Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “network reduction”

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 991 records · Page 55
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Hierarchically Templated Synthesis of 3D‐Printed Crosslinked Cyclodextrins for Lycopene Harvesting

Abstract Plants produce a wide range of bioactive phytochemicals, such as antioxidants and vitamins, which play crucial roles in aging prevention, inflammation reduction, and reducing the risk of cancer. Selectively harvesting these phytochemicals, such as lycopene, from tomatoes through the adsorption method is cost‐effective and energy efficient. In this work, a templated synthesis of 3D‐printed crosslinked cyclodextrin polymers featuring nanotubular structures for highly selective lycopene harvesting is reported. Polypseudorotaxanes formed by triethoxysilane‐based telechelic polyethylene glycols and α‐cyclodextrins (α‐CDs) are designed as the template to (1) synthetically access urethane‐based nanotubular structures at the molecular level, and (2) construct 3D‐printed architectures with designed macroscale voids. The polypseudorotaxane hydrogels showed good rheological properties for direct ink writing, and the 3D‐printed hydrogels were converted to the desired α‐CD polymer network through a three‐step postprinting transformation. The obtained urethane‐crosslinked α‐CD monoliths possess nanotubular structures and 3D‐printed voids. They selectively adsorb lycopene from raw tomato juice, protecting lycopene from photo‐ or thermo‐degradations. This work highlights the hierarchically templated synthesis approach in developing functional 3D‐printing materials by connecting the bottom‐up molecular assembly and synthesis with the top‐down 3D architecture control and fabrication.

Chemistry↗
DOE OSTI2024

Unusual Electrical Conductivity Enhancement in Stable n‐Type Carbon Nanotube Networks

Organic molecule-doped n-type single-walled carbon nanotube (SWCNT) networks are promising candidates for advanced energy applications, such as flexible thermoelectrics and photovoltaics. Yet charge transport in n-type SWCNTs is limited by two factors: i) charge localization impeding inter-tube transport caused by disordered mesostructure of randomly oriented SWCNTs and ii) reduction of charge carrier concentration driven by oxidation. Herein, studied the relationship between the mesostructure and thermoelectric properties of n-type SWCNTs obtained by surfactant-functionalization and polymer-dopant grafting. Surprisingly, the electrical conductivity of the polymer-doped SWCNTs keeps increasing with increasing polymer content, even after the saturation of carrier concentration, resulting in 12x higher conductivity on polymer-doping compared to surfactant-functionalization. While hopping transport typically dominates in disordered systems, it is shown that a bridging effect from the polymer causes unusual band-like conduction in polymer-doped SWCNTs. Additionally, since surfactants are essential to prevent oxidation and retain n-type over a long duration, shows that SWCNTs obtained through a dual-functionalization strategy using both polymer-dopant and surfactant, demonstrates a long-term stable high n-type thermoelectric power factor, when the surfactant amount is carefully controlled. Besides thermoelectrics, the findings are of general interest to developing stable and conductive n-type SWCNTs for various energy and electronic applications.

carbon nanotubes↗
DOE OSTI2014

A probabilistic graphical model based stochastic input model construction

Model reduction techniques have been widely used in modeling of high-dimensional stochastic input in uncertainty quantification tasks. However, the probabilistic modeling of random variables projected into reduced-order spaces presents a number of computational challenges. Due to the curse of dimensionality, the underlying dependence relationships between these random variables are difficult to capture. In this work, a probabilistic graphical model based approach is employed to learn the dependence by running a number of conditional independence tests using observation data. Thus a probabilistic model of the joint PDF is obtained and the PDF is factorized into a set of conditional distributions based on the dependence structure of the variables. The estimation of the joint PDF from data is then transformed to estimating conditional distributions under reduced dimensions. To improve the computational efficiency, a polynomial chaos expansion is further applied to represent the random field in terms of a set of standard random variables. This technique is combined with both linear and nonlinear model reduction methods. Numerical examples are presented to demonstrate the accuracy and efficiency of the probabilistic graphical model based stochastic input models. - Highlights: • Data-driven stochastic input models without the assumption of independence of the reduced random variables. • The problem is transformed to a Bayesian network structure learning problem. • Examples are given in flows in random media.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2018

