Search NASA⌕ Search

SEARCH · Search NASA

Results for “machine learning operations”

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 1,801 records · Page 100

Supporting 64-bit global indices in Epetra and other Trilinos packages :

The Trilinos Project is an effort to facilitate the design, development, integration and ongoing support of mathematical software libraries within an object-oriented framework. It is intended for large-scale, complex multiphysics engineering and scientific applications [2, 4, 3]. Epetra is one of its basic packages. It provides serial and parallel linear algebra capabilities. Before Trilinos version 11.0, released in 2012, Epetra used the C++ int data-type for storing global and local indices for degrees of freedom (DOFs). Since int is typically 32-bit, this limited the largest problem size to be smaller than approximately two billion DOFs. This was true even if a distributed memory machine could handle larger problems. We have added optional support for C++ long long data-type, which is at least 64-bit wide, for global indices. To save memory, maintain the speed of memory-bound operations, and reduce further changes to the code, the local indices are still 32-bit. We document the changes required to achieve this feature and how the new functionality can be used. We also report on the lessons learned in modifying a mature and popular package from various perspectives design goals, backward compatibility, engineering decisions, C++ language features, effects on existing users and other packages, and build integration.

Jhurani, Chetan↗

Physics-Informed Neural Network (PINN) Prediction of Mixed Mass-Heat-Crystallization Limited Methane Hydrate Formation and Dissociation in Micro-Confinement

The creation and use of Physics-Informed Neural Networks (PINNs) for simulating the dynamics of methane hydrate formation and dissociation will be presented. The PINN framework's main benefit is its capacity to impose physical consistency with only a partial comprehension of the governing equations. This makes the algorithm especially useful for systems with little experimental evidence or a lack of theoretical knowledge. A strong basis for forecasting methane hydrate behavior over the verified operating ranges of 30.0-80.9 bar pressure and 1.0-4.0 K sub-cooling conditions is provided by the combination of conductive heat transfer equations and mixed mass-transfer–crystallization kinetics. PINNs were more accurate at predicting the mixed mass-heat-crystallization limited kinetics than conventional Artificial Neural Networks (ANNs), demonstrating remarkable predictive accuracy for methane hydrate production over the ANN model. The efficiency of incorporating physical limitations from first principles into machine learning frameworks for methane hydrate crystallizations is reinforced by these findings. For hydrate-related applications in energy generation, carbon sequestration, and climate modelling, our study establishes PINNs as a computational tool that is both scalable and efficient. The proven capacity to close the gap between conventional physics-based simulations and solely data-driven models creates new opportunities for expedited hydrate research and practical applications.

Hartman, Ryan L [NYU Tandon School of Engineering]↗

Probing the Sun with Imaging Spectrographs

EUV and X-ray images of the Sun have revolutionized our understanding of our closest star. With them, we can probe the morphology and temperature structure of the solar atmosphere and see how they evolve as a function of space and time. However, image data cannot be used to determine line-of-sight velocities, abundances, or densities. This information is required to calculate the energy budget of eruptive events, provide boundary conditions for global solar models, and explore fundamental processes occurring in the solar atmosphere. For those diagnostics, we require spectroscopy. Because the structures on the Sun are extended sources, most modern-day spectrographs observe the Sun through long narrow slits. Two-dimensional, spectrally pure solar images with velocity, abundance, and density information are built up by stepping the slit over regions of interest. This mode of operation implies that two-dimensional information is highly limited by cadence and the temporal evolution and spatial structure of these parameters can never be truly separated. Both spatial and spectral information can be obtained in a single snapshot with slitless spectrographs, which were often used in the 1950-1970s, but were abandoned due to the difficulty of unfolding the overlapping spatial and spectral information. Thanks to advances in computer processing speeds and machine learning algorithms, there have been several techniques developed to complete the spatial/spectral unfolding, unlocking the full capability of slitless spectrographs for solar observations. The goal of this talk is to give a broad overview of the capability of such instruments and demonstrate their usefulness in the next decade of solar observatories and beyond.

