Search NASA⌕ Search

SEARCH · Search NASA

Results for “Complex Network Data”

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 1,045 records · Page 58

Enabling site-specific well leakage risk estimation during geologic carbon sequestration using a modular deep-learning-based wellbore leakage model

Geologic carbon sequestration (GCS) is a promising technology for mitigating net carbon emissions and growing climate concern by storing CO 2 in reservoirs. Oil and gas brownfields are an attractive option for CO 2 storage, but these sites have many historical wellbores from petroleum production and can be a potential leakage pathway for CO 2 or formation brine. Therefore, risk management of GCS operations requires an assessment of potential well leakage. Due to the high uncertainty of the system, stochastic approaches are ideal for quantifying the range of risk behaviors, but they must be computationally efficient in the face of complex physics. Here, we develop a new physics-centric deep learning wellbore model to predict the leakage of CO 2 and brine through leaky wellbores. Multi-physics numerical simulations were used to generate data sets, and physics-informed features were introduced. Neural networks were optimized with an automated searching algorithm. Feature analysis quantifies the impact of each feature on model prediction and confirms the role of physics-inspired parameters. The model shows high predictive performance across a wide range of geologic and injection conditions and well attributes. In conclusion, a case study illustrates how the model is applied to assess well leakage in GCS operations.

58 GEOSCIENCES↗

Molecular arrangement in water: random but not quite.

Water defines life on Earth from the cellular to the terrestrial level. Yet the molecular level arrangement in water is not well understood, posing problems in comprehending its very special chemical, physical and biological properties. Here we present high-resolution x-ray diffraction data for water clearly showing that its molecular arrangement exhibits specific correlations that are consistent with the presence of rings of H{sub 2}O molecules linked together by hydrogen bonds into tetrahedral-like units from a continuous network. This level of molecular arrangement complexity is beyond what a simple 'two-state' model of water (Bernal and Fowler 1933 J. Chem. Phys.1 515-48) could explain. It may not be explained by the recently put forward 'chains-clusters of completely uncorrelated molecules' model (Wernet et al 2004 Science 304 995-9) either. Rather it indicates that water is homogeneous down to the molecular level where different water molecules form tetrahedral units of different perfection and/or participate in rings of different sizes, thus experiencing different local environments. The local diversity of this tetrahedral network coupled to the flexibility of the hydrogen bonds that hold it together may explain well the rich phase diagram of water and why it responds non-uniformly to external stimuli such as, for example, temperature and pressure.

08 HYDROGEN↗

A Survey of Open-Source Tools for Transmission and Distribution Systems Research

This work presents a review of open-source electric power transmission and distribution systems analysis tools suitable for use by industry professionals and academic researchers. Due to the high complexity of the electric grid, there exist numerous tools and extensive research pertaining to nearly every aspect of the design, operation, and control of transmission and distribution networks. In addition to the commercial tools, a wide range of free, open-source tools, models, and data usable by the scientific community for related research have been developed by different organizations, including both international and US universities and national laboratories. However, due to the absence of a catalog of available tools, models and data, researchers often lack a knowledge of existing capabilities and may develop duplicative software and tools. Increasing awareness of these available resources seeks to accelerate their broader use, leading to more efficient and standardized grid analysis. This review paper (which is part of a larger survey effort that studied over 400 tools in the transmission, distribution, buildings, and electric vehicles space) outlines selected open-source resources that have been developed in power transmission and distribution systems research. It is anticipated that this work can serve as a guide for industry and academic researchers alike, ensuring that research efforts are well-channeled.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Data from the Development Evolution of a Vehicle for Custom Control

In order to develop custom controllers intended to operate vehicles on a live highway, a series of data collection-focused tests were performed at increasing stages of complexity. Modern vehicles with features like Adaptive Cruise Control (ACC) feature a rich set of sensors and drive-by-wire mechanisms. The presented stages of data collection begins with the analysis of raw data provided by various vehicles, and eventually results in spoofing Controller Area Network (CAN) protocols for sending control commands to operate a vehicle. This paper covers the data and technical efforts needed at various stages. The raw data and tools to plot the data are also publicly available.

