Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “network dynamics”

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 1,567 records · Page 87
All sourcesNASA NTRSDOE OSTI
NASA NTRS2017

A 3D Active Learning Application for NeMO-Net, the NASA Neural Multi-Modal Observation and Training Network for Global Coral Reef Assessment

NeMO-Net, the NASA neural multi-modal observation and training network for global coral reef assessment, is an open-source deep convolutional neural network and interactive active learning training software aiming to accurately assess the present and past dynamics of coral reef ecosystems through determination of percent living cover and morphology as well as mapping of spatial distribution. We present an interactive video game prototype for tablet and mobile devices where users interactively label morphology classifications over mm-scale 3D coral reef imagery captured using fluid lensing to create a dataset that will be used to train NeMO-Nets convolutional neural network. The application currently allows for users to classify preselected regions of coral in the Pacific and will be expanded to include additional regions captured using our NASA FluidCam instrument, presently the highest-resolution remote sensing benthic imaging technology capable of removing ocean wave distortion, as well as lower-resolution airborne remote sensing data from the ongoing NASA CORAL campaign. Active learning applications present a novel methodology for efficiently training large-scale Neural Networks wherein variances in identification can be rapidly mitigated against control data. NeMO-Net periodically checks users input against pre-classified coral imagery to gauge their accuracy and utilize in-game mechanics to provide classification training. Users actively communicate with a server and are requested to classify areas of coral for which other users had conflicting classifications and contribute their input to a larger database for ranking. In partnering with Mission Blue and IUCN, NeMO-Net leverages an international consortium of subject matter experts to classify areas of confusion identified by NeMO-Net and generate additional labels crucial for identifying decision boundary locations in coral reef assessment.

NeMO=NET↗
DOE OSTI2021

Impact of DER Communication Delay in AGC: Cyber-Physical Dynamic Co-Simulation: Preprint

Distributed energy resource (DER) frequency regulations are promising technologies for future grid operation. Unlike conventional generators, DERs might require open communication networks to exchange signals with control centers, possibly through DER aggregators; therefore, the impacts of the communication variations including the latency on the system frequency stability need to be investigated. This paper develops a cyber-physical dynamic simulation (CPDS) model based on the Hierarchical Engine for Large-Scale Co-Simulation (HELICS) to evaluate the impact of the communication variations, such as delays in DER frequency regulations. The feasible delay range can be obtained under different AGC parameter settings. The results show that the risk of instability generally increases with the communication delay and that the difference of AGC provided from DER and conventional generation units should be recognized when designing frequency regulation controllers.

61 RADIATION PROTECTION AND DOSIMETRY↗
DOE OSTI2022

Effects of pore connectivity and tortuosity on the dynamics of fluids confined in sub-nanometer pores

Dynamical behavior of fluids under nano-pore confinement is studied extensively as it has important implications for several industrial as well as geological processes. Pore network in many porous materials exhibits a varied degree of inter connections. The extent of this pore connectivity may affect the structural and dynamical behavior of the confined fluid. However, studies of fluid confinement addressing these effects systematically are lacking. In this paper we report molecular dynamics simulation studies addressing the effects of pore connectivity on the dynamics of two representative fluids – CO 2 and ethane in silicalite by systematically varying the degree of pore connectivity through selectively blocking some pore space with immobile methane molecules. By selectively turning off the pore spaces in the shape of straight, or tortuous zigzag channels, we also probe the effects of pore tortuosity. In general, pore connectivity is found to facilitate both the translational as well as rotational dynamics of both fluids, while the intermolecular modes of vibration in both fluids remain largely unaffected. The effects of providing connections between a set of straight or zigzag channel-like pores are however more nuanced. Pore tortuosity facilitates the rotational motion, but suppresses the translational motion of CO 2 , while its effects on the rotational and translational motion of ethane are less pronounced. The intermolecular vibrational modes of both fluids shift to higher energies with an increase in the number of tortuous pores. The results reported here provide a detailed molecular level understanding of the effects of pore connectivity on the dynamics of fluids and thus have implications for applications like fluid separation.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