Self-Supported Hydrous Iridium–Nickel Oxide Two-Dimensional Nanoframes for High Activity Oxygen Evolution Electrocatalysts

Oxygen evolution reaction (OER) electrocatalysts with high activity, high stability, and low costs are needed for proton-exchange membrane (PEM) electrolyzers. Based on the high cost and limited supply of iridium, approaches that result in iridium-based OER catalysts with increased catalytic activity are of significant interest. Here, we report a carbon-free, self-supported hydrous iridium-nickel oxide two-dimensional nanoframe structure synthesized by thermal treatment of iridium-decorated nickel oxide nanosheets under reducing conditions and subsequent chemical leaching in acid. The catalyst nanoarchitecture contains an interconnected network of metallic iridium-nickel alloy domains with hydrous iridium oxide and nickel oxide located in the surface region. The electrochemical oxidation step maintains the three-dimensional nanoarchitecture and results in a thin (~5 A) oxide/hydroxide surface layer. The temperature used for thermal reduction was found to strongly affect the catalyst surface structure and OER activity. Utilizing a lower thermal reduction temperature of 200 degrees C was determined to provide a higher relative surface concentration of hydroxides and nickel oxide and result in higher OER activities compared with materials treated at 300 degrees C. The 200 degrees C-treated hydrous iridium-nickel oxide electrocatalyst showed 15 times higher initial OER mass activity than commercial IrO 2 , and the activity remained 10 times higher than IrO 2 after accelerated durability testing. Density functional theory (DFT) calculations and analysis of the experimental Tafel slopes support that the second electron transfer step is the rate-limiting step for the reaction. The DFT calculations demonstrate that Ni substitution on the IrO 2 surface lowers the activation energy for adsorbed intermediates of the second electron transfer step of the OER reaction. This work establishes that noble metal-decorated metal oxide nanosheets can be transformed into high surface area, carbon-free electrocatalyst nanostructures with high catalytic activities and molecular accessibility and reveals the importance of using controlled thermal reduction temperatures to alter the surface structure and OER activity.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Adaptative Survival of Aspergillus fumigatus to Echinocandins Arises from Cell Wall Remodeling Beyond β-1,3-glucan Synthesis Inhibition

Antifungal echinocandins inhibit the biosynthesis of β-1,3-glucan, a major and essential polysaccharide component of the fungal cell wall. However, the efficacy of echinocandins against the pathogen Aspergillus fumigatus is limited. Here, we use solid-state nuclear magnetic resonance (ssNMR) and other techniques to show that echinocandins induce dynamic changes in the assembly of mobile and rigid polymers within the A. fumigatus cell wall. The reduction of β-1,3- glucan induced by echinocandins is accompanied by a concurrent increase in levels of chitin, chitosan and highly polymorphic α-1,3-glucans, whose physical association with chitin maintains cell wall integrity and modulates water permeability. The rearrangement of the macromolecular network is dynamic and controls the permeability and circulation of the drug throughout the cell wall. Thus, our results indicate that echinocandin treatment triggers compensatory rearrangements in the cell wall that may help A. fumigatus to tolerate the drugs’ antifungal effects.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

Using molten salts to probe outer-coordination sphere effects on lanthanide( III )/( II ) electron-transfer reactions

