models for quantitative microscopy analysis. Pre-trained weights are available for download and can be used in place of ImageNet weights with just a couple lines of code.
machine learning↗
Data-driven cyber-attack detection for photovoltaic systems: A transfer learning approach
With increasing exposure to software-based sensing and control, power systems are facing higher risks of cyber/physical attacks. Here, to ensure system stability and minimize the potential economic losses, it is imperative to monitor the operating states and detect those attacks at the early stage. In this paper, a transfer learning method is proposed to detect cyber-attacks in photovoltaic (PV) systems with much less training data. First of all, two PV systems with a different number of PV inverters and power ratings are analyzed and their attack models are studied. Next, an attack detection Convolutional Neural Network (CNN) model was trained with rich amount of data from PV #1. Then, transfer learning was proposed to transfer the well-trained features from PV #1 to PV #2. Lastly, the attack detection model on PV #2 was trained based on the transferred CNN model. The experiment results show that the proposed transfer learning method achieves better accuracy and a faster convergence rate with a much less training dataset than conventional deep learning.
14 SOLAR ENERGY↗
Simplified Model and Approach to Transform Infrared Surface Temperature to Film Effectiveness in a Conjugate Heat Transfer Experiment
In the pursuit of more efficient gas turbines, film cooling is a critical technology. This article describes a simplified engineering model based on a one-dimensional thermal resistance network. The model is used to relate film-cooling effectiveness and heat transfer augmentation to local overall cooling effectiveness in a conjugate flat plate experiment. Here, this article presents experimental proof-of-concept data to demonstrate the potential for this model. In contrast to previous approaches, neither the wall heat flux nor the adiabatic wall temperature is required to estimate the local film-cooling performance parameters. The model predicts surface temperatures that are within the experimental uncertainties over the range for which the model is trained and to within five percent when the model is extrapolated to higher coolant channel Reynolds numbers. This article is relevant to conjugate test rigs that can measure the hot-surface temperature distribution with and without film cooling. This information may also be relevant to designers as a method to approximate surface temperatures or used as an approximate heat transfer model for optimization studies.
advanced turbines↗
Blending Pipeline Analysis Tool for Hydrogen (BlendPATH) Documentation and User Manual
The Blending Pipeline Analysis Tool for Hydrogen (BlendPATH) is a flexible, open-source Python tool designed to provide users with case-by-case analysis capabilities to identify the necessary modifications to repurpose existing natural gas transmission pipeline networks to transport hydrogen as a blend of a user-specified volume fraction of hydrogen or as a pure stream and to estimate the associated capital and operating expenditures resulting from those modifications. This tool is intended to be applied during the initial screening stage of a prospective project when pipeline developers compile transmission pipeline technical documentation and history but prior to performing detailed pipeline inspections. Performing analysis with BlendPATH during this initial screening stage can provide the user with an understanding of promising opportunities and probable economic outcomes of repurposing their pipeline network for hydrogen before proceeding with detailed pipeline materials testing and pipeline inspections. BlendPATH consists of multiple modules to simulate, assess, and modify existing natural gas transmission pipeline network designs to be compatible with hydrogen as a blend or pure stream. The tool employs an open-source gas network hydraulic model to simulate existing, user-specified transmission pipeline networks, and it applies ASME B31.12 to assess the pipe segments within these networks for compatibility with hydrogen and to modify the networks to achieve compatibility where the existing infrastructure is inadequate. Users can specify ASME B31.12 design options for pipeline assessment and can select from multiple methods for pipeline modification. This report details the functionalities of BlendPATH Version 2.0.2 and demonstrates an example of applying BlendPATH to a case study. Potential and intended users of this framework include natural gas pipeline developers and operators and researchers at both public and private institutions. BlendPATH is publicly available at https://github.com/NREL/BlendPATH.
08 HYDROGEN↗
Calculation of flow about posts and powerhead model
A three dimensional analysis of the non-uniform flow around the liquid oxygen (LOX) posts in the Space Shuttle Main Engine (SSME) powerhead was performed to determine possible factors contributing to the failure of the posts. Also performed was three dimensional numerical fluid flow analysis of the high pressure fuel turbopump (HPFTP) exhaust system, consisting of the turnaround duct (TAD), two-duct hot gas manifold (HGM), and the Version B transfer ducts. The analysis was conducted in the following manner: (1) modeling the flow around a single and small clusters (2 to 10) of posts; (2) modeling the velocity field in the cross plane; and (3) modeling the entire flow region with a three dimensional network type model. Shear stress functions which will permit viscous analysis without requiring excessive numbers of computational grid points were developed. These wall functions, laminar and turbulent, have been compared to standard Blasius solutions and are directly applicable to the cylinder in cross flow class of problems to which the LOX post problem belongs.