Nonlocal Effect of Percolated Particle Networks on Viscoelasticity of Polymer–Filler Nanocomposites: A Mesoscale Simulation Study

With nanoparticles (NPs) as fillers, polymer nanocomposites (PNCs) usually exhibit enhanced mechanical properties. However, a direct connection between the microscopic structural relaxation and macroscopic mechanical properties of PNCs remains to be established. To investigate the micro-to-macro connection, we develop a mesoscale model, in which the NPbridging polymer chains are represented by a dynamic bonded interaction between NPs, and the bulk polymer matrix is implicitly modeled by overdamped Langevin dynamics. Extensive equilibrium simulations are performed to quantify the microscopic dynamics of model PNCs. Systematic analyses of modified Rouse dynamics, dynamic structure factor, and relaxation modulus uncover that the microscopic relaxation dynamics of PNCs are significantly decelerated across different length scales because of nonlocal effects of percolated particle networks a phenomenon that has not been adequately captured in prior simulation studies. We find that NPvolume- fraction and NP-bonding-energy barrier are the two critical variables that affect bulk viscoelasticity the most. The proposed mesoscale model is versatile and provides a powerful framework for studying structure−property relations of different PNCs.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Physics Informed Reinforcement Learning for Power Grid Control using Augmented Random Search

Wide adoption of deep reinforcement learning need to overcome several challenges in energy system domain, including scalability, learning from limited samples, and high-dimensional continuous state and action spaces. In this paper, we integrated physics-based information from the normal generator operation state formula in the reinforcement learning agent's neural network loss function, and applied an augmented random search agent to optimize the generator control under dynamic contingency. Simulation results demonstrated the reliability performance improvements in training speed, reward convergence, sampling efficiency, scalability, and transferability.

physics informed ML, Physics Informed Neural Netwo↗
DOE OSTI2020

Automated Realistic Testbed Synthesis for Power System Communication Networks based on Graph Metrics

Communication networks are integral to modern power grid operations and are becoming increasingly critical as grid dynamics speed up and as more controls become closed-loop in form. Focusing on the interdependence between the physical grid and the communication system, we identify several key characteristics and typical graph properties based on analysis of a real communication system for a power grid. Moreover, an automated process was provided for synthetic testbed in the NS-3 simulator for the power system test case, and its network characteristics have been further derived for power system monitoring and control applications

Graph Analysis, Synthetic Testbed, communication n↗
DOE OSTI2020

Layer-Parallel Training of Deep Residual Neural Networks

Residual neural networks (ResNets) are a promising class of deep neural networks that have shown excellent performance for a number of learning tasks, e.g., image classification and recognition. Mathematically, ResNet architectures can be interpreted as forward Euler discretizations of a nonlinear initial value problem whose time-dependent control variables represent the weights of the neural network. Hence, training a ResNet can be cast as an optimal control problem of the associated dynamical system. For similar time-dependent optimal control problems arising in engineering applications, parallel-in-time methods have shown notable improvements in scalability. This paper demonstrates the use of those techniques for efficient and effective training of ResNets. The proposed algorithms replace the classical (sequential) forward and backward propagation through the network layers with a parallel nonlinear multigrid iteration applied to the layer domain. This adds a new dimension of parallelism across layers that is attractive when training very deep networks. From this basic idea, we derive multiple layer-parallel methods. The most efficient version employs a simultaneous optimization approach where updates to the network parameters are based on inexact gradient information in order to speed up the training process. Finally, using numerical examples from supervised classification, we demonstrate that the new approach achieves a training performance similar to that of traditional methods, but enables layer-parallelism and thus provides speedup over layer-serial methods through greater concurrency.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Efficient Training of Deep Neural Operator Networks via Randomized Sampling

