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 811 records · Page 45

TPSAS-NF1676L-35475-DND

This project would identify a methodology and implement a living solution to map, both visually and utilizing some form of database, the complex network of stakeholders that the Disasters Program routinely interacts with to maximize efficiency and minimize confusion and overlapping effort during disaster responses. This project's solution will take into account factors such as stakeholder data production type, geospatial data maturity, geographic areas of interest, federal mandates, type of relationship, national priorities and many additional relevant attributes.

Lauren Cutler↗

Scientific Objectives, Measurement Needs, and Challenges Motivating the PARAGON Aerosol Initiative

Aerosols are involved in a complex set of processes that operate across many spatial and temporal scales. Understanding these processes, and ensuring their accurate representation in models of transport, radiation transfer, and climate, requires knowledge of aerosol physical, chemical, and optical properties and the distributions of these properties in space and time. To derive aerosol climate forcing, aerosol optical and microphysical properties and their spatial and temporal distributions, and aerosol interactions with clouds, need to be understood. Such data are also required in conjunction with size-resolved chemical composition in order to evaluate chemical transport models and to distinguish natural and anthropogenic forcing. Other basic parameters needed for modeling the radiative influences of aerosols are surface reflectivity and three-dimensional cloud fields. This large suite of parameters mandates an integrated observing and modeling system of commensurate scope. The Progressive Aerosol Retrieval and Assimilation Global Observing Network (PARAGON) concept, designed to meet this requirement, is motivated by the need to understand climate system sensitivity to changes in atmospheric constituents, to reduce climate model uncertainties, and to analyze diverse collections of data pertaining to aerosols. This paper highlights several challenges resulting from the complexity of the problem. Approaches for dealing with them are offered in the set of companion papers.

optical depth↗

Data distribution service-based interoperability framework for smart grid testbed infrastructure

This study presents the design and implementation of a communication and control infrastructure for smart grid operation. The proposed infrastructure enhances the reliability of the measurements and control network. The advantages of utilizing the data-centric over message-centric communication approach are discussed in the context of smart grid applications. The data distribution service (DDS) is used to implement a data-centric common data bus for the smart grid. This common data bus improves the communication reliability, enabling distributed control and smart load management. These enhancements are achieved by avoiding a single point of failure while enabling peer-to-peer communication and an automatic discovery feature for dynamic participating nodes. The infrastructure and ideas presented in this paper were implemented and tested on the smart grid testbed. A toolbox and application programing interface for the testbed infrastructure are developed in order to facilitate interoperability and remote access to the testbed. This interface allows control, monitoring, and performing of experiments remotely. Furthermore, it could be used to integrate multidisciplinary testbeds to study complex cyber-physical systems (CPS).

97 MATHEMATICS AND COMPUTING↗

Deep learning as a tool for neural data analysis: Speech classification and cross-frequency coupling in human sensorimotor cortex

A fundamental challenge in neuroscience is to understand what structure in the world is represented in spatially distributed patterns of neural activity from multiple single-trial measurements. This is often accomplished by learning a simple, linear transformations between neural features and features of the sensory stimuli or motor task. While successful in some early sensory processing areas, linear mappings are unlikely to be ideal tools for elucidating nonlinear, hierarchical representations of higher-order brain areas during complex tasks, such as the production of speech by humans. Here, we apply deep networks to predict produced speech syllables from a dataset of high gamma cortical surface electric potentials recorded from human sensorimotor cortex. We find that deep networks had higher decoding prediction accuracy compared to baseline models. Having established that deep networks extract more task relevant information from neural data sets relative to linear models (i.e., higher predictive accuracy), we next sought to demonstrate their utility as a data analysis tool for neuroscience. We first show that deep network's confusions revealed hierarchical latent structure in the neural data, which recapitulated the underlying articulatory nature of speech motor control. We next broadened the frequency features beyond high-gamma and identified a novel high-gamma-to-beta coupling during speech production. Finally, we used deep networks to compare task-relevant information in different neural frequency bands, and found that the high-gamma band contains the vast majority of information relevant for the speech prediction task, with little-to-no additional contribution from lower-frequency amplitudes. Together, these results demonstrate the utility of deep networks as a data analysis tool for basic and applied neuroscience.

