SEARCH · Engineering Papers
Results for “Networking”
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.
Geometric sensitivity of electrochemical fin shape on three dimensional microstructure network conductivity analysis
Not Available
Competing endogenous RNA network associated with oxygen-induced retinopathy: Expression of the network and identification of the MALAT1/miR-124–3p/EGR1 regulatory axis
Retinopathy of prematurity (ROP) is a severe retinal dysfunction in prematurely born babies. The relationship between non-coding RNAs and retinopathy of prematurity (ROP) remain unclear. Microarray analysis of lncRNAs, miRNAs, and mRNAs was conducted in a mouse model of ROP. A competing endogenous RNA (ceRNA) network was constructed. The relationship among MALAT1, miR-124–3p, and Early growth response protein 1 (EGR1) was assessed in hypoxia-induced primary human umbilical vein endothelial cells (HUVECs) and ROP mouse model. In the study, we found 2252 lncRNAs, 1239 mRNAs, and 36 miRNAs were differentially regulated. ceRNA network consisting of 21 lncRNAs, 10 miRNAs, and 19 mRNAs was established. Of the most down-regulated miRNAs, miR-124–3p was selected for additional study. miR-124–3p ceased the migration and proliferation of primary HUVECs in hypoxic conditions, and directly suppressed EGR1. Additionally, MALAT1 directly sponged miR-124–3p. Knockdown of MALAT1 decreased EGR1 expression and inhibited the migration and proliferation of primary HUVECs in hypoxia. Furthermore, these changes were rescued by depletion of miR-124–3p. In vivo, intravitreal injection of miR-124–3p, shMALAT1 decreased EGR1 expression and markedly suppressed retinal neovascularization in OIR models. Intravitreal injection of shMALAT1 and miR-124–3p antagomir at the same time can promote retinal neovascularization, which reversed the suppression of retinal neovascularization functioned by shMALAT1. In conclusion, the expression profiles of lncRNAs and miRNAs and the ceRNA network in a mouse model of ROP may be indicative of the underlying mechanisms of retinal angiogenesis and neural activity. The MALAT1/miR-124–3p/EGR1 regulatory axis is partly responsible for retinal neovascularization, which may provide a novel theoretical basis for the pathogenesis of ROP.
Versatile Synthetic Platform for Polymer Membrane Libraries Using Functional Networks
Poly(ethylene glycol) (PEG) diacrylate copolymer networks containing pentafluorophenyl active esters can be quantitatively substituted with a wide variety of primary amines, enabling the development of a versatile synthetic platform for the preparation of polymer hydrogel and membrane libraries. Here, by tuning the starting network, a high degree of control over cross-linking density, water uptake, and functional group incorporation can be reproducibly achieved, which is vital for elucidating structure–property relationships in ion transporting membranes. From the same cross-linked scaffold, a diversity of basic, acidic, and solute-chelating moieties were obtained through functionalization, which allows for tailored uptake of basic and acidic organic dyes and metal chloride salts. Ion permeation and sorption measurements for a series of polymer networks with controlled cross-linking density and varied imidazole grafting densities illustrate the ability of this platform to isolate the effect of chemical functionality on ion transport from the effects of cross-linking density and water content.
Anisotropic and Interconnected Nanoporous Materials from Randomly End-Linked Copolymer Networks
Microphase separation within randomly end-linked copolymer networks (RECNs) provides access to disordered bicontinuous morphologies over a wide composition range of the constituent network strands. Here, we rely on end-linking of telechelic hydroxyl-terminated polystyrene (PS) and poly(D,Llactide) (PLA) chains of equal molecular weight, with a tetrafunctional isocyanate cross-linker in a good solvent for both strands, followed by solvent removal to induce microphase separation, and finally etching of the PLA phase to yield nanoporous materials. Transmission electron microscopy (TEM) tomographic reconstructed 3D images along with gravimetric measurements and small-angle X-ray scattering (SAXS) indicate the formation of highly interconnected structures over a range of ~40-70 vol % of PLA, while N2 adsorption measurements indicate narrowly distributed pore sizes that can be tuned by varying the strand molecular weights. Stretching of the PS/PLA copolymer networks above the glass transition temperatures of both components prior to etching the PLA phase provides a straightforward means to introduce controlled anisotropy into the 3D interconnected porous materials.