Neural operators (NOs) employ deep neural networks to learn the mappings between infinitedimensional function spaces. Deep operator network (DeepONet), a popular NO architecture, has demonstrated success in the real-time prediction of complex dynamics across various scientific and engineering applications. In this work, we introduce a random sampling technique to be adopted during the training of DeepONet, aimed at improving the generalization ability of the model, while significantly reducing the computational time. The proposed approach targets the trunk network of the DeepONet model that outputs the basis functions corresponding to the spatiotemporal locations of the bounded domain on which the physical system is defined. While constructing the loss function, DeepONet training traditionally considers a uniform grid of spatiotemporal points at which all the output functions are evaluated for each iteration. This approach leads to a larger batch size, resulting in poor generalization and increased memory demands, due to the limitations of the stochastic gradient descent (SGD) optimizer. The proposed random sampling over the inputs of the trunk net mitigates these challenges, improving generalization and reducing the memory requirements during training, resulting in significant computational gains. We validate our hypothesis through three benchmark examples, demonstrating substantial reductions in training time while achieving comparable or lower overall test errors relative to the traditional training approach. Our results indicate that incorporating randomization in the trunk network inputs during training enhances the efficiency and robustness of DeepONet, offering a promising avenue for improving the framework’s performance in modeling complex physical systems.

Karumuri, Sharmila [Department of Civil & Systems ↗
NASA NTRS1989

Analysis of random drop for gateway congestion control

Lately, the growing demand on the Internet has prompted the need for more effective congestion control policies. Currently No Gateway Policy is used to relieve and signal congestion, which leads to unfair service to the individual users and a degradation of overall network performance. Network simulation was used to illustrate the character of Internet congestion and its causes. A newly proposed gateway congestion control policy, called Random Drop, was considered as a promising solution to the pressing problem. Random Drop relieves resource congestion upon buffer overflow by choosing a random packet from the service queue to be dropped. The random choice should result in a drop distribution proportional to the bandwidth distribution among all contending TCP connections, thus applying the necessary fairness. Nonetheless, the simulation experiments demonstrate several shortcomings with this policy. Because Random Drop is a congestion control policy, which is not applied until congestion has already occurred, it usually results in a high drop rate that hurts too many connections including well-behaved ones. Even though the number of packets dropped is different from one connection to another depending on the buffer utilization upon overflow, the TCP recovery overhead is high enough to neutralize these differences, causing unfair congestion penalties. Besides, the drop distribution itself is an inaccurate representation of the average bandwidth distribution, missing much important information about the bandwidth utilization between buffer overflow events. A modification of Random Drop to do congestion avoidance by applying the policy early was also proposed. Early Random Drop has the advantage of avoiding the high drop rate of buffer overflow. The early application of the policy removes the pressure of congestion relief and allows more accurate signaling of congestion. To be used effectively, algorithms for the dynamic adjustment of the parameters of Early Random Drop to suite the current network load must still be developed.

Hashem, Emam Salaheddin↗
DOE OSTI2022

Deep Reinforcement Learning based Model-free On-line Dynamic Multi-Microgrid Formation to Enhance Resilience

Multi-microgrid formation (MMGF) is a promising solution for enhancing power system resilience. This paper proposes a new deep reinforcement learning (RL) based model-free on-line dynamic MMGF scheme. Additionally, the dynamic MMGF problem is formulated as a Markov decision process, and a complete deep RL framework is specially designed for the topologytransformable micro-grids. In order to reduce the large action space caused by flexible switch operations, a topology transformation method is proposed and an action-decoupling Q-value is applied. Then, a convolutional neural network (CNN) based multi-buffer double deep Q-network (CM-DDQN) is developed to further improve the learning ability of the original DQN method. The proposed deep RL method provides real-time computing to support the on-line dynamic MMGF scheme, and the scheme handles a long-term resilience enhancement problem using an adaptive on-line MMGF to defend changeable conditions. The effectiveness of the proposed method is validated using a 7-bus system and the IEEE 123-bus system. The results show strong learning ability, timely response for varying system conditions and convincing resilience enhancement.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2020

Optimizing Signal Timing Control for Large Urban Traffic Networks Using an Adaptive Linear Quadratic Regulator Control Strategy