Amy Winebarger↗

The AP1000{sup R} China projects move forward to construction completion and equipment installation

The AP1000 design is the only Generation III+ technology to receive design certification from the U.S. Nuclear Regulatory Commission. This evolutionary design provides the highest safety and performance standards and has several distinct advantages over other designs, including improved operations and reduced construction schedule risks through the use of modern, modular, engineering principles that allow construction and fabrication tasks traditionally performed in sequence to be undertaken in parallel. Since the first granting of Design Certification in 2005 by the NRC, the AP1000 design has been modified to meet emergent NRC requirements such as those requiring the design to withstand the impact of an aircraft crash. Both domestic and foreign utilities have turned to the Westinghouse AP1000 plant design to meet their near - and long-term sustainable energy needs. The first ever deployment of this advanced U.S. nuclear power technology began in China in 2007 with the award of a contract to build four AP1000 units, constructed in pairs at the coastal sites of Sanmen (Zhejiang Province) and Haiyang (Shandong Province). Currently, all four units are at an advanced stage of construction. The commercial operation date for Sanmen Unit 1 is November 2013 followed by Haiyang Unit 1 being operational in May 2014. Construction and equipment manufacture is at an advanced stage. Sanmen Unit 1 equipment that has been delivered includes the reactor vessel, the reactor vessel closure head, the passive residual heat removal heat exchanger, the integrated head package, the polar crane, and the refueling machine. The steam generators are also completed. The RV was installed within the containment vessel building in September 2011. The installation of this major equipment will allow the setting of the containment vessel top head. Haiyang Unit 1 is also achieving significant progress. Significant benefits continue to be realized as a result of lessons learned and experience gained from the first-of-a-kind activities for Sanmen Unit 1 and AP1000 equipment design and manufacturing. For example, the nuclear island basemat at Haiyang Units 1 and 2 and Sanmen Unit 2 was laid in less time than that of Sanmen Unit 1, the ultra-large steam generator and RV forging lead times were reduced for the follow on units, and the fabrication of the auxiliary building module for Haiyang Unit 1 took less time than for the Sanmen first unit. These benefits are also being realized by the United States AP1000 project construction and fabrication modules, and equipment. Some difficulties arise from building this first-of-a-kind (advanced passive) type of plant; however, these difficulties are being overcome and the overall schedule remains achievable. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Subsurface Characterization and Machine Learning Predictions at Brady Hot Springs: Preprint

Subsurface data analysis, reservoir modeling, and machine learning (ML) techniques have been applied to the Brady Hot Springs (BHS) geothermal field in Nevada, USA to further characterize the subsurface and assist with optimizing reservoir management. Hundreds of reservoir simulations have been conducted in TETRAD-G and CMG STARS to explore different injection and production fluid flow rates and allocations and to develop a training data set for ML. This process included simulating the historical injection and production since 1979 and prediction of future performance through 2040. ML networks were created and trained using TensorFlow based on multilayer perceptron (MLP), long short-term memory (LSTM), and convolutional neural network (CNN) architectures. These networks took as input selected flow rates, injection temperatures, and historical field operation data and produced estimates of future production temperatures. This approach was first successfully tested on a simplified single fracture doublet system, followed by the application to the BHS reservoir. Using an initial BHS dataset with 37 simulated scenarios, the trained and validated network predicted the production temperature for 6 production wells with the mean absolute percentage error of less than 8%. In a complementary analysis effort, the principal component analysis applied to 13 BHS geological parameters revealed that vertical fracture permeability shows the strongest correlation with fault density and fault intersection density. A new BHS reservoir model was developed considering the fault intersection density as proxy for permeability. This new reservoir model helps to explore under-exploited zones in the reservoir. A data gathering plan to obtain additional subsurface data was developed; it includes temperature surveying for three idle injection wells, at which the reservoir simulations indicate high bottom-hole temperatures. The collected data assist with calibrating the reservoir model and may lead to converting these wells to producers to access under-exploited zones in the reservoir. Data gathering activities are planned for the first quarter of 2021.

40 EE - Geothermal Technologies Office (EE-4G)↗

Predicting Storm Outages Through New Representations of Weather and Vegetation

This paper presents new developments in an outage prediction model (OPM) for an electric distribution network in the Northeastern United States and assesses their significance to the OPM performance. The OPM uses regression tree models fed by numerical weather prediction outputs, spatially distributed information on soil, vegetation, electric utility assets, and historical power outage data to forecast the number and spatial distribution of outages across the power distribution grid. New modules introduced hereby consist in 1) a storm classifier based on weather variables; 2) a multimodel optimization of regression tree output; and 3) a post-processing routine for more accurately describing tree-leaf conditions. Model implementations are tested through leave-one-storm-out cross-validations performed on 120 storms of varying intensity and characteristics. The results demonstrate that the median absolute percentage error of the new OPM version decreased from 130% to 59% for outage predictions at the service territory level, and the OPM skills for operational forecasts are consistent with the skills based on historical storm analyses.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Safe Reinforcement Learning for Emergency Load Shedding of Power Systems