99 GENERAL AND MISCELLANEOUS↗

FNETVision: A WAMS Big Data Knowledge Discovery System

With the growing awareness of the power system stability, wide area measurement system (WAMS) has been widely studied and deployed. In WAMS, a phasor data concentrator (PDC) collects and persists the GPS-time-synchronized phasor data from hundreds of phasor measurement units (PMU). Due to the size and complexity of the modern power systems, the number of deployed PMUs has been increasing and their data size becomes increasingly large. The large data volume challenges the analysis, and the knowledge discovery of the measurement data. This paper takes the pioneering distribution level WAMS- the frequency monitoring network FNET/GridEye as an example. It develops an adaptive visualization system- FNETVision to visualize the large-volume WAMS data by interconnection. The case studies suggest, based on the FNETVision system, the system administrator can quickly dig out issues of the FNET system and accordingly pinpoint the cause of these issues.

Wang, Weikang↗

Non-Intrusive Appliance Identification with Appliance-Specific Networks

The problem of noninstrusive load monitoring (NILM) is usually formulated as a single-channel blind source separation task, whose successful solution enable fast and convenient load identification and energy disaggregation. When applied at test time, NILM algorithms aim to identify the operating characteristics of individual appliances from an aggregate power measurement of the entire house. Recent advances in deep learning gave rise to many methods that mostly focus on learning a direct mapping from aggregate measurement to individual appliance power. However, these methods are not only computationally expensive, but they often suffer from overfitting and do not generalize very well. In this article, we propose a novel NILM method that leverages advances in statistical learning that have not been properly applied in this domain before. The proposed method consists of three stages: first, a Bayesian nonparametric learning-based approach for appliance state extraction; second, synthetic minority oversampling technique for data augmentation and mitigating the heavy imbalance in switching events; and third, appliance-specific lightweight long short-term memory networks for status classification for each appliance. Here, we adopt a “differential” input (the difference before and after the switching event) to reduce the complexity of network training and make the proposed method robust to multiappliance switching events. Experiments are conducted to demonstrate the effectiveness of the proposed method, achieving superior performance when compared to recent methods. An ablation study is conducted to demonstrate the effectiveness of each module of our method. Finally, we investigate the quality of generated synthetic samples.

42 ENGINEERING↗

Benchmarking Parametric and Machine Learning Models for Genomic Prediction of Complex Traits

The usefulness of genomic prediction in crop and livestock breeding programs has prompted efforts to develop new and improved genomic prediction algorithms, such as artificial neural networks and gradient tree boosting. However, the performance of these algorithms has not been compared in a systematic manner using a wide range of datasets and models. Using data of 18 traits across six plant species with different marker densities and training population sizes, we compared the performance of six linear and six non-linear algorithms. First, we found that hyperparameter selection was necessary for all non-linear algorithms and that feature selection prior to model training was critical for artificial neural networks when the markers greatly outnumbered the number of training lines. Across all species and trait combinations, no one algorithm performed best, however predictions based on a combination of results from multiple algorithms (i.e., ensemble predictions) performed consistently well. While linear and non-linear algorithms performed best for a similar number of traits, the performance of non-linear algorithms vary more between traits. Although artificial neural networks did not perform best for any trait, we identified strategies (i.e., feature selection, seeded starting weights) that boosted their performance to near the level of other algorithms. Our results highlight the importance of algorithm selection for the prediction of trait values.

59 BASIC BIOLOGICAL SCIENCES↗

An Efficient Storage-Driven Machine Learning Model for Performance in the Era of Multimodal Scientific Data

Scientific workflows are increasingly relying on machine learning (ML), simulation, and hybrid techniques to predict, understand, and optimize the behavior of complex experiments. High-performance computing has greatly improved researchers’ ability to acquire diverse data modalities in these workflows. Recent studies suggest that the performance of machine learning models can be improved by integrating data from various sources. Unfortunately, these workloads pose unprecedent pressure on the network storage to meet the demands associated with accessing these multimodal data. To mitigate the impact of intensive IO, we propose a solution that utilizes a multi-tier High-Performance Computing (HPC) distributed storage and data processing framework, placing computation where the data resides for better performance. By adopting this project, the scientific community will gain new opportunities to explore multimodal storage-driven possibilities, integrating multiple scientific data sources with advanced streaming frameworks. Additionally, our framework effectively utilizes computing resources and bridges the gaps identified by HPC experts. Our proposed approach tackles scalability and persistence challenges by leveraging native persistency, which has posed difficulties in traditional approaches. Furthermore, we seek to enhance fault-tolerance and load-balance of computations by leveraging real-time streaming in diverse scientific computing environments, thereby propelling advanced scientific computing research into the next generation.