Anderson, P. G.↗
Simplified Model and Approach to Transform Infrared Surface Temperature to Film Effectiveness in a Conjugate Heat Transfer Experiment
This paper describes a simplified engineering model based on a one-dimensional thermal resistance network. The model is used to develop a new method to relate film cooling effectiveness and heat transfer augmentation to local overall cooling effectiveness in a conjugate flat plate experiment. This paper presents experimental proof-of-concept data to demonstrate the potential for this model. In contrast to previous approaches, neither the wall heat flux nor the adiabatic wall temperature is required to estimate the local film cooling performance parameters. The model predicts surface temperatures that are within the experimental uncertainties over the range for which the model is trained, and to within five percent when the model is extrapolated to higher coolant channel Reynolds numbers. This paper is relevant to conjugate test rigs that can measure the hot surface temperature distribution with and without film cooling. This information may also be relevant to designers as a method to approximate surface temperatures or used as an approximate heat transfer model for optimization studies.
advanced gas turbines↗
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↗
The Airspace Concepts Evaluation System Architecture and System Plant
The Airspace Concepts Evaluation System is a simulation of the National Airspace System. It includes models of flights, airports, airspaces, air traffic controls, traffic flow managements, and airline operation centers operating throughout the United States. It is used to predict system delays in response to future capacity and demand scenarios and perform benefits assessments of current and future airspace technologies and operational concepts. Facilitation of these studies requires that the simulation architecture supports plug and play of different air traffic control, traffic flow management, and airline operation center models and multi-fidelity modeling of flights, airports, and airspaces. The simulation is divided into two parts that are named, borrowing from classical control theory terminology, control and plant. The control consists of air traffic control, traffic flow management, and airline operation center models, and the plant consists of flight, airport, and airspace models. The plant can run open loop, in the absence of the control. However, undesired affects, such as conflicts and over congestions in the airspaces and airports, can occur. Different controls are applied, "plug and played", to the plant. A particular control is evaluated by analyzing how well it managed conflicts and congestions. Furthermore, the terminal area plants consist of models of airports and terminal airspaces. Each model consists of a set of nodes and links which are connected by the user to form a network. Nodes model runways, fixes, taxi intersections, gates, and/or other points of interest, and links model taxiways, departure paths, and arrival paths. Metering, flow distribution, and sequencing functions can be applied at nodes. Different fidelity model of how a flight transits are can be used by links. The fidelity of the model can be adjusted by the user by either changing the complexity of the node/link network-or the way that the link models how the flights transit from one node to the other.
Windhorst, Robert↗
Link Scheduling in Satellite Networks via Machine Learning Over Riemannian Manifolds
Low Earth Orbit (LEO) satellites play a crucial role in enhancing global connectivity, serving a complementary solution to existing terrestrial systems. In wireless networks, scheduling is a vital process that allocates time-frequency resources to users for interference management. However, LEO satellite networks face significant challenges in scheduling their links towards ground users due to the satellites’ mobility and overlapping coverage. This paper addresses the dynamic link scheduling problem in LEO satellite networks by considering spatio-temporal correlations introduced by the satellites’ movements. The first step in the proposed solution involves modeling the network over Riemannian manifolds, thanks to their representation as symmetric positive definite matrices. We introduce two machine learning (ML)-based link scheduling techniques that model the dynamic evolution of satellite positions and link conditions over time and space. To accurately predict satellite link states, we present a recurrent neural network (RNN) over Riemannian manifolds, which captures spatio-temporal characteristics over time. Furthermore, we introduce a separate model, the convolutional neural network (CNN) over Riemannian manifolds, which captures geometric relationships between satellites and users by extracting spatial features from the network topology across all links. Simulation results demonstrate that both RNN and CNN over Riemannian manifolds deliver comparable performance to the fractional programming-based link scheduling (FPLinQ) benchmark. Remarkably, unlike other ML-based models that require extensive training data, both models only need 30 training samples to achieve over 99% of the sum rate while maintaining similar computational complexity relative to the benchmark.
42 ENGINEERING↗