Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Distributed computing, machine learning”

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 577 records · Page 32

LEGION: Lightweight Expandable Group of Independently Operating Nodes

LEGION is a lightweight C-language software library that enables distributed asynchronous data processing with a loosely coupled set of compute nodes. Loosely coupled means that a node can offer itself in service to a larger task at any time and can withdraw itself from service at any time, provided it is not actively engaged in an assignment. The main program, i.e., the one attempting to solve the larger task, does not need to know up front which nodes will be available, how many nodes will be available, or at what times the nodes will be available, which is normally the case in a "volunteer computing" framework. The LEGION software accomplishes its goals by providing message-based, inter-process communication similar to MPI (message passing interface), but without the tight coupling requirements. The software is lightweight and easy to install as it is written in standard C with no exotic library dependencies. LEGION has been demonstrated in a challenging planetary science application in which a machine learning system is used in closed-loop fashion to efficiently explore the input parameter space of a complex numerical simulation. The machine learning system decides which jobs to run through the simulator; then, through LEGION calls, the system farms those jobs out to a collection of compute nodes, retrieves the job results as they become available, and updates a predictive model of how the simulator maps inputs to outputs. The machine learning system decides which new set of jobs would be most informative to run given the results so far; this basic loop is repeated until sufficient insight into the physical system modeled by the simulator is obtained.

Burl, Michael C.↗

Integrating Learning and Physics based Computation for Fast Online Transient Analysis

In this work, a novel method that integrates learning and physics based computation is developed for greatly accelerating the simulation of full power system transient trajectories. To solve the dynamic algebraic equations, the method replaces the time-consuming dynamic computation for generator dynamics with trained predictors, while retaining the time-efficient algebraic computation of solving AC-power flow (PF) for power systems. In particular, a predictor is trained for each generator, and the system trajectories are computed by alternating steps of calling the predictors and solving AC-PF. The proposed method also allows fully parallelizable training strategies and a flexible trade-off between training time and testing accuracy. Comprehensive evaluations of the proposed method for transient/dynamic contingency analysis of the New York/New England 16-machine 68-bus power systems demonstrate excellent performance and significant acceleration of computation.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Physics-Informed Machine Learning Model for Ceramic Matrix Composite Creep

A physics-informed recurrent neural network (RNN) based surrogate model is developed to emulate the nonlinear, time-dependent constitutive behavior of ceramic matrix composites (CMCs) driven by matrix damage and constituent creep at the microscale. Physics-informed constraints are introduced into the surrogate model through regularization to ground the prediction in physics and improve its predictive capabilities. Training data is generated using the high-fidelity generalized method of cells (HFGMC) approach which calls appropriate creep and damage models for each of the constituents. This coupling permits simulating the nonlinear behavior of CMCs based on constituent response at the microscale along with microstructural features such as fiber and porosity volume fraction and fiber radius. The microscale repeating unit cell is loaded under creep fatigue conditions to replicate the material loading experienced in a turbine engine. Therefore, the RNN-based surrogate model is tasked with predicting, as a function of variable input stress sequence, temperature, and microstructural features, the resulting strain history response while satisfying physical constraints related to creep rate, isochoric inelastic deformation, and strain energy density. The trained surrogate model is shown to effectively match the strain history over quantified distributions of microstructural features and relevant loading regimes and temperatures. Neural network based surrogate models can offer efficient alternatives to running computationally intensive multiscale material models to simulate the nonlinear response of large structural models. Therefore, the presented work provides evidence towards the feasibility of developing, training, and running such models for CMCs with complex microstructures, nonlinear time-dependent material response, and under non-monotonic loading conditions.