97 MATHEMATICS AND COMPUTING↗

Graph learning for particle accelerator operations

Particle accelerators play a crucial role in scientific research, enabling the study of fundamental physics and materials science, as well as having important medical applications. This study proposes a novel graph learning approach to classify operational beamline configurations as good or bad. By considering the relationships among beamline elements, we transform data from components into a heterogeneous graph. We propose to learn from historical, unlabeled data via our self-supervised training strategy along with fine-tuning on a smaller, labeled dataset. Additionally, we extract a low-dimensional representation from each configuration that can be visualized in two dimensions. Leveraging our ability for classification, we map out regions of the low-dimensional latent space characterized by good and bad configurations, which in turn can provide valuable feedback to operators. This research demonstrates a paradigm shift in how complex, many-dimensional data from beamlines can be analyzed and leveraged for accelerator operations.

43 PARTICLE ACCELERATORS↗

Understanding Contaminant Migration within a Dynamic River Corridor through Field Experiments and Reactive Transport Modeling

The behavior of a persistent uranium plume within an extended river corridor at the DOE Hanford site is dominantly controlled by river stage fluctuations in the adjacent Columbia River. The plume behavior is further complicated by substantial heterogeneity in physical and geochemical properties of the host aquifer sediments. Multi-scale field and laboratory experiments and reactive transport modeling were integrated to understand the complex plume behavior influenced by highly variable hydrologic and geochemical conditions in time and space. In this paper, we (1) describe multiple data sets from field-scale uranium adsorption and desorption experiments performed at our experimental well-field, (2) develop a reactive transport model that incorporates hydrologic and geochemical heterogeneities characterized from multi-scale and multi-type datasets and a surface complexation reaction network based on laboratory studies, and (3) compare the modeling and observation results to provide insights on how to refine the conceptual model and reduce prediction uncertainties. The experimental results revealed significant spatial variability in uranium adsorption/desorption behavior, while modeling demonstrated that ambient hydrologic and geochemical conditions and heterogeneities in sediment physical and chemical properties both contributed to complex plume behavior and its persistence. This research underscores the great challenges in adequately characterizing this type of site to model the reactive transport processes over scales of 10 m or more. Our analysis provides important insights into the characterization, understanding, modeling, and remediation of groundwater contaminant plumes influenced by dynamic surface water and groundwater interactions.

Chen, Xingyuan↗

Benchmarking parametric and machine learning models for genomic prediction of complex traits

The usefulness of genomic prediction in crop and livestock breeding programs has prompted efforts to develop new and improved genomic prediction algorithms, such as artificial neural networks and gradient tree boosting. However, the performance of these algorithms has not been compared in a systematic manner using a wide range of datasets and models. Using data of 18 traits across six plant species with different marker densities and training population sizes, we compared the performance of six linear and six non-linear algorithms. First, we found that hyperparameter selection was necessary for all non-linear algorithms and that feature selection prior to model training was critical for artificial neural networks when the markers greatly outnumbered the number of training lines. Across all species and trait combinations, no one algorithm performed best, however predictions based on a combination of results from multiple algorithms (i.e. ensemble predictions) performed consistently well. While linear and non-linear algorithms performed best for a similar number of traits, the performance of non-linear algorithms vary more between traits. Although artificial neural networks did not perform best for any trait, we identified strategies (i.e. feature selection, seeded starting weights) that boosted their performance to near the level of other algorithms. Our results highlight the importance of algorithm selection for the prediction of trait values.