We discuss how traffic signal control is important for intersection safety and efficiency. However, most traffic signal control methods are designed for individual intersections or corridors. Although some adaptive control systems have been developed, the methods used are often proprietary and not published, making it difficult to evaluate their effectiveness. This study proposes an adaptive multi-input and multi-output traffic signal control method that not only can improve network-wide traffic operations in terms of reduced traffic delay and energy consumption, but also is more computationally feasible than existing centralized signal control methods. Considering intersection interactions, a linear dynamic traffic system model was built and adaptively updated to reflect how the signal control input of each intersection affects network-wide vehicle travel delay. Based on the system model, an adaptive linear-quadratic regulator (LQR) was designed to minimize both traffic delay and incremental changes in the control input. The proposed control method was evaluated in a microscopic traffic simulation environment with a 35-intersection network of Bellevue City, Washington. Simulation results show that the proposed method had shorter average traffic delays in the network when compared with the traffic delays controlled by the state-of-the-art max-pressure, self-organizing traffic lights, and independent deep Q network methods.

42 ENGINEERING↗
DOE OSTI2024

Exact enforcement of temporal continuity in sequential physics-informed neural networks

The use of deep learning methods in scientific computing represents a potential paradigm shift in engineering problem solving. One of the most prominent developments is Physics-Informed Neural Networks (PINNs), in which neural networks are trained to satisfy partial differential equations (PDEs). While this method shows promise, the standard version has been shown to struggle in accurately predicting the dynamic behavior of time-dependent problems. To address this challenge, methods have been proposed that decompose the time domain into multiple segments, employing a distinct neural network in each segment and directly incorporating continuity between them in the loss function of the minimization problem. In this work we introduce a method to exactly enforce continuity between successive time segments via a solution ansatz. This hard constrained sequential PINN (HCS-PINN) method is simple to implement and eliminates the need for any loss terms associated with temporal continuity. The method is tested for a number of benchmark problems involving both linear and non-linear PDEs. Examples include various first order time dependent problems in which traditional PINNs struggle, namely advection, Allen–Cahn, and Korteweg–de Vries equations. Furthermore, second and third order time-dependent problems are demonstrated via wave and Jerky dynamics examples, respectively. Notably, the Jerky dynamics problem is chaotic, making the problem especially sensitive to temporal accuracy. Finally, the numerical experiments conducted with the proposed method demonstrated superior convergence and accuracy over both traditional PINNs and the soft-constrained counterparts.

42 ENGINEERING↗
DOE OSTI2022

Easy-plane spin Hall nano-oscillators as spiking neurons for neuromorphic computing

Here we show analytically using a macrospin approximation that easy-plane spin Hall nano-oscillators excited by a spin current polarized perpendicularly to the easy plane have phase dynamics analogous to that of Josephson junctions. Similarly to Josephson junctions, they can reproduce the spiking behavior of biological neurons that is appropriate for neuromorphic computing. To take advantage of typical spin-orbit torques, we use a nanoconstriction geometry, in which the magnetostatic interaction and magnetocrystalline anisotropy are tuned to create an easy plane that includes the interface normal direction. We perform micromagnetic simulations of such oscillators realized in this geometry and show that the easy-plane spiking dynamics is preserved in this experimentally feasible architecture. Finally we simulate two elementary neural network blocks that implement operations essential for neuromorphic computing. First, we show that output spikes energies from two neurons can be summed and injected into a following layer neuron and second, we demonstrate that outputs can be multiplied by synaptic weights implemented by locally modifying the anisotropy.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Rapid Cis–Trans Coevolution Driven by a Novel Gene Retroposed from a Eukaryotic Conserved CCR4–NOT Component in Drosophila