ceramic matrix composites↗

Improving ideal MHD equilibrium accuracy with physics-informed neural networks

We present a novel approach to compute three-dimensional magnetohydrodynamic equilibria with isotropic pressure profiles and nested surfaces by parametrizing Fourier modes with artificial neural networks (NNs). The full nonlinear global force residual of single equilibria across the volume in real space is then minimized with first order optimizers and compared to equilibria computed by conventional solvers. Already, we observe competitive computational cost to arrive at the same minimum residuals computable with existing codes. With increased computational cost, lower minima of the residual are computable with the NNs than with any other tested solver, establishing a new lower bound for the force residual. We use minimally complex NNs, and we expect significant improvements for solving not only single equilibria with NNs, but also for creating NN models valid over continuous distributions of equilibria.

ideal magnetohydrodynamics↗

Property space mapping of Pseudomonas aeruginosa permeability to small molecules

Two membrane cell envelopes act as selective permeability barriers in Gram-negative bacteria, protecting cells against antibiotics and other small molecules. Significant efforts are being directed toward understanding how small molecules permeate these barriers. In this study, we developed an approach to analyze the permeation of compounds into Gram-negative bacteria and applied it to Pseudomonas aeruginosa, an important human pathogen notorious for resistance to multiple antibiotics. The approach uses mass spectrometric measurements of accumulation of a library of structurally diverse compounds in four isogenic strains of P. aeruginosa with varied permeability barriers. We further developed a machine learning algorithm that generates a deterministic classification model with minimal synonymity between the descriptors. This model predicted good permeators into P. aeruginosa with an accuracy of 89% and precision above 58%. The good permeators are broadly distributed in the property space and can be mapped to six distinct regions representing diverse chemical scaffolds. We posit that this approach can be used for more detailed mapping of the property space and for rational design of compounds with high Gram-negative permeability.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Steam Generator Model Design Parameter Sensitivity Study Using Advanced Optimization Tools

This study focuses on design parameter sensitivity studies pertaining to a steam generator (SG) model, using both Python and machine-learning tools. The SG model is a mathematical representation (including fluid flow and heat transfer equations/models/correlations) of a steam-generating unit in a pressurized water reactor (PWR)-type small modular reactor (SMR) system. Design studies involve changing the model’s input design parameters (e.g., temperature, pressure, mass flow rate) to observe the resulting effects on the output of the system (e.g., heat transfer coefficient [HTC], Nusselt number, heat transfer performance). Sensitivity studies analyze the degree to which system output and/or desired parameters (e.g., HTC or heat transfer performance) are sensitive to changes in input parameters. By using machine-learning tools such as the Risk Analysis Virtual Environment (RAVEN) developed at Idaho National Laboratory (INL), detailed design parametric sensitivity studies and model optimization were performed. Six input parameters—namely, the pressure, temperature, and mass flow rate for the inlet of the primary-side (hot fluid) and secondary-side (cold fluid) of the SG—were randomly perturbed via RAVEN’s Monte Carlo Sampler module, using uniform distributions (±1% relative changes). The analysis results give valuable insights into SG system performance and optimization, and provide justification for researching optimized sensor placement to effectively monitor and obtain experimental data.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Steam Generator Model Design Parameter Sensitivity Study Using Advanced Optimization Tools

This study focuses on design parameter sensitivity studies pertaining to a steam generator (SG) model, using both Python and machine-learning tools. The SG model is a mathematical representation (including fluid flow and heat transfer equations/models/correlations) of a steam-generating unit in a pressurized water reactor (PWR)-type small modular reactor (SMR) system. Design studies involve changing the model’s input design parameters (e.g., temperature, pressure, mass flow rate) to observe the resulting effects on the output of the system (e.g., heat transfer coefficient [HTC], Nusselt number, heat transfer performance). Sensitivity studies analyze the degree to which system output and/or desired parameters (e.g., HTC or heat transfer performance) are sensitive to changes in input parameters. By using machine-learning tools such as the Risk Analysis Virtual Environment (RAVEN) developed at Idaho National Laboratory (INL), detailed design parametric sensitivity studies and model optimization were performed. Six input parameters—namely, the pressure, temperature, and mass flow rate for the inlet of the primary-side (hot fluid) and secondary-side (cold fluid) of the SG—were randomly perturbed via RAVEN’s Monte Carlo Sampler module, using uniform distributions (±1% relative changes). The analysis results give valuable insights into SG system performance and optimization, and provide justification for researching optimized sensor placement to effectively monitor and obtain experimental data.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Flexible Pilot Jobs Framework for Distributed High Throughput Computing