Controlling structure and reactivity by manipulating the outer-coordination sphere around a given reagent represents a longstanding challenge in chemistry. Despite advances toward solving this problem, it remains difficult to experimentally interrogate and characterize outer-coordination sphere impact. Here, this work describes an alternative approach that quantifies outer-coordination sphere effects. It shows how molten salt metal chlorides (MCl n ; M = K, Na, n = 1; M = Ca, n = 2) provided excellent platforms for experimentally characterizing the influence of the outer-coordination sphere cations (M n+ ) on redox reactions accessible to lanthanide ions; Ln 3+ + e 1– → Ln 2+ (Ln = Eu, Yb, Sm; e 1– = electron). As a representative example, X-ray absorption spectroscopy and cyclic voltammetry results showed that Eu 2+ instantaneously formed when Eu 3+ dissolved in molten chloride salts that had strongly polarizing cations (like Ca 2+ from CaCl 2 ) via the Eu 3+ + Cl 1– → Eu 2+ + ½Cl 2 reaction. Conversely, molten salts with less polarizing outer-sphere M 1+ cations (e.g., K 1+ in KCl) stabilized Ln 3+ . For instance, the Eu 3+ /Eu 2+ reduction potential was >0.5 V more positive in CaCl 2 than in KCl. In accordance with first-principle molecular dynamics (FPMD) simulations, we postulated that hard M n+ cations (high polarization power) inductively removed electron density from Ln n+ across Ln–Cl···M n+ networks and stabilized electron-rich and low oxidation state Ln 2+ ions. Conversely, less polarizing M n+ cations (like K 1+ ) left electron density on Lnn+ and stabilized electron-deficient and high-oxidation state Ln 3+ ions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

On-Chip Batteries as Distributed Energy Sources in Heterogeneous 2.5D/3D Integrated Circuits

Energy efficiency in digital systems faces challenges due to the constraints imposed by small-scale transistors. Moreover, the growing demand for portable consumer electronics necessitates the use of compact energy sources. To address these challenges, heterogeneous 3D IC technology has emerged as a promising solution for the former. Regarding the latter, we propose the concept of distributed batteries within a heterogeneous 3D IC. This approach involves utilizing multiple smaller batteries with different specifications among different modules of 3D ICs. This approach optimizes performance and overcomes limitations associated with both 3D ICs and conventional power delivery methods. Distributed batteries play a vital role in effectively managing the heat generated by energy sources and modules within a 3D IC. Furthermore, they contribute to achieving a uniform distribution of heat throughout the entire structure, which ultimately ensures the optimal performance of the batteries and modules. The simulation results indicate a 40 percent enhancement in achieving a more even distribution of generated heat. Additionally, the proposed distributed battery techniques improve power delivery, enhance reliability, and enable optimized voltage regulation while improving efficiency. In addition to the primary benefits, alternative configurations of the proposed approach can offer extra energy storage capacity and act as efficient electromagnetic shields, resulting in an impressive reduction of external electromagnetic noises by 60 dB.

47 OTHER INSTRUMENTATION↗
DOE OSTI2021

Identification and Testing of Electric Vehicle Fast Charger Cybersecurity Mitigations

Fast-charging infrastructure for electric vehicles (EVs) is needed to enable and achieve the national goals of transitioning the vehicle fleet toward more electrification. Idaho National Laboratory, Oak Ridge National Laboratory, and the National Renewable Energy Laboratory (NREL) have jointly worked to identify, evaluate, and mitigate potential cyber-related consequences associated with fast charger systems. NREL contributed by considering cyberattack scenarios and consequences associated with integrating distributed energy resources (DERs) at fast-charging stations. The dynamic nature of fast-charger load profiles would encourage site operators to incorporate solar for energy cost reduction and energy storage for peak demand cost management at future charging facilities with multiple fast chargers at a site. These energy resources would be monitored and coordinated via a site energy management controller with data exchange between devices and local power metering infrastructure; thus, networking between devices and the design of the system becomes important in the overall cybersecurity posture. In addition, component vendors and system operators might have remote interfaces to any of these systems. It is therefore important to understand the breadth of the cyberattack surface and potential strategies to mitigate impacts. This project has focused on components and protocols expected to be found within a local charging site that includes multiple chargers and DER resources. Our methods and results are summarized in this final report.