60 APPLIED LIFE SCIENCES↗

IMPROVED GROUND TRUTH IN SOUTHERN ASIA USING IN-COUNTRY DATA, ANALYST WAVEFORM REVIEW, AND ADVANCED ALGORITHMS

A new catalog of seismicity at magnitudes above 2.5 for the period 1923-2008 in the Iran region is assembled from arrival times reported by global, regional, and local seismic networks. Using in-country data we have formed new events, mostly at lower magnitudes that were not previously included in standard global earthquake catalogs. The magnitude completeness of the catalog varies strongly through time, complete to about magnitude 4.2 prior to 1998 and reaching a minimum of about 3.6 during the period 1998-2005. Of the 25,722 events in the catalog, most of the larger events have been carefully reviewed for proper phase association, especially for depth phases and to eliminate outlier readings, and relocated. To better understand the quality of the data set of arrival times reported by Iranian networks that are central to this study, many waveforms for events in Iran have been re-picked by an experienced seismic analyst. Waveforms at regional distances in this region are often complex. For many events this makes arrival time picks difficult to make, especially for smaller magnitude events, resulting in reported times that can be substantially improved by an experienced analyst. Even when the signal/noise ratio is large, re-picking can lead to significant differences. Picks made by our analyst are compared with original picks made by the regional networks. In spite of the obvious outliers, the median (-0.06 s) and spread (0.51 s) are small, suggesting that reasonable confidence can be placed in the picks reported by regional networks in Iran. This new catalog has been used to assess focal depth distributions throughout Iran. A principal result of this study is that the geographic pattern of depth distributions revealed by the relatively small number of earthquakes (~167) with depths constrained by waveform modeling (+/- 4 km) are now in agreement with the much larger number of depths (~1229) determined using reanalysis of ISC arrival-times (+/-10 km), within their respective errors. This is a significant advance, as outliers and future events with apparently anomalous depths can be readily identified and, if necessary, further investigated. The patterns of reliable focal depth distributions have been interpreted in the context of Middle Eastern active tectonics. Most earthquakes in the Iranian continental lithosphere occur in the upper crust, less than about 25-30 km in depth, with the crustal shortening produced by continental collision apparently accommodated entirely by thickening and distributed deformation rather than by subduction of crust into the mantle. However, intermediate-depth earthquakes associated with subducted slab do occur across the central Caspian Sea and beneath the Makran coast. A multiple-event relocation technique, specialized to use different kinds of near-source data, is used to calibrate the locations of 24 clusters containing 901 events drawn from the seismicity catalog. The absolute locations of these clusters are fixed either by comparing the pattern of relocated earthquakes with mapped fault geometry, by using one or more cluster events that have been accurately located independently by a local seismic network or aftershock deployment, by using InSAR data to determine the rupture zone of shallow earthquakes, or by some combination of these near-source data. This technique removes most of the systematic bias in single-event locations done with regional and teleseismic data, resulting in 624 calibrated events with location uncertainties of 5 km or better at the 90% confidence level (GT590). For 21 clusters (847 events) that are calibrated in both location and origin time we calculate empirical travel times, relative to a standard 1-D travel time model (ak135), and investigate event to station travel-time anomalies as functions of epicentral distance and azimuth. Substantial travel-time anomalies are seen in the Iran region which make accurate locations impossible unless observing stations are at very short distances (less than about 200 km) or travel-time models are improved to account for lateral heterogeneity in the region. Earthquake locations in the Iran region by international agencies, based on regional and teleseismic arrival time data, are systematically biased to the southwest and have a 90% location accuracy of 18-23 km, with the lower value achievable by applying limits on secondary azimuth gap. The data set of calibrated locations reported here provides an important constraint on travel-time models that would begin to account for the lateral heterogeneity in Earth structure in the Iran region, and permit seismic networks, especially the regional ones, to obtain in future more accurate locations of the earthquakes in the region.

58 GEOSCIENCES↗

Systematic modification of functionality in disordered elastic networks through free energy surface tailoring

A combined machine learning–physics–based approach is explored for molecular and materials engineering. Specifically, collective variables, akin to those used in enhanced sampled simulations, are constructed using a machine learning model trained on data gathered from a single system. Through the constructed collective variables, it becomes possible to identify critical molecular interactions in the considered system, the modulation of which enables a systematic tailoring of the system’s free energy landscape. To explore the efficacy of the proposed approach, we use it to engineer allosteric regulation and uniaxial strain fluctuations in a complex disordered elastic network. Its successful application in these two cases provides insights regarding how functionality is governed in systems characterized by extensive connectivity and points to its potential for design of complex molecular systems.