Experimental particle physics has been at the forefront of analyzing the world’s largest datasets for decades. The high-energy physics (HEP) community was among the first to develop suitable software and computing tools for this purpose. GlideinWMS is a Glidein-based workload management system whose purpose is to provide experiments like CMS at CERN, DUNE at Fermilab, and others, a way to access and efficiently use vast amounts of computing resources. This system wants to provide a simple way to submit jobs to a set of computing resources, that will be provided to users behind the scenes. Glideins are the pilot jobs executed on the worker nodes at the grid sites, performing operations such as hardware detection, environment setup, and error handling. After all these operations, they will launch the actual user job. Many grid sites are supported, such as shared clusters, Google CE, and AWS. My internship aimed to design and code a flexible pilot jobs framework that will replace the one used by GlideinWMS, developing a modular and flexible skeleton of the Glidein and adding further functionalities. My project also focused on the application of machine learning techniques as support to this management system.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Automated Quantitative Analysis of Extended Irradiation Defects - Dislocations, Voids and Precipitates in Neutron Irradiated HT-9 Steel

Structural alloys that make up the core components of nuclear reactors are exposed to radiation that can substantially alter the microstructure and mechanical properties of the material. Irradiation can cause the formation of extended defects including dislocation line and loops, voids and various types of precipitates. To understand the irradiation effects, it is critical to characterize these defects individually and collectively to quantify how they evolve throughout irradiation stages. Transmission electron microscopy (TEM) is a standard tool for providing extensive size, number density and spatial distribution of extended irradiation defects, however, manual measurement of defect size and number density requires time-intensive effort to analyze enough defects that provide a statistically meaningful result and are prone to errors of subjectivity. In this work, we developed a set of MATLAB algorithms offering robust and automated identification and quantification of extended irradiation defects like dislocation lines, voids and precipitates commonly observed in neutron irradiated cladding alloys [1]. Using a neutron-irradiated HT-9 martensitic alloy irradiated at 412°C to 111.8 dpa, we demonstrate the capability of these algorithms for quantitative defect analysis. Figure 1 presents defect images of the HT-9 using an advanced diffraction contrast imaging scanning transmission electron microscopy (DCI STEM) that suppresses bend contours and provides better defect contrast compared to conventional TEM [2]. The first step of defect analysis is defect identification, a process that can be digitally achieved by contrast thresholding (for simple defect contrast) and by machine learning computer vision for defect semantic segmentation [3]. Here, labeled binary defect images (Figures 2a and 2c) were used. For dislocation density quantification, we employed an optimized circular-grid intersection method first introduced by Dr. John Steeds [4]. This method is shown in Figures 2a and 2b, where the number of intersections of dislocation lines with circular traces was measured. This produces an unbiased sampling of dislocation lines, regardless of their orientation. This process was optimized with automated digital intersection number and grid length registration. Interestingly, it was found that a relatively large number of circles in a set circular gird is was needed to produce an invariant result (Figure 2b), with a low systematic measurement error. This demonstrates the main advantage of computer-based defect analysis, in performing repetitive operations quickly with a reduced systematic error. For the quantitative analysis of voids and precipitates a challenge encountered with automated defect analysis is feature overlap, particularly when the density of voids/precipitates is high. To mitigate this problem, as shown in Figure 2d, we modified the popular circular Hough transformation and treated ‘circular shape’ as the particle descriptor for voids/precipitates, efficiently resolving the overlapping particles. In summary, this automated digital quantification together with machine learning enabled defect identification [3] opens new opportunities for standardizing the characterization workflow for quantitative analysis of extended irradiation defects.