Young, or newly evolved, genes arise ubiquitously across the tree of life, and they can rapidly acquire novel functions that influence a diverse array of biological processes. Previous work identified a young regulatory duplicate gene in Drosophila, Zeus that unexpectedly diverged rapidly from its parent, Caf40, an extremely conserved component in the CCR4–NOT machinery in post-transcriptional and post-translational regulation of eukaryotic cells, and took on roles in the male reproductive system. This neofunctionalization was accompanied by differential binding of the Zeus protein to loci throughout the Drosophila melanogaster genome. However, the way in which new DNA-binding proteins acquire and coevolve with their targets in the genome is not understood. Here, by comparing Zeus ChIP-Seq data from D. melanogaster and D. simulans to the ancestral Caf40 binding events from D. yakuba, a species that diverged before the duplication event, we found a dynamic pattern in which Zeus binding rapidly coevolved with a previously unknown DNA motif, which we term Caf40 and Zeus-Associated Motif (CAZAM), under the influence of positive selection. Interestingly, while both copies of Zeus acquired targets at male-biased and testis-specific genes, D. melanogaster and D. simulans proteins have specialized binding on different chromosomes, a pattern echoed in the evolution of the associated motif. Using CRISPR-Cas9-mediated gene knockout of Zeus and RNA-Seq, we found that Zeus regulated the expression of 661 differentially expressed genes (DEGs). Our results suggest that the evolution of young regulatory genes can be coupled to substantial rewiring of the transcriptional networks into which they integrate, even over short evolutionary timescales. Our results thus uncover dynamic genome-wide evolutionary processes associated with new genes.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Exploring urban typologies using comprehensive analysis of transportation dynamics

Abstract As urban areas continue to expand and develop, categorizing cities into typologies offers a valuable framework for understanding metropolitan dynamics and fostering inter-city collaboration. However, existing typologies related to urban mobility have limitations, failing to consider cities within a single large urban region and often overlooking crucial dimensions such as trip demand and traffic flow. In this paper, we introduce a transportation-focused characterization for cities within a large urban region, specifically the San Francisco Bay Area, California. We incorporate over 40 metrics across five transportation dimensions: trip demand, road network, multi-modal network, traffic flow, and land use. Specifically, for the trip demand dimension, we include metrics capturing residents’ trip characteristics, such as mode share, intra-city trips, and inter-city trips. Additionally, we analyze the purpose of trips entering the city to gain a deeper understanding of incoming trip patterns. In the traffic flow dimension, we examine metrics like vehicle miles traveled, delay, and congestion to assess the traffic conditions on the street network. These, combined with other dimensions, provide a comprehensive view of a city’s transportation dynamics. Using unsupervised machine learning clustering methods, we identified eight distinct typologies for the Bay Area: Live Work Cities; Job and Activity Magnet Cities; Anchor Cities; Multi-modal Cities; Hyper-connected Cities; Low-density Residential Cities; Medium-density Residential Cities; and Mixed-use Residential Cities. Our findings show that many clusters are strongly influenced by trip demand and traffic flow metrics. Finally, we examine the practicality of this typology and its potential to guide collaborative transportation management strategies. The typologies provide a foundation for dialogue among Bay Area cities, focusing on evaluating shared characteristics and leveraging successes or challenges to develop unified strategies for transportation management.

Kuncheria, Anu↗
DOE OSTI2022

Toward wide-spectrum antivirals against coronaviruses: Molecular characterization of SARS-CoV-2 NSP13 helicase inhibitors

To date, effective therapeutic treatments that confer strong attenuation against coronaviruses (CoVs) remain elusive. Among potential drug targets, the helicase of CoVs is attractive due to its sequence conservation and indispensability. We rely on atomistic molecular dynamics simulations to explore the structural coordination and dynamics associated with the SARS-CoV-2 Nsp13 apo enzyme, as well as their complexes with natural ligands. A complex communication network is revealed among the five domains of Nsp13, which is differentially activated because of the presence of the ligands, as shown by shear strain analysis, principal components analysis, dynamical cross-correlation matrix analysis, and water transport analysis. The binding free energy and the corresponding mechanism of action are presented for three small molecules that were shown to be efficient inhibitors of the previous SARS-CoV Nsp13 enzyme. Together, our findings provide critical fresh insights for rational design of broad-spectrum antivirals against CoVs.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Effects of input gradient regularization on neural networks time-series forecasting of thermal power systems