42 ENGINEERING↗

Long short-term memory embedded nudging schemes for nonlinear data assimilation of geophysical flows

Reduced rank nonlinear filters are increasingly utilized in data assimilation of geophysical flows, but often require a set of ensemble forward simulations to estimate forecast covariance. On the other hand, predictor-corrector type nudging approaches are still attractive due to their simplicity of implementation when more complex methods need to be avoided. However, optimal estimate of nudging gain matrix might be cumbersome. In this paper, we put forth a fully nonintrusive recurrent neural network approach based on a long short-term memory (LSTM) embedding architecture to estimate the nudging term, which plays a role not only to force the state trajectories to the observations but also acts as a stabilizer. Furthermore, our approach relies on the power of archival data and the trained model can be retrained effectively due to power of transfer learning in any neural network applications. In order to verify the feasibility of the proposed approach, we perform twin experiments using Lorenz 96 system. Our results demonstrate that the proposed LSTM nudging approach yields more accurate estimates than both extended Kalman filter (EKF) and ensemble Kalman filter (EnKF) when only sparse observations are available. With the availability of emerging AI-friendly and modular hardware technologies and heterogeneous computing platforms, we articulate that our simplistic nudging framework turns out to be computationally more efficient than either the EKF or EnKF approaches.

42 ENGINEERING↗

Large-scale compute-intensive analysis via a combined in-situ and co-scheduling workflow approach, In: SC '15 Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis

Large-scale simulations can produce hundreds of terabytes to petabytes of data, complicating and limiting the efficiency of workflows. Traditionally, outputs are stored on the file system and analyzed in post-processing. With the rapidly increasing size and complexity of simulations, this approach faces an uncertain future. Trending techniques consist of performing the analysis in-situ, utilizing the same resources as the simulation, and/or off-loading subsets of the data to a compute-intensive analysis system. We introduce an analysis framework developed for HACC, a cosmological N-body code, that uses both in-situ and co-scheduling approaches for handling petabyte-scale outputs. We compare different analysis set-ups ranging from purely off-line, to purely in-situ to in-situ/co-scheduling. The analysis routines are implemented using the PISTON/VTK-m framework, allowing a single implementation of an algorithm that simultaneously targets a variety of GPU, multi-core, and many-core architectures.

Computer Science↗

Learning Only on Boundaries: A Physics-Informed Neural Operator for Solving Parametric Partial Differential Equations in Complex Geometries

Recently, deep learning surrogates and neural operators have shown promise in solving partial differential equations (PDEs). However, they often require a large amount of training data and are limited to bounded domains. In this work, we present a novel physics-informed neural operator method to solve parameterized boundary value problems without labeled data. By reformulating the PDEs into boundary integral equations (BIEs), we can train the operator network solely on the boundary of the domain. This approach reduces the number of required sample points from $O(N^d)$ to $O(N^{d-1}$), where $d$ is the domain’s dimension, leading to a significant acceleration of the training process. Additionally, our method can handle unbounded problems, which are unattainable for existing physics-informed neural networks (PINNs) and neural operators. Finally, our numerical experiments show the effectiveness of parameterized complex geometries and unbounded problems.

97 MATHEMATICS AND COMPUTING↗

The HEP.TrkX Project: deep neural networks for HL-LHC online and offline tracking

Particle track reconstruction in dense environments such as the detectors of the High Luminosity Large Hadron Collider (HL-LHC) is a challenging pattern recognition problem. Traditional tracking algorithms such as the combinatorial Kalman Filter have been used with great success in LHC experiments for years. However, these state-of-the-art techniques are inherently sequential and scale poorly with the expected increases in detector occupancy in the HL-LHC conditions. The HEP.TrkX project is a pilot project with the aim to identify and develop cross-experiment solutions based on machine learning algorithms for track reconstruction. Machine learning algorithms bring a lot of potential to this problem thanks to their capability to model complex non-linear data dependencies, to learn effective representations of high-dimensional data through training, and to parallelize easily on high-throughput architectures such as GPUs. This contribution will describe our initial explorations into this relatively unexplored idea space. Furthermore, we will discuss the use of recurrent (LSTM) and convolutional neural networks to find and fit tracks in toy detector data.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