42 ENGINEERING↗
DOE OSTI2026

Scaling kinetic Monte-Carlo simulations of grain growth with combined convolutional and graph neural networks

Graph neural networks (GNN) have emerged as a promising machine learning method for microstructure simulations such as grain growth. However, accurate modeling of realistic grain boundary networks requires large simulation cells, which GNN has difficulty scaling up to. To alleviate the computational costs and memory footprint of GNN, we suggest a hybrid architecture combining a convolutional neural network (CNN) based bijective autoencoder to compress the spatial dimensions, and a GNN that evolves the microstructure in the latent space of reduced spatial sizes. Our results demonstrate that the new design significantly reduces computational costs with using fewer message passing layer (from 12 down to 3) compared with GNN alone. The reduction in computational cost becomes more pronounced as the spatial size increases, indicating strong computational scalability. For the largest mesh evaluated (160 3 ), our method reduces memory usage and runtime in inference by 117× and 115×, respectively, compared with GNN-only baseline. More importantly, it shows higher accuracy and stronger spatiotemporal capability than the GNN-only baseline, especially in long-term testing. Such combination of scalability and accuracy is essential for simulating realistic material microstructures over extended time scales. The improvements can be attributed to the bijective autoencoder’s ability to compress information losslessly from spatial domain into a high dimensional feature space, thereby producing more expressive latent features for the GNN to learn from, while also contributing its own spatiotemporal modeling capability. Training data are generated from stochastic grain growth simulations, providing realistic variability for learning robust microstructure evolution. Comprehensive system validation confirms that the model is accurate, robust, and scalable.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Position-specific carbon isotopes of Murchison amino acids elucidate extraterrestrial abiotic organic synthesis networks

The Murchison meteorite is a well-studied carbonaceous chondrite with relatively high concentrations of amino acids thought to be endogenous to the meteorite, in part because they are characterized by carbon isotope (δ 13 C) values higher than those typical of terrestrial amino acids. Past studies have proposed that extraterrestrial amino acids in the Murchison meteorite could have formed by Strecker synthesis (for α-amino acids), Michael addition (for β-amino acids), or reductive amination, but a lack of constraints have prevented confident discrimination among these possibilities, or assignment of specific formation pathways to each of several specific amino acids. Position-specific carbon isotope analysis differentiates amongst these mechanisms by relating molecular sites to isotopically distinct carbon sources and by constraining isotope effects associated with elementary chemical reactions. Prior measurements of the position-specific carbon isotopic composition of α-alanine from the Murchison CM chondrite demonstrated that alanine’s high δ 13 C VPDB value is attributable to the amine carbon (δ 13 C VPDB = +142 ± 20‰), consistent with Strecker synthesis drawing on 13 C-rich carbonyl groups in precursors (Chimiak et al., 2021). Here, we measured the δ 13 C composition of fragment ions generated by electron impact ionization of derivatized α-alanine, β-alanine, and aspartic acid from Murchison via gas chromatography-Fourier transform mass spectrometry. α-Alanine’s amine carbon yielded δ 13 C VPDB = +109 ± 21‰, which is consistent with the previously measured value and with formation from 13 C-rich precursors. β-Alanine’s amine carbon presents a lower δ 13 C VPDB = +33 ± 24‰, which supports formation from 13 C-rich precursors but potentially via a Michael addition mechanism rather than Strecker synthesis. Aspartic acid’s amine carbon has δ 13 C VPDB = -14 ± 5‰, suggesting synthesis from precursors distinct from those that generated the alanine isomers. Further, these measurements indicate that Murchison amino acids are a mixture of compounds made from different synthesis mechanisms, though some subsets likely drew on the same substrates; this conclusion highlights the complexity of extraterrestrial organic synthesis networks and the potential of emerging methods of isotope ratio analysis to elucidate the details of those networks.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Sustainable, degradable and malleable low-dielectric-constant thermosets derived from biomass for recyclable green electronics