Sainju, Rajat↗

Monitoring Airspace Complexity and Determining Contributing Factors

The national airspace has evolved over many years to accommodate increased traffic demand [1] while simultaneously maintaining one of the safest forms of transportation [2], [3]. One of the reasons for this success is the ability of the system and the operators to adapt and accommodate to situations that routinely disrupt optimal operations. These situations may include: adverse weather, delays, early arrivals, equipment outages, and other factors that are outside the operators ability to control. These factors can lead to states where automation is unable to properly handle these issues and therefore air traffic controllers and pilots have to intervene, ultimately increasing communication between operators resulting in higher workload. As controller workload increases to handle sub-optimal operating conditions this can be viewed as an increase in complexity. The reasoning for this is because humans are now required to make tactical decisions in response to external factors, resulting in a departure from the strategic plan where operations would be more efficiently managed. Human operators control airspace complexity under rigid regulations that are constantly changing. The airspace is divided into sectors and the number of aircraft assigned to each controller is limited for safe handling. There has been past work that devised airspace complexity metrics in commercial aviation and related these metrics to controller workload (e.g., [4],[5]). The upper bounds on the system load are pre-determined. Such bounds on complexity make for a safe system, but the system cannot scale and adapt to autonomous, dense, and heterogeneous traffic, including the many types of Unmanned Aerial Vehicles (UAVs) envisioned to be added to the operations. We hypothesize that, as traffic density and heterogeneity grow, and other key metrics change, there will be phase transitions at which the way traffic should be managed changes significantly [6]. We offer a method for in-time detection of contributing factors that lead to phase transitions, characterized by increased complexity. To the best of our knowledge, there is no tool similar to our proposed effort that identifies such contributing factors or precursor patterns. To define the scope we are proposing to measure complexity from the viewpoint of the Terminal Radar Approach Control Facilities (TRACON) controller’s perspective. In particular we are analyzing arrivals into KSFO. With safety as the top concern for airspace operators, it is important to recognize that as density and heterogeneity grow, the focus of the system will change. Times of the day when the airspace has low density and heterogeneity, the flights will follow more efficient paths where the aircraft move on established routes that are more or less directly to the destination. However, when density and heterogeneity increases, the system will begin changing focus to avoiding conflicts and collisions and route the flights in a more flexible way. Higher flexibility requires more communication and coordination between controllers and pilots which the current automation is unable to handle. This paper proposes a novel approach that monitors airspace complexity at multiple scales, uses a Machine Learning-based tool that predicts when operations will transition to a regime of greater complexity, and identifies actions that can reduce the complexity while still maintaining efficient and safe operations. We demonstrate our proposed approach using data from multiple complementary sources. This includes, but is not limited to: historical aircraft surveillance data from NASA’s Sherlock Data Warehouse [7], METAR weather data, and airport configuration data from Aviation System Performance Metrics (ASPM). The surveillance data flight paths are sampled at a variable sample rate — increasing as the aircraft approaches the airport. This is due to how Sherlock manages flight track stitching between different radar facilities which have different sampling rates. The weather and performance data are logged at defined intervals throughout the day at a courser refresh rate. In addition to the logged data and metrics, we leverage pre-defined Standard Terminal Arrival Routes (STARs) procedures to characterize the path of each flight. Each flight files for one of these routes in the flight plan well before entering the terminal airspace, and approximately follows the route until it leaves the STAR, typically on the final fix of a runway transition. However, most flights do not always fly the full STAR procedure to completion [8], but the majority do adhere to the fixes within the common route of the procedure. Our approach leverages fixes in the common route of each of the STARs to build a reference path to the airport. This allows us to characterize the flight paths in what we are defining as the “maneuvering area” (the airspace between the STAR and before the flight is lined up on the runway’s final approach) to determine how off nominal the flights are to calculate its complexity score. Determining airspace complexity is a concept that does not have a concrete answer. In designing this