NASA's Biosentinel Mission: Lessons Learned and What's Next

In the last two years, two BioSentinel payloads were launched to space. The ISS mission launched in December 2021, and returned to the ground in August 2022 after successfully completing eight biological experiments while validating the different instruments. On the other hand, the deep space mission launched onboard Artemis I in November 2022, and is currently in a heliocentric orbit over 20 million kilometers away from the Earth. Even though all hardware subsystems were validated in deep space, the microfluidic subsystems experienced anomalies throughout the initial 6-month mission. The main goals of this presentation are (1) to present flight data from the deep space payload, including biology, fluidics, electronics, data processing, and mission operations, and (2) to discuss the lessons learned – what worked and what did not – from this unique complex mission, and how these lessons are aiding in the development of the Lunar Exploration Instrument for space biology Applications (LEIA) mission, launching to the lunar surface on a commercial lander in 2026. As of the writing of this abstract, the satellite continues to work nominally, communicating to Earth via the Deep Space Network (DSN) twice per week. Importantly, the mission received and extension to continue recording data on the deep space radiation environment on its way to solar maximum (i.e., higher probability of solar particle events). BioSentinel is supported by NASA Exploration Systems Development Mission Directorate (ESDMD).

BioSentinel↗

Ad Hoc modeling, expert problem solving, and R&T program evaluation

A simplified cost and time (SCAT) analysis program utilizing personal-computer technology is presented and demonstrated in the case of the NASA-Goddard end-to-end data system. The difficulties encountered in implementing complex program-selection and evaluation models in the research and technology field are outlined. The prototype SCAT system described here is designed to allow user-friendly ad hoc modeling in real time and at low cost. A worksheet constructed on the computer screen displays the critical parameters and shows how each is affected when one is altered experimentally. In the NASA case, satellite data-output and control requirements, ground-facility data-handling capabilities, and project priorities are intricately interrelated. Scenario studies of the effects of spacecraft phaseout or new spacecraft on throughput and delay parameters are shown. The use of a network of personal computers for higher-level coordination of decision-making processes is suggested, as a complement or alternative to complex large-scale modeling.

Silverman, B. G.↗

A Study of Physics-Informed Deep Learning for System Fluid Dynamics Closures

The sub-grid-scale (SGS) physics models, or so-called closure relations (CR), are essential in thermal-fluid modeling and simulation codes, ranging from large-eddy simulation to system-level thermo-fluid dynamic problems. While details vary, all thermo-fluid codes solve coupled mass-momentum-energy conservation partial differential equations (PDE). Traditionally, the analytical form of closure models has to be known, and then we calibrate the model using data from a set of applicable experiments. Characteristically, it takes long effort to gain insights and mechanistic understanding needed for developing closure models. Such a model is a compact form of data. The long time required for new model development limits the pace of model application while dealing with new geometries, new coolants or system conditions, particularly in newly designed systems. Alternative to mechanistic and semi-analytical models, it is possible to discover the pattern of closure relations by applying the modern machine learning methods to the body of relevant experimental data. Deep learning (DL) is a relatively new branch in the machine learning family, and a major breakthrough has been achieved since Hinton introduces a fast algorithm to efficiently train a feed-forward neural network (NN) with multilayer perceptrons. Generally, any NN with more than two layers is considered as deep learning, and it contains a large number of hyper-parameters to achieve pattern recognition or regression for complex datasets. Hornik shows that multilayer NN is compatible with universal approximation theorem. There is no constraint for a multilayer NN to capture the properties of any measurable information. This capability makes DL attractive for closure models in a system thermal-hydraulic code. As system geometry and/or operating conditions change, new physical experiments (increasingly sophisticated) are conducted to support evaluation and revision of closure models. Such a so-called data-driven modeling relaxes the structural form of system codes to directly assimilate data instead of hard-wiring the model to pre-calibrated correlations. The objective of this research is to enable cost-effective development of a DL-based closure that reflects the data. The previous work on applying machine learning techniques to fluid dynamics focused on statistical learning with basic NN from direct numerical simulations (DNS) dataset. In the present work, we show that DL-based closure models can cause the fluid calculation to diverge if multilayer NN is used. To address the issue, we investigate conditions by which the DL-based closure models work compatibly, stably, and effectively with PDE-constrained forward prediction problems. Optimally meeting the compatibility conditions constitutes a physics-informed deep learning (PIDL) strategy. We refer to this process as PDE-constrained machine learning or PDE-ML. To develop the technical basis for PDE-ML using DL-based fluid dynamics closure, numerical experiments are performed for manufactured problems of increasing complexity. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Linking Advanced Visualization and MATLAB for the Analysis of 3D Gene Expression Data