Energy-efficient Mott activation neuron for full-hardware implementation of neural networks
To circumvent the von Neumann bottleneck, substantial progress has been made towards in-memory computing with synaptic devices. However, compact nanodevices implementing non-linear activation functions are required for efficient full-hardware implementation of deep neural networks. Here, in this work, we present an energy-efficient and compact Mott activation neuron based on vanadium dioxide and its successful integration with a conductive bridge random access memory (CBRAM) crossbar array in hardware. The Mott activation neuron implements the rectified linear unit function in the analogue domain. The neuron devices consume substantially less energy and occupy two orders of magnitude smaller area than those of analogue complementary metal–oxide semiconductor implementations. The LeNet-5 network with Mott activation neurons achieves 98.38% accuracy on the MNIST dataset, close to the ideal software accuracy. We perform large-scale image edge detection using the Mott activation neurons integrated with a CBRAM crossbar array. Our findings provide a solution towards large-scale, highly parallel and energy-efficient in-memory computing systems for neural networks.
Citation network datasets for benchmarking spiking graph neural networks on experimental neuromorphic hardware
Spiking neural networks (SNNs) running on neuromorphic computers offer an energy-efficient alternative for AI tasks. Recently, spiking graph neural networks (S-GNNs) have been shown to produce encouraging results on benchmark citation network datasets such as Cora, CiteSeer, and PubMed for node classification tasks. These S-GNNs were run on SNN simulators only because they contain up to tens of thousands of neurons and up to millions of synapses, translating poorly to neuromorphic hardware. Therefore, in this paper, we create a suite of benchmark datasets from the CiteSeer dataset that can be accommodated on current neuromorphic hardware platforms. Our contribution consists of a collection of three datasets. First, we have an induced subgraph of CiteSeer, which we call MiniSeer, containing 2110 papers, 3604 binary features, and 6 topics. Second, MicroSeer is a very small dataset consisting of 84 papers, 1227 features, and 6 topics. Lastly, BiteSeer is a collection of 15 binary classification datasets. We present creation of these datasets along with accuracies, running times, and spike counts when simulated. We believe that our results in this paper will be used by the neuromorphic community to benchmark, test, and develop neuromorphic hardware and simulators.
Hydrogen adsorption and desorption with 3D silicon nanotube-network and film-network structures: Monte Carlo simulations
Hydrogen is clean, sustainable, and renewable, thus is viewed as promising energy carrier. However, its industrial utilization is greatly hampered by the lack of effective hydrogen storage and release method. Carbon nanotubes (CNTs) were viewed as one of the potential hydrogen containers, but it has been proved that pure CNTs cannot attain the desired target capacity of hydrogen storage. In this paper, we present a numerical study on the material-driven and structure-driven hydrogen adsorption of 3D silicon networks and propose a deformation-driven hydrogen desorption approach based on molecular simulations. Two types of 3D nanostructures, silicon nanotube-network (Si-NN) and silicon film-network (Si-FN), are first investigated in terms of hydrogen adsorption and desorption capacity with grand canonical Monte Carlo simulations. It is revealed that the hydrogen storage capacity is determined by the lithium doping ratio and geometrical parameters, and the maximum hydrogen uptake can be achieved by a 3D nanostructure with optimal configuration and doping ratio obtained through design optimization technique. For hydrogen desorption, a mechanical-deformation-driven-hydrogen-release approach is proposed. Compared with temperature/pressure change-induced hydrogen desorption method, the proposed approach is so effective that nearly complete hydrogen desorption can be achieved by Si-FN nanostructures under sufficient compression but without structural failure observed. The approach is also reversible since the mechanical deformation in Si-FN nanostructures can be elastically recovered, which suggests a good reusability. This study may shed light on the mechanism of hydrogen adsorption and desorption and thus provide useful guidance toward engineering design of microstructural hydrogen (or other gas) adsorption materials.
Direct evidence that an extended hydrogen-bonding network influences activation of pyridoxal 5'-phosphate in aspartate aminotransferase
Abstract not provided
DeepGraphONet: A Deep Graph Operator Network to Learn and Zero-Shot Transfer the Dynamic Response of Networked Systems
This article develops a deep graph operator network (DeepGraphONet) framework that learns to approximate the dynamics of a complex system (e.g., the power grid or traffic) with an underlying subgraph structure. Here, we build our DeepGraphONet by fusing the ability of graph neural networks to exploit spatially correlated graph information and deep operator networks to approximate the solution operator of dynamical systems. The resulting DeepGraphONet can then predict the dynamics within a given short/medium-term time horizon by observing a finite history of the graph state information. Furthermore, we design our DeepGraphONet to be resolution independent. That is, we do not require the finite history to be collected at the exact/same resolution. In addition, to disseminate the results from a trained DeepGraphONet, we design a zero-shot learning strategy that enables using it on a different subgraph. Finally, empirical results on the transient stability prediction problem of power grids and traffic flow forecasting problem of a vehicular system illustrate the effectiveness of the proposed DeepGraphONet.