In view of their excellent thermal and chemical resistance, favorable adhesion properties, high tensile strength, and other advantages, epoxy thermosets have numerous industrial applications such as coating, adhesive, and composite material production. However, bisphenol A (BPA)-based resins, which account for a significant fraction of the above thermosets, adversely affect human health by interfering with the normal functioning of the endocrine system and cannot be easily decomposed and recycled. Herein, degradable and sustainable bio-based epoxy thermosets with crosslinked network structures are prepared by the epoxidation of isosorbide with epichlorohydrin and the curing of the produced isosorbide diglycidyl ether (ISDGE) with cyclic lactones through a cationic ring-opening reaction. The ISDGE thermosets fully decompose into soluble and recyclable products under mildly basic conditions within three days, which is ascribed to the presence of ester moieties within the polymer network structure. Compared to a representative BPA-based epoxy, ISDGE-based thermosets exhibit lower dielectric constants and are more flexible. Moreover, the glass fibers in ISDGE-based prepregs can be fully recovered after on-demand degradation. Furthermore, our work provides promising eco-friendly alternatives to conventional epoxy thermosets and paves the way for the reduction of plastic waste generation and the development of recyclable green electronics.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRS2018

Decadal Changes in Summertime Reactive Oxidized Nitrogen and Surface Ozone over the Southeast United States

Widespread efforts to abate ozone (O3) smog have significantly reduced emissions of nitrogen oxides (NOx) over the past 2 decades in the Southeast US, a place heavily influenced by both anthropogenic and biogenic emissions. How reactive nitrogen speciation responds to the reduction in NOx emissions in this region remains to be elucidated. Here we exploit aircraft measurements from ICARTT (International Consortium for Atmospheric Research on Transport and Transformation - July-August 2004), SENEX (Southeast Nexus - June-July 2013), and SEAC4RS (Studies of Emissions and Atmospheric Composition, Clouds and Climate Coupling by Regional Surveys - August-September 2013) and long-term ground measurement networks alongside a global chemistry-climate model to examine decadal changes in summertime reactive oxidized nitrogen (RON) and ozone over the Southeast US. We show that our model can reproduce the mean vertical profiles of major RON species and the total (NO (sub y)) in both 2004 and 2013. Among the major RON species, nitric acid (HNO3) is dominant (approximately 42-45 percent), followed by NOx (31 percent), total peroxy nitrates (Sigma PNs; 14 percent), and total alkyl nitrates (Sigma ANs; 9-12 percent) on a regional scale. We find that most RON species, including NOx, Sigma PNs, and HNO3, decline proportionally with decreasing NOx emissions in this region, leading to a similar decline in NO (sub y). This linear response might be in part due to the nearly constant summertime supply of biogenic VOC (Volatile Organic Compounds) emissions in this region. Our model captures the observed relative change in RON and surface ozone from 2004 to 2013. Model sensitivity tests indicate that further reductions of NOx emissions will lead to a continued decline in surface ozone and less frequent high-ozone events.

Li, Jingyi↗
DOE OSTI2022

A 1D–0D–3D coupled model for simulating blood flow and transport processes in breast tissue