Three-dimensional gene expression PointCloud data generated by the Berkeley Drosophila Transcription Network Project (BDTNP) provides quantitative information about the spatial and temporal expression of genes in early Drosophila embryos at cellular resolution. The BDTNP team visualizes and analyzes Point-Cloud data using the software application PointCloudXplore (PCX). To maximize the impact of novel, complex data sets, such as PointClouds, the data needs to be accessible to biologists and comprehensible to developers of analysis functions. We address this challenge by linking PCX and Matlab via a dedicated interface, thereby providing biologists seamless access to advanced data analysis functions and giving bioinformatics researchers the opportunity to integrate their analysis directly into the visualization application. To demonstrate the usefulness of this approach, we computationally model parts of the expression pattern of the gene even skipped using a genetic algorithm implemented in Matlab and integrated into PCX via our Matlab interface.

59 BASIC BIOLOGICAL SCIENCES↗

A Scalable, Out-of-Band Diagnostics Architecture for International Space Station Systems Support

The computational infrastructure of the International Space Station (ISS) is a dynamic system that supports multiple vehicle subsystems such as Caution and Warning, Electrical Power Systems and Command and Data Handling (C&DH), as well as scientific payloads of varying size and complexity. The dynamic nature of the ISS configuration coupled with the increased demand for payload support places a significant burden on the inherently resource constrained computational infrastructure of the ISS. Onboard system diagnostics applications are hosted on computers that are elements of the avionics network while ground-based diagnostic applications receive only a subset of available telemetry, down-linked via S-band communications. In this paper we propose a scalable, out-of-band diagnostics architecture for ISS systems support that uses a read-only connection for C&DH data acquisition, which provides a lower cost of deployment and maintenance (versus a higher criticality readwrite connection). The diagnostics processing burden is off-loaded from the avionics network to elements of the on-board LAN that have a lower overall cost of operation and increased computational capacity. A superset of diagnostic data, richer in content than the configured telemetry, is made available to Advanced Diagnostic System (ADS) clients running on wireless handheld devices, affording the crew greater mobility for troubleshooting and providing improved insight into vehicle state. The superset of diagnostic data is made available to the ground in near real-time via an out-of band downlink, providing a high level of fidelity between vehicle state and test, training and operational facilities on the ground.

Fletcher, Daryl P.↗

Identifying apparent local stable isotope equilibrium in a complex non-equilibrium system

Although being out of equilibrium, biomolecules in organisms have the potential to approach isotope equilibrium locally because enzymatic reactions are intrinsically reversible. A rigorous approach that can describe isotope distribution among biomolecules and their apparent deviation from equilibrium state is lacking, however. Applying the concept of distance matrix in graph theory, we propose that apparent local isotope equilibrium among a subset of biomolecules can be assessed using an apparent fractionation difference (|Δα|) matrix, in which the differences between the observed isotope composition (δ') and the calculated equilibrium fractionation factor (1000lnβ) can be more rigorously evaluated than by using a previous approach for multiple biomolecules. We tested our |Δα| matrix approach by re-analyzing published data of different amino acids (AAs) in potato and in green alga. Our re-analysis shows that biosynthesis pathways could be the reason for an apparently close-to-equilibrium relationship inside AA families in potato leaves. Different biosynthesis/degradation pathways in tubers may have led to the observed isotope distribution difference between potato leaves and tubers. The analysis of data from green algae does not support the conclusion that AAs are further from equilibrium in glucose-cultured green algae than in the autotrophic ones. Application of the |Δα| matrix can help us to locate potential reversible reactions or reaction networks in a complex system such as a metabolic system. The same approach can be broadly applied to all complex systems that have multiple components, e.g. geochemical or atmospheric systems of early Earth or other planets.