This study proposes using neural networks, specifically gated recurrent unit (GRU), long-short-term memory (LSTM), and transformer networks, to improve control strategies in a 450 MW coal-fired power plant. However, neural networks face issues of becoming overly dependent on just a few variables to make predictions, which negatively impacts control decisions that rely on the model to determine the value of all manipulated variables. The paper introduces regularization techniques, including noise injection and input gradient regularization, during the training phase. Here, the work presents novel contributions in adapting neural networks to control industrial systems and applying regularization techniques from computer vision to industrial process control. Results demonstrate the effectiveness of input gradient regularization in reducing model dependence on subsets of variables, emphasizing the balance between fidelity and controllability. Further exploration is recommended, including the development of recurrent transformers, closed-loop control testing, and a sensitivity analysis on computer models to provide further insight.

20 FOSSIL-FUELED POWER PLANTS↗
NASA NTRS2002

Structure and Dynamics of the Quiet Solar Chromosphere

The grant supported research on the structure of the quiet, nonmagnetic chromosphere and on wave excitation and propagation in both the nonmagnetic chromosphere and the magnetic network. The work on the structure of the chromosphere culminated in the recognition that between two competing views of the solar chromosphere, older models by Avrett and collaborators (referred to as VAL) and the newer, dynamical model by Carlsson & Stein (referred to as CS), the clear decision is in favor of the older models, and this in spite of the evident lack of physics, which does not include wave motion and oscillations. The contrast between the static VAL models and the dynamical CS model can be stated most succinctly by comparing the temperature variation implied by the VAL models and the temperature fluctuations of the CS model, which are, respectively, of the order of 10% for the VAL model (at heights where hydrogen is 50% ionized) and a factor of 10 (at the upper boundary of their chromospheric model). The huge fluctuations of the CS model have never been observed, whereas the smaller temperature variations of the VAL models are consistent with ground-based and space-based observations. While it should be obvious which model describes the Sun and which one fails, the case is far from settled in the minds of solar physicists. Thus, much educational work remains to be done and, of course, more research to develop arguments that make the case more convincing. The research on waves and oscillations has been based on a unified theory of excitation of acoustic waves in the field-free atmosphere and of transverse and longitudinal waves in magnetic flux tubes located in the magnetic network by noting, first, that impulsive excitation of all these waves in gravitationally stratified media leads to oscillations at the respective cutoff frequencies and, second, that the observed oscillation frequencies in the nonmagnetic and magnetic parts of the chromosphere match corresponding cutoff frequencies in the upper solar photosphere. The dynamical simulations by Carlsson and Stein have been most instructive and of fundamental importance for understanding wave propagation in a stratified medium by their "flaws", the most important of which is an intensity excess at the H2v emission peak in the H line of Ca II, which is surprising since the observed intensity should have been an upper limit to the simulated intensity. The only plausible explanation for a predicted intensity that is higher than observed is that energy is spread horizontally in upward propagation in the Sun, but not in the plane-wave modeling of CS (as well as by almost everyone else). Investigations of the horizontal size of the region disturbed by the upward-propagating shock in the acoustic-wave propagation implies that the waves in H2v bright-point oscillations emanate from a point source with a diameter corresponding to the width of an intergranular lane, about 100 km, and reach a size of about 4000 km in the upper layers of the chromosphere. Linear, analytic modeling of waves emanating from a point source in a stratified atmosphere shows that the upward-expanding propagation channel does not have sharp boundaries and that the shape of the wave front depends on the order of the wave behind the initial pulse. Otherwise, the behavior of the linear waves resembles that of the nonlinear shock waves observed in the Sun. Research that needs to be done to firm up the conclusions reached above concerns the numerical simulations of nonlinear waves and oscillations in a three-dimensional stratified atmosphere with impulsive excitation, and observations linking directly the horizontal size of the disturbed area in upward propagation to individual waves.

Kalkofen, Wolfgang↗