In this work, we present mixed dimensional models for simulating blood flow and transport processes in breast tissue and the vascular tree supplying it. These processes are considered, to start from the aortic inlet to the capillaries and tissue of the breast. Large variations in biophysical properties and flow conditions exist in this system necessitating the use of different flow models for different geometries and flow regimes. In total, we consider four different model types. First, a system of 1D nonlinear hyperbolic partial differential equations (PDEs) is considered to simulate blood flow in larger arteries with highly elastic vessel walls. Second, we assign 1D linearized hyperbolic PDEs to model the smaller arteries with stiffer vessel walls. The third model type consists of ODE systems (0D models). It is used to model the arterioles and peripheral circulation. Finally, homogenized 3D porous media models are considered to simulate flow and transport in capillaries and tissue within the breast volume. Sink terms are used to account for the influence of the venous and lymphatic systems. Combining the four model types, we obtain two different 1D–0D–3D coupled models for simulating blood flow and transport processes: The first model results in a fully coupled 1D–0D–3D model covering the complete path from the aorta to the breast combining a generic arterial network with a patient specific breast network and geometry. The second model is a reduced one based on the separation of the generic and patient specific parts. The information from a calibrated fully coupled model is used as inflow condition for the patient specific sub-model allowing a significant computational cost reduction. Several numerical experiments are conducted to calibrate the generic model parameters and to demonstrate realistic flow simulations compared to existing data on blood flow in the human breast and vascular system. Moreover, we use two different breast vasculature and tissue data sets to illustrate the robustness of our reduced sub-model approach.

multiscale blood flow models↗
NASA NTRS2023

Considering Soil Moisture in Models of Climate Impacts on Child Health in Farming-Centric Countries

Soil moisture reflects the amount of water available to crops in the top layer of soil. As such, considering soil moisture provides important insight into water availability and ultimately crop yields in agricultural settings. In studies of climate change, food security, and health, however, soil moisture is rarely empirically considered despite its connection to crop health and yields. In this project, we aim to advance understanding of climate impacts on food security by incorporating soil moisture into quantitative models of child health. Combining spatially referenced health survey data from the Demographic and Health Surveys for 2005 and 2010 in Senegal and 2007, 2011, and 2014 in Bangladesh, with soil moisture data from the Famine Early Warning System Network Land Data Assimilation System, we explore the linkages between sub-annual and sub-seasonal climate conditions and child malnutrition in two rainfed agriculture dependent countries—Bangladesh and Senegal. Results suggest that soil moisture, measured on very short time scales, may be associated with reductions in anthropometric weight-for-height z-scores, but the relationship is highly dependent upon geographic context.

Child health↗
DOE OSTI2017

Fault diagnosis and severity estimation in nuclear power plants - 196

Signed Directed Graph (SDG) is a type of fault diagnosis method with a capability to display the complex relationship between plant parameters and reveal the potential danger and the propagation path when a fault occurs in a nuclear plant. However, the fault diagnosis method based on qualitative SDG model inevitably has a number of disadvantages, and some of these disadvantages include its inability to identify reoccurring faults; ambiguous reasoning that can result in low resolution; performance that relies on the accuracy of the model; slow diagnosis rate for large systems and poor real-time diagnosis, among others. In this paper, we propose an SDG fault diagnosis methodology based on Granular Computing Theory, referred to as Granular Computing Signed Directed Graph (GCSDG). This paper shows how knowledge reduction nature of granular computing based fault diagnosis can simplify the decision table, reduce the allocation of resources, scale-down the complexity of the problem and improve the efficiency of data processing. Then, we estimate the severity of the fault using Back Propagation (BP) neural network approach. The proposed methodology was verified by considering certain faults in the primary loop of a PWR, using PCTRAN. The result shows a considerable improvement in fault diagnosis and the severity estimated serves as a useful information for the operator, providing a firm foundation for further intervention. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2024

Using spatio-temporal graph neural networks to estimate fleet-wide photovoltaic performance degradation patterns