metric, we consider what increases the workload for the air traffic controllers. Consequently more specialized vectoring maneuvers results in higher workload. Accordingly, we start with a theory: each flight has a direct path it takes from the STAR’s common route to the final approach’s outer marker fix for the flight’s landing runway. It is important to note that the direct path is only used as a reference. If the majority of the flights have a large consistent offset as compared to other routes it does not necessarily mean that those flights have higher complexity. We are merely building a distribution based on this direct path for that particular STAR and runway pair to determine the normal mode of operations for that route. Flights that are in the upper tail of these distributions will result in higher complexity scores and flights that fly in the median will represent the normal mode of operations and therefore will have lower complexity scores. Since flights following each STAR route take different paths to the airport, we have a different distribution for each STAR route and therefore can model these distributions to compute a complexity score from their respective normalized distributions. To evaluate the effectiveness of our proposed airspace complexity metric we will compare against an established approach based on trajectory clustering [9]. This unsupervised learning technique consists of the following steps: (1) identify the general maneuvering areas (waypoints) by performing $\kappa$-means or DBSCAN clustering on locations where aircraft frequently turn based on the surveillance radar track data, (2) map flight trajectories onto sequences of waypoints, and (3) cluster the sequences based on their common subsequences. From a high-level perspective, this baseline model learns nominal operations in the airspace through the sequence of waypoints that are representative of where aircraft change direction and defines deviations from the nominal operations as “complex.” Therefore, more deviations from the nominal operations correspond to higher complexity values. For our validation, we re-implemented this technique and tune model hyper-parameters to correctly detect waypoints for the arrival traffic into the San Francisco bay area. We will compute the complexity measure over a one-year period using our proposed technique as well as the baseline. Our validation will be based on each technique’s ability to detect a set of undesirable outcomes (e.g., go-arounds, holding patterns, average time in the airspace, etc.). Since our current complexity metric is derived from the offset from the direct reference path, it’s important to understand what causes these offsets. In many of the flights with high offset distance, flights performing holding patterns and S turns can be observed. These maneuvering tactics are utilized to add distance between the aircraft and the destination runway to prevent multiple flights from having conflicting arrival times. In order to predict a rise in complexity (or the precursor to complexity), it’s necessary to be able to identify these potential conflicts (which in turn, result in higher offsets). To do this, we define a “representative flight” for each STAR route and runway pair. This flight is approximately the path the flight would take if there was a clear path with no other flights in the airspace — including the time remaining to the airport. We first identify the flights for a given STAR runway pair using the offset to the reference path distributions that fall between the 44-55 percentiles. This yields the flights that conform to the most normal mode of operation. Each of these flights is partitioned based on the percent complete from the entry point into the maneuvering areas from 0\% – 100\% complete. Then for each percent “bin”, we take the median value of the flight’s latitude/longitude coordinates, airspeed, and (non causal) time remaining to the airport to construct a lookup table for each percent complete bin on a given route. As a flight enters the maneuvering area, we can find the estimated arrival time of a flight to the airport by finding the closest point to the representative path’s percent complete bin (relative to the flight’s current position at any snapshot in the airspace) and therefore retrieve the corresponding remaining time left on the “representative path”. We assume that the flight will follow the representative path to completion when deriving these estimates. We can then compare these estimated arrival times against other flights for the same snapshot in time to identify potential conflicts. If more flights are estimated to arrive within a tolerance window than there are runways available, then we have a potential conflict. We can use this derived measure along with other factors expected to add disruption to the operation such as weather and runway configuration changes as an input to machine learning tools to detect precursors that increases in our complexity measure. This novel method will assist in uncovering insights into the contributing factors that lead to increased complexity that may allow for in-time responses to avoid reaching a high complexity state in the airspace.