59 BASIC BIOLOGICAL SCIENCES↗

Operations for Learning with Graphical Models

This paper is a multidisciplinary review of empirical, statistical learning from a graphical model perspective. Well-known examples of graphical models include Bayesian net- works, directed graphs representing a Markov chain, and undirected networks representing a Markov field. These graphical models are extended to model data analysis and empirical learning using the notation of plates. Graphical operations for simplifying and manipulating a problem are provided including decomposition, differentiation, and the manipulation of probability models from the exponential family. These operations adapt existing techniques from statistics and automatic differentiation to graphs. Two standard algorithm schemes for learning are reviewed in a graphical framework: Gibbs sampling and the expectation maximization algorithm. Some algorithms are developed in this graphical framework including a generalized version of linear regression, techniques for feed-forward networks, and learning Gaussian and discrete Bayesian networks from data. The paper concludes by sketching some implications for data analysis and summarizing some popular algorithms that fall within the framework presented. The main original contributions here are the decomposition techniques and the demonstration that graphical models provide a framework for understanding and developing complex learning algorithms.

Buntine, Wray L.↗

Simultaneous Retrieval of Aerosol and Surface Optical Properties from Combined Airborne- and Ground-Based Direct and Diffuse Radiometric Measurements

This paper presents a new method for simultaneously retrieving aerosol and surface reflectance properties from combined airborne and ground-based direct and diffuse radiometric measurements. The method is based on the standard Aerosol Robotic Network (AERONET) method for retrieving aerosol size distribution, complex index of refraction, and single scattering albedo, but modified to retrieve aerosol properties in two layers, below and above the aircraft, and parameters on surface optical properties from combined datasets (Cloud Absorption Radiometer (CAR) and AERONET data). A key advantage of this method is the inversion of all available spectral and angular data at the same time, while accounting for the influence of noise in the inversion procedure using statistical optimization. The wide spectral (0.34-2.30 m) and angular range (180 ) of the CAR instrument, combined with observations from an AERONET sunphotometer, provide sufficient measurement constraints for characterizing aerosol and surface properties with minimal assumptions. The robustness of the method was tested on observations made during four different field campaigns: (a) the Southern African Regional Science Initiative 2000 over Mongu, Zambia, (b) the Intercontinental Transport Experiment-Phase B over Mexico City, Mexico (c) Cloud and Land Surface Interaction Campaign over the Atmospheric Radiation Measurement (ARM) Central Facility, Oklahoma, USA, and (d) the Arctic Research of the Composition of the Troposphere from Aircraft and Satellites (ARCTAS) over Elson Lagoon in Barrow, Alaska, USA. The four areas are dominated by different surface characteristics and aerosol types, and therefore provide good test cases for the new inversion method.

Gatebe, C. K.↗

Large-Scale Solar Panel Mapping from Aerial Images Using Deep Convolutional Networks

Up-to-date maps of installed solar photovoltaic panels are a critical input for policy and financial assessment of solar distributed generation. However, such maps for large areas are not available. With high coverage and low cost, aerial images enable large-scale mapping, bit it is highly difficult to automatically identify solar panels from images, which are small objects with varying appearances dispersed in complex scenes. We introduce a new approach based on deep convolutional networks, which effectively learns to delineate solar panels in aerial scenes. The approach has successfully mapped solar panels in imagery covering 200 square kilometers in two cities, using only 12 square kilometers of training data that are manually labeled.

Yuan, Jiangye↗

PersGNN: Applying Topological Data Analysis and Geometric Deep Learning to Structure-Based Protein Function Prediction

Understanding protein structure-function relationships is a key challenge in computational biology, with applications across the biotechnology and pharmaceutical industries. While it is known that protein structure directly impacts protein function, many functional prediction tasks use only protein sequence. In this work, we isolate protein structure to make functional annotations for proteins in the Protein Data Bank in order to study the expressiveness of different structure-based prediction schemes. We present PersGNN - an end-to-end trainable deep learning model that combines graph representation learning with topological data analysis to capture a complex set of both local and global structural features. While variations of these techniques have been successfully applied to proteins before, we demonstrate that our hybridized approach, PersGNN, outperforms either method on its own as well as a baseline neural network that learns from the same information. PersGNN achieves a 9.3% boost in area under the precision recall curve (AUPR) compared to the best individual model, as well as high F1 scores across different gene ontology categories, indicating the transferability of this approach.