Accurate estimation of photovoltaic (PV) system performance is crucial for determining its feasibility as a power generation technology and financial asset. PV-based energy solutions offer a viable alternative to traditional energy resources due to their superior Levelized Cost of Energy (LCOE). A significant challenge in assessing the LCOE of PV systems lies in understanding the Performance Loss Rate (PLR) for large fleets of PV systems. Estimating the PLR of PV systems becomes increasingly important in the rapidly growing PV industry. Precise PLR estimation benefits PV users by providing real-time monitoring of PV module performance, while explainable PLR estimation assists PV manufacturers in studying and enhancing the performance of their products. However, traditional PLR estimation methods based on statistical models have notable drawbacks. Firstly, they require user knowledge and decision-making. Secondly, they fail to leverage spatial coherence for fleet-level analysis. Additionally, these methods inherently assume the linearity of degradation, which is not representative of real world degradation. To overcome these challenges, we propose a novel graph deep learning-based decomposition method called the Spatio-Temporal Graph Neural Network for fleet-level PLR estimation (PV-stGNN-PLR). PV-stGNN-PLR decomposes the power timeseries data into aging and fluctuation components, utilizing the aging component to estimate PLR. PV-stGNN-PLR exploits spatial and temporal coherence to derive PLR estimation for all systems in a fleet and imposes flatness and smoothness regularization in loss function to ensure the successful disentanglement between aging and fluctuation. We have evaluated PV-stGNN-PLR on three simulated PV datasets consisting of 100 inverters from 5 sites. Experimental results show that PV-stGNN-PLR obtains a reduction of 33.9% and 35.1% on average in Mean Absolute Percent Error (MAPE) and Euclidean Distance (ED) in PLR degradation pattern estimation compared to the state-of-the-art PLR estimation methods.

14 SOLAR ENERGY↗
DOE OSTI2023

Confidentiality-preserving machine learning algorithms for soft-failure detection in optical communication networks

Automated fault management is at the forefront of next-generation optical communication networks. The increase in complexity of modern networks has triggered the need for programmable and software-driven architectures to support the operation of agile and self-managed systems. In these scenarios, the European Telecommunications Standards Institute zero-touch network and service management approach is imperative. The need for machine learning algorithms to process the large volume of telemetry data brings safety concerns as distributed cloud-computing solutions become the preferred approach for deploying reliable communication network automation. This paper’s contribution is twofold. First, we propose a simple yet effective method to guarantee the confidentiality of the telemetry data based on feature scrambling. The method allows the operation of third-party computational services without direct access to the full content of the collected data. Additionally, the effectiveness of four unsupervised machine learning algorithms for soft-failure detection is evaluated when applied to the scrambled telemetry data. The methods are based on factor analysis, principal component analysis, nonlinear principal component analysis, and singular value decomposition. Most dimensionality reduction algorithms have the common property that they can maintain similar levels of fault classification performance while hiding the data structure from unauthorized access. Evaluations of the proposed algorithms demonstrate this capability.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Atomistic Insights into Nitrogen-Cycle Electrochemistry: A Combined DFT and Kinetic Monte Carlo Analysis of NO Electrochemical Reduction on Pt(100)

Electrocatalytic denitrification is a promising technology for the removal of NOx species in groundwater. However, a lack of understanding of the molecular pathways that control the overpotential and product distribution have limited the development of practical electrocatalysts, and additional atomic-level insights are needed to advance this field. Adsorbed NO has been identified as a key intermediate in the NOx electroreduction network, and the elementary steps by which it decomposes to NH4+, N2, NH3OH+, or N2O remain a subject of debate. Herein, we report a combined density functional theory (DFT) and kinetic Monte Carlo (kMC) study of this reaction on Pt(100), a catalytic surface that is known to be suitable for the activation of strong covalent bonds, in acidic electrolytes. This approach describes the effects of coverage-dependent adsorbate–adsorbate interactions, water-mediated protonation kinetics and thermodynamics, and transient potential sweeps, on reaction rates and selectivities. The results predict NO stripping curves in excellent agreement with experiments while, at the same time, providing a mechanistic interpretation of observed current peaks. Furthermore, production of NH4+ products is traced to the rapid kinetics of N–O bond breaking in reactive intermediates, whereas rapid hydrogenation of surface N* species prevent competing pathways from forming either N2 or N2O. The combined DFT-kMC methodology thus provides a unique tool to describe the mechanism and energetics of platinum-catalyzed electroreduction in the nitrogen cycle, and this approach should also find application to related electrocatalytic processes that are of technological and environmental interest.

Chun, Hee-Joon↗