complexity↗

Parallel sorting algorithm classification: is manual instrumentation necessary?

Understanding parallel algorithms is crucial for accelerating scientific simulations on complex, distributed memory, high-performance computers. Modern algorithm classification approaches learn semantics directly from source code to differentiate between algorithms, however, accessing source code is not always possible. We can learn about parallel algorithms from observing their performance, as programs running the same algorithms and using the same hardware should exhibit similar performance characteristics. We present an approach to learn algorithm classes from parallel performance data directly in order to classify algorithms without access to the source code. We extend previous work to enable classifying parallel sorting algorithms using automatic instrumentation instead of requiring manual region annotations in the source code. In this work, we design and demonstrate a study for classification of parallel sorting algorithms using parallel performance data collected from automatic instrumentation, and evaluate the performance of our new methodology on classification. We leverage Caliper to collect the performance data, Thicket for our exploratory data analysis (EDA), and PyTorch and Scikit-learn to evaluate the effectiveness of random forests, support vector machines (SVMs), decision trees, neural networks, and logistic regressions on parallel performance data. Additionally, we study noise in parallel performance data, whether the removal of noise and pre-processing of the data is necessary to accurately classify parallel sorting algorithms, and determine the effectiveness of features created from performance data. In conclusion, we demonstrate classification accuracy for these five different models of up to 97.7% across four different parallel algorithm classes.

Algorithm Classification↗

IMPACT OF ATMOSPHERIC CHROMATIC EFFECTS ON WEAK LENSING MEASUREMENTS

Current and future imaging surveys will measure cosmic shear with statistical precision that demands a deeper understanding of potential systematic biases in galaxy shape measurements than has been achieved to date. We use analytic and computational techniques to study the impact on shape measurements of two atmospheric chromatic effects for ground-based surveys such as the Dark Energy Survey and the Large Synoptic Survey Telescope (LSST): (1) atmospheric differential chromatic refraction and (2) wavelength dependence of seeing. We investigate the effects of using the point-spread function (PSF) measured with stars to determine the shapes of galaxies that have different spectral energy distributions than the stars. We find that both chromatic effects lead to significant biases in galaxy shape measurements for current and future surveys, if not corrected. Using simulated galaxy images, we find a form of chromatic “model bias” that arises when fitting a galaxy image with a model that has been convolved with a stellar, instead of galactic, PSF. We show that both forms of atmospheric chromatic biases can be predicted (and corrected) with minimal model bias by applying an ordered set of perturbative PSF-level corrections based on machine-learning techniques applied to six-band photometry. Catalog-level corrections do not address the model bias. We conclude that achieving the ultimate precision for weak lensing from current and future ground-based imaging surveys requires a detailed understanding of the wavelength dependence of the PSF from the atmosphere, and from other sources such as optics and sensors. The source code for this analysis is available at https://github.com/DarkEnergyScienceCollaboration/chroma.

79 ASTRONOMY AND ASTROPHYSICS↗

Robust Decentralized Learning Using ADMM With Unreliable Agents

Many signal processing and machine learning problems can be formulated as consensus optimization problems which can be solved efficiently via a cooperative multi-agent system. However, the agents in the system can be unreliable due to a variety of reasons: noise, faults and attacks. Providing erroneous updates leads the optimization process in a wrong direction, and degrades the performance of distributed machine learning algorithms. This paper considers the problem of decentralized learning using ADMM in the presence of unreliable agents. First, we rigorously analyze the effect of erroneous updates (in ADMM learning iterations) on the convergence behavior of the multi-agent system. We show that the algorithm linearly converges to a neighborhood of the optimal solution under certain conditions and characterize the neighborhood size analytically. Next, we provide guidelines for network design to achieve a faster convergence to the neighborhood. Here, we also provide conditions on the erroneous updates for exact convergence to the optimal solution. Finally, to mitigate the influence of unreliable agents, we propose ROAD , a robust variant of ADMM, and show its resilience to unreliable agents with an exact convergence to the optimum.