Swenson, Nicolas↗

Hierarchical Resilience Planning for Networked Microgrids: A Case Study of Puerto Rico

Microgrids can be designed to enhance the energy resilience of communities and critical infrastructures, such as hospitals, data centers, and communication networks, which are vulnerable to frequent weather-related disruption. Coordinating multiple microgrids in a network can leverage the geographical diversity of load and generation resources while enabling resilient and cost-effective planning of the distribution system. Designing a networked microgrid is complex, involving intricate technical assessment, cost-benefit analysis, site-specific requirements, and the evaluation of existing resources. Therefore, this paper proposes a hierarchical resilience planning framework and performs an extensive techno-economic analysis for the design of a networked microgrid. Hierarchical resilience planning involves technology sizing at an individual community level to meet the critical load and satisfy resilience criteria, and resource optimization at networked microgrid level to provide a higher level of resilience and energy adequacy. A real-world case of Puerto Rico's cooperative microgrid “Microrred de la Montaña” is investigated considering localized electricity tariffs, site-specific demand profiles, solar generation, and existing hydro resources. Multiple optimization scenarios are developed based on the resiliency requirement to estimate the capacity of solar photovoltaic and battery energy storage (BES) to be installed at each substation. The results provide the optimal sizing for individual community and networked microgrid to withstand 1day and 3-day outages along with the criteria for critical load.

13 - HYDRO ENERGY↗

Mortandad Canyon seismic modeling (Progress Report)

The goal of this project is to assess site amplification for LANL through modeling of site and topography effects. Such assessment is difficult due to the lack of data documenting site effects for large normal faulting earthquakes such as the Pajarito Fault System that was found to be governing the seismic hazard at LANL. LANL site is also characterized by unique geologic settings, a series of mesa at the edge of the huge volcanic complex of the Jemez caldera. The latter has produced a series of successive and intertwined deposits of volcanic flow and sedimentary layer. Such layering of material with different elastic properties is expected to create large amplification effects magnified further by the presence at the surface of tuff materials. In 2014, a seismic campaign was performed in the Mortandad Canyon with a network of sensors on and near a local mesa. That campaign was designed to provide experimental data to estimation and prediction of amplification and site effects at LANL. An analysis on the seismic noise records of that campaign shows evidence of amplification of the seismic amplitude for the stations on the top of the mesa. Our goal is to reproduce the effect observed (both retrieving the resonance frequencies and amplification levels). We also plan to test if amplification effects observed for seismic noise can be extended to shaking produced by earthquakes. We are using a 3D modeling tool based on the Spectral Element Method (SEM) that is a particular type of high-order finite element modeling. Two types of seismic sources will be considered: (1) earthquakes for purpose of seismic hazard evaluation, (2) ambient noise. We show 2 series of numerical computations: (1) seismic waves are generated by random sources to simulate ambient noise. (2) seismic waves are planar waves to simulate a distant earthquake. We plan in the future to model rupture scenario on the Pajarito Fault System.

58 GEOSCIENCES↗

Single-Point Extrapolation to the Complete Basis Set Limit through Deep Learning

Machine learning (ML) offers an attractive method for making predictions about molecular systems while circumventing the need to run expensive electronic structure calculations. Once trained on ab initio data, the promise of ML is to deliver accurate predictions of molecular properties that were previously computationally infeasible. In this work, we develop and train a graph neural network model to correct the basis set incompleteness error (BSIE) between a small and large basis set at the RHF and B3LYP levels of theory. Our results show that, when compared to fitting to the total potential, an ML model fitted to correct the BSIE is better at generalizing to systems not seen during training. We test this ability by training on single molecules while evaluating on molecular complexes. We also show that ensemble models yield better behaved potentials in situations where the training data is insufficient. However, even when only fitting to the BSIE, acceptable performance is only achieved when the training data sufficiently resemble the systems one wants to make predictions on. As a result, the test error of the final model trained to predict the difference between the cc-pVDZ and cc-pV5Z potential is 0.184 kcal/mol for the B3LYP density functional, and the ensemble model accurately reproduces the large basis set interaction energy curves on the S66x8 dataset.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