59 BASIC BIOLOGICAL SCIENCES↗

High-Performance Computing for Real-Time Grid Analysis and Operation

Power grids worldwide are undergoing an unprecedented transition as a result of grid evolution meeting information revolution. The grid evolution is largely driven by the desire for green energy. Emerging grid technologies such as renewable generation, smart loads, plug-in hybrid vehicles, and distributed generation provide opportunities to generate energy from green sources and to manage energy use for better system efficiency. With utility companies actively deploying these technologies, a high level of penetration of these new technologies is expected in the next 5-10 years, bringing in a level of intermittency, uncertainties, and complexity that the grid did not see nor design for. On the other hand, the information infrastructure in the power grid is being revolutionized with large-scale deployment of sensors and meters in both the transmission and distribution networks. The future grid will have two-way flows of both electrons and information. The challenge is how to take advantage of the information revolution: pull the large amount of data in, process it in real time, and put information out to manage grid evolution. Without addressing this challenge, the opportunities in grid evolution will remain unfulfilled. This transition poses grand challenges in grid modeling, simulation, and information presentation. The computational complexity of underlying power grid modeling and simulation will significantly increase in the next decade due to an increased model size and a decreased time window allowed to compute model solutions. High-performance computing is essential to enable this transition. The essential technical barrier is to vastly increase the computational speed so operation response time can be reduced from minutes to seconds and sub-seconds. The speed at which key functions such as state estimation and contingency analysis are conducted (typically every 3-5 minutes) needs to be dramatically increased so that the analysis of contingencies is both comprehensive and real time. An even bigger challenge is how to incorporate dynamic information into real-time grid operation. Today’s online grid operation is based on a static grid model and can only provide a static snapshot of current system operation status, while dynamic analysis is conducted offline because of low computational efficiency. The offline analysis uses a worst-case scenario to determine transmission limits, resulting in under-utilization of grid assets. This conservative approach does not necessarily lead to reliability. Many times, actual power grid scenarios are not studied, and they will push the grid over the edge and resulting in outages and blackouts. This chapter addresses the HPC needs in power grid analysis and operations. Example applications such as state estimation and contingency analysis are given to demonstrate the value of HPC in power grid applications. Future research directions are suggested for high performance computing applications in power grids to improve the transparency, efficiency, and reliability of power grids.

Contingency Analysis↗

NeuroSEM: A hybrid framework for simulating multiphysics problems by coupling PINNs and spectral elements

Multiphysics problems that are characterized by complex interactions among fluid dynamics, heat transfer, structural mechanics, and electromagnetics, are inherently challenging due to their coupled nature. While experimental data on certain state variables may be available, integrating these data with numerical solvers remains a significant challenge. Physics-informed neural networks (PINNs) have shown promising results in various engineering disciplines, particularly in handling noisy data and solving inverse problems in partial differential equations (PDEs). However, their effectiveness in forecasting nonlinear phenomena in multiphysics regimes, particularly involving turbulence, is yet to be fully established. Here, this study introduces NeuroSEM, a hybrid framework integrating PINNs with the highfidelity Spectral Element Method (SEM) solver, Nektar++. NeuroSEM leverages the strengths of both PINNs and SEM, providing robust solutions for multiphysics problems. PINNs are trained to assimilate data and model physical phenomena in specific subdomains, which are then integrated into the Nektar++ solver. We demonstrate the efficiency and accuracy of NeuroSEM for thermal convection in cavity flow and flow past a cylinder. The framework effectively handles data assimilation by addressing those subdomains and state variables where the data is available. We applied NeuroSEM to the Rayleigh-B´enard convection system, including cases with missing thermal boundary conditions and noisy datasets. Finally, we applied the proposed NeuroSEM framework to real particle image velocimetry (PIV) data to capture flow patterns characterized by horseshoe vortical structures. Our results indicate that NeuroSEM accurately models the physical phenomena and assimilates the data within the specified subdomains. The framework’s plug-and-play nature facilitates its extension to other multiphysics or multiscale problems. Furthermore, NeuroSEM is optimized for efficient execution on emerging integrated GPU-CPU architectures. This hybrid approach enhances the accuracy and efficiency of simulations, making it a powerful tool for tackling complex engineering challenges in various scientific domains.

42 ENGINEERING↗