97 MATHEMATICS AND COMPUTING↗

A generative artificial intelligence framework for long-time plasma turbulence simulations

Generative deep learning techniques are employed in a novel framework for the construction of surrogate models capturing the spatiotemporal dynamics of 2D plasma turbulence. The proposed Generative Artificial Intelligence Turbulence (GAIT) framework enables the acceleration of turbulence simulations for long-time transport studies. GAIT leverages a convolutional variational auto-encoder and a recurrent neural network to generate new turbulence data from existing simulations, extending the time horizon of transport studies with minimal computational cost. The application of the GAIT framework to plasma turbulence using the Hasegawa–Wakatani (HW) model is presented, evaluating its performance via various analyses. Very good agreement is found between the GAIT and the HW models in the spatiotemporal Fourier and Proper Orthogonal Decomposition spectra, the flow topology characterized by the Okubo–Weiss parameter, and the time autocorrelation function of turbulent fluctuations. Excellent agreement has also been obtained in the probability distribution function of particle displacements and the effective turbulent diffusivity. In-depth analyses of the latent space of turbulent states, choice of hyperparameters and alternative deep learning models for the time prediction are presented. Our results highlight the potential of Artificial Intelligence-based surrogate models to overcome the computational challenges in turbulence simulation, which can be extended to other situations such as geophysical fluid dynamics.

Artificial intelligence↗

Operator learning for predicting multiscale bubble growth dynamics

We report simulating and predicting multiscale problems that couple multiple physics and dynamics across many orders of spatiotemporal scales is a great challenge that has not been investigated systematically by deep neural networks (DNNs). Herein, we develop a framework based on operator regression, the so-called deep operator network (DeepONet), with the long-term objective to simplify multiscale modeling by avoiding the fragile and time-consuming “hand-shaking” interface algorithms for stitching together heterogeneous descriptions of multiscale phenomena. To this end, as a first step, we investigate if a DeepONet can learn the dynamics of different scale regimes, one at the deterministic macroscale and the other at the stochastic microscale regime with inherent thermal fluctuations. Specifically, we test the effectiveness and accuracy of the DeepONet in predicting multirate bubble growth dynamics, which is described by a Rayleigh–Plesset (R–P) equation at the macroscale and modeled as a stochastic nucleation and cavitation process at the microscale by dissipative particle dynamics (DPD). First, we generate data using the R–P equation for multirate bubble growth dynamics caused by randomly time-varying liquid pressures drawn from Gaussian random fields (GRFs). Our results show that properly trained DeepONets can accurately predict the macroscale bubble growth dynamics and can outperform long short-term memory networks. We also demonstrate that the DeepONet can extrapolate accurately outside the input distribution using only very few new measurements. Subsequently, we train the DeepONet with DPD data corresponding to stochastic bubble growth dynamics. Although the DPD data are noisy and we only collect sparse data points on the trajectories, the trained DeepONet model is able to predict accurately the mean bubble dynamics for time-varying GRF pressures. Taken together, our findings demonstrate that DeepONets can be employed to unify the macroscale and microscale models of the multirate bubble growth problem, hence providing new insight into the role of operator regression via DNNs in tackling realistic multiscale problems and in simplifying modeling with heterogeneous descriptions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

PopGNN: Graph Neural Network-Based Flexible Future Population Forecasting Model

Accurate population forecasts is important to plan critical infrastructure and services, from housing and education to healthcare and transport. However, traditional population prediction studies have only employed traditional machine learning models limited to capture complex spatial interdependencies and patterns. Althogh recently computer vision-based framework was introduced with with promising accuracy, it has critical limitations for real-world planning applications: it function only at fixed spatial resolutions, restricting their use in diverse boundaries such as census tracts, neighborhoods, or administrative zones. Therefore, this study suggests a Graph Neural Network (GNN)-based population prediction framework, called PopGNN. This model recorded remarkable performance compared with state-of-the-art models and traditional baseline models in the grid and administrative boundaries. Furthermore, our framework achieved comparable predictive accuracy to a computer vision-based model in both the South Korea and Tennessee case studies. Consequently, this study is valuable in that a single model can provide accurate population forecasts that address diverse planning demands, ranging from granular grid-level estimates for precise service allocation and facility location planning to aggregate administrative-level forecasts for macro-scale regional policy and resource distribution.