Topological Power Grid Statistics from a Network-of-Networks Perspective
In this work we provide novel tools to evaluate the ``realism'' of a power system topology based combining ideas from the study of complex networks with electrical system parameters. Specifically, we show that the recent network-of-networks approach of Halappanavar, et al.\ can be used to gain insight into reasonable ranges for topological properties of a power system such as, diameter, average shortest path length, core size, and clustering.
A Fast and Scalable Genetic Algorithm-Based Approach for Planning of Microgrids in Distribution Networks
As a result of climate change, extreme weather events are occurring more frequently and with increasing impact. This trend poses a significant challenge for distribution utilities and system operators to ensure that there is uninterrupted power supply to critical loads in their networks; thus, the level of proactive preparation of the distribution system to be able to handle severe impacts of extreme weather events represents the system's resilience. One method that distribution system planners can use to prepare for future extreme events is to plan multiple microgrids which can use local generation as much as possible to supply critical loads. But partitioning an existing distribution system such that multiple feasible islands are planned and which are capable of supporting critical loads is still challenging for distribution systems - first, because of the size of the network graph partitioning problem and, second, because of the difficulty in properly formulating the desired attributes of such islands or microgrids. Therefore, this paper presents a genetic algorithm based approach that facilitates incorporating multiple objectives for grid partitioning by formulating two types of problems - node allocation and edge elimination - and it considers multiple topological and resilience-enhancing objectives. The performance of the proposed genetic algorithm-based approach is numerically evaluated on multiple test systems as well as on a real distribution feeder in Colorado, United States.
Physics-Informed Generative Adversarial Networks for Stochastic Differential Equations
We developed a new class of physics-informed generative adversarial networks (PI-GANs) to solve forward, inverse, and mixed stochastic problems in a unified manner based on a limited number of scattered measurements. Unlike standard GANs relying solely on data for training, here we encoded into the architecture of GANs the governing physical laws in the form of stochastic differential equations (SDEs) using automatic differentiation. In particular, we applied Wasserstein GANs with gradient penalty (WGAN-GP) for its enhanced stability compared to vanilla GANs. We first tested WGAN-GP in approximating Gaussian processes of different correlation lengths based on data realizations collected from simultaneous reads at sparsely placed sensors. We obtained good approximation of the generated stochastic processes to the target ones even if there is a mismatch between the input noise dimensionality and the effective dimensionality of the target stochastic processes. We also studied the overfitting issue for both the discriminator and the generator, and we found that overfitting occurs also in the generator in addition to the discriminator as previously reported. Subsequently, we considered the solution of elliptic SDEs requiring approximations of three stochastic processes, namely the solution, the forcing, and the diffusion coefficient. Here again, we assumed data collected from simultaneous reads at a limited number of sensors for the multiple stochastic processes. Furthermore, three generators were used for the PI-GANs: two of them were feed forward deep neural networks (DNNs), while the other one was the neural network induced by the SDE. For the case where we have one group of data, we employed one feed forward DNN as the discriminator, while for the case of multiple groups of data we employed multiple discriminators in PI-GANs. We solved forward, inverse, and mixed problems without changing the framework of PI-GANs, obtaining both the means and the standard deviations of the stochastic solution and the diffusion coefficient in good agreement with benchmarks. In this work, we have demonstrated the effectiveness of PI-GANs in solving SDEs for about 120 dimensions. In principle, PI-GANs could tackle very high dimensional problems given more sensor data with low-polynomial growth in computational cost.
Software-Defined Network Solutions for Science Scenarios: Performance Testing Framework and Measurements
Explore the source record for details and available documents.
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.
Hardware-in-the-loop Testing of Network Protectors for Low-Voltage Networks with Distributed Energy Resources
In this report, we developed and validated a network protector relay digital twin model and interfaced a commonly used network protector relay hardware with our real-time simulation system. Hardware-in-the-loop protection studies are performed to assess the impact of distributed energy resources (DER) and benchmark a rate-of-change-based mitigation strategy. Simulation results suggest that the network protector reverse trip and auto-reclose functions are negatively impacted by the high distributed energy resource penetration. To accommodate DER backfeed while remaining secure and reliable for faults on primary feeders, we recommend options for a rate-of-change-based blocking scheme and a protection setting change. Finally, future mitigation ideas and standard revisions are discussed.