97 MATHEMATICS AND COMPUTING↗

MassiveGNN: Efficient Training via Prefetching for Massively Connected Distributed Graphs

Graph Neural Networks (GNN) are indispensable in learning from graph-structured data, yet their rising computational costs, especially on massively connected graphs, pose significant challenges in terms of execution performance. To tackle this, distributed-memory solutions such as partitioning the graph to concurrently train multiple replicas of GNNs are in practice. However, approaches requiring a partitioned graph usually suffer from communication overhead and load imbalance, even under optimal partitioning and communication strategies due to irregularities in the neighborhood minibatch sampling. This paper proposes practical trade-offs for improving the sampling and communication overheads for representation learn- ing on distributed graphs (using popular GraphSAGE architecture) by developing a parameterized prefetch and eviction scheme on top of the state-of-the-art Amazon DistDGL distributed GNN framework, demonstrating about 15–40% improvement in end-to-end training performance on the NERSC Perlmutter supercomputer for various OGB datasets.

Machine Leanring, high performance comptuing, grap↗

Machine Learning Based Aerosol and Ocean Color Joint Retrieval Algorithm for Multiangle Polarimeters over Coastal Waters

NASA’s Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) mission, recently launched in February 2024, carries two multiangle polarimeters (MAPs): the UMBC Hyper-Angular Rainbow Polarimeter (HARP2) and SRON Spectropolarimeter for Planetary Exploration One (SPEXone). Measurements from these MAPs will greatly advance ocean ecosystem and aerosol studies as their measurements contain rich information on microphysical properties of aerosols and hydrosols. The Multi-Angular Polarimetric Ocean coLor (MAPOL) joint retrieval algorithm has been developed to retrieve aerosol and ocean color information, which uses a vector radiative transfer (RT) model as the forward model. The RT model is computationally expensive, which makes processing a large amount of data challenging. FastMAPOL was developed to expedite retrieval using neural networks to replace the RT forward models. As a prototype study, FastMAPOL was initially limited to open ocean applications where the ocean Inherent Optical Properties (IOPs) were parameterized in terms of one parameter: chlorophyll-a concentration (Chla). In this study we further expand the FastMAPOL joint retrieval algorithm to incorporate NN based forward models for coastal waters, which use multi-parameter bio-optical models. In addition, aerosols are represented by six components, i.e., fine mode non absorbing insoluble (FNAI), brown carbon (BrC), black carbon (BC), fine mode non absorbing soluble (FNAS), sea salt (SS) and non-spherical dust (Dust). Sea salt and dust are coarse mode aerosols, while the other components are in the fine mode. The sizes and spectral refractive indices are fixed for each aerosol component, while their abundances are retrievable. The multi-parameter bio-optical model and aerosol components are chosen to represent the coastal marine environment. The retrieval algorithm is applied to synthetic measurements in three different configurations ofMAPs in the PACE mission: HARP2 observations only, SPEXone observations only and combined HARP2 and SPEXone observations. The retrieval results from synthetic measurements show that for aerosol retrieval the SPEXone-only configuration works equally well with the HAPR2-only configuration. On the other hand, for ocean color retrieval the SPEXone instrument provides better information due to its larger spectral coverage. For the surface parameters (wind speed), HARP2 measurements provide better information due to its wide field of view. Combined measurement configuration HARP2+SPEXone performed the best to retrieve all aerosol, ocean color and surface parameters. We also studied the impact of sun glint to aerosol and ocean color retrievals. The retrieval test revealed that wind speed and absorbing aerosol retrieval improves significantly when including measurements at glint geometries. Furthermore, the retrieval algorithm is equipped with modules for atmospheric correction and bidirectional reflectance distribution (BRDF) correction to obtain the remote sensing reflectance, which enables ocean biogeochemistry studies using the PACE polarimeter data.

PACE↗