The paradigm shift in the electric power grid necessitates a revisit of existing control methods to ensure the grid’s security and resilience. In particular, the increased uncertainties and rapidly changing operational conditions in power systems have revealed outstanding issues in terms of either speed, adaptiveness, or scalability of the existing control methods for power systems. On the other hand, the availability of massive real-time data can provide a clearer picture of what is happening in the grid. Recently, deep reinforcement learning (RL) has been regarded and adopted as a promising approach leveraging massive data for fast and adaptive grid control. However, like most existing machine learning (ML)- based control techniques, RL control usually cannot guarantee the safety of the power systems. In this paper, we introduce a novel method for safe RL-based load shedding of power systems that can enhance the safe voltage recovery of the electric power grid after experiencing faults. Numerical simulation on the IEEE 39-bus testcase is performed to demonstrate the effectiveness of the proposed safe RL emergency control, as well as its adaptive capability to faults not seen in the training.

reinforcement learning, emergency control, power g↗

Assimilation of Multiscale Data into Multifidelity Biogeochemical Models (Final Report)

Quantitative predictions of subsurface processes rely on computational models that capture, with different degrees of fidelity, complex interactions between hydrologic and biogeochemical processes. Molecular- and pore-scale models provide a high-fidelity representation of these processes but are impractical at the field scale. Reduced complexity (e.g., field-scale or data-driven) models sacrifice some degree of fidelity in favor of computational efficiency. Uncertainty and assimilation of data into model predictions, pose a question of model selection: Given a significant difference in computational cost, can a lower-fidelity model be preferable to its higher-fidelity counterpart? Since a predictive model must be computable in reasonable time it has to operate at the field scale, with molecular- and pore-scale data (obtained either from simulations or measurements) determining both the model's structure and parameters. Availability of such multi-resolution data raises a question hitherto undressed in hydrogeology: Can a coarse model dynamically ``learn'' its own structure (e.g., adjust the reaction pathways in its transport module) as more fine-scale data become available during simulations? Our results in multifidelity simulations, data assimilation, and machine learning led us to hypothesize that the answer to these questions is ``yes''. The overarching goal of this project was to confirm this hypothesis by developing scalable, computationally efficient tools for assimilation of data into multiscale models, in which fine-scale data and simulations dynamically inform and autonomously modify coarse-scale models. Our research led to nine manuscripts, three of which have been published and the other six are currently under review.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Final Report in Response to ARPA-E Contract Award Number DE-AR0001157

Autonomous operations and maintenance (O&M) by robots has been identified as a key technology to facilitate both the safe operation, and reduced operational costs associated with future nuclear reactors. Radiation and thermal environments, particularly associated with molten salt reactor designs, excludes the use of human proximity and so there is a need to train robots for tasks that have not yet been fully identified, or for accidents that may occur in the future. This program sought to develop a generic methodology that could be applied to train ‘any’ robot, to perform ‘any’ task, through the use of machine learning (ML) training performed in a virtual reality (VR) environment that simulates the physically perceived task(s). The VR environment allows us to construct ‘any’ future task and the ML approach, which included reinforcement learning (RL), allowed us to generate extensive data sets that can be used to establish control algorithms to thereby control the physical robot in the physical environment.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Hazard Detection Detector Cards

This report presents a comprehensive summary of five advanced anomaly detection tools developed and deployed by Oak Ridge National Laboratory in support of the VA’s Health Information Technology modernization. These detectors—Order Path Tracker, Trend Watcher, Pain Pointer, Performance Monitor, and Patient Record Flag Detector—leverage statistical and machine learning methods to monitor workflow disruptions, detect anomalies in care sequences and volumes, identify bottlenecks, and track system-level performance metrics across VistA and Millennium systems. All detectors have been integrated into the Health Data Analytics Platform (HDAP), with most having completed deployment and testing using live data from targeted stations in cardiology and oncology domains. This work enhances VA’s capacity for proactive system surveillance, promotes patient safety, and informs data-driven operational improvements across the EHR ecosystem.

97 MATHEMATICS AND COMPUTING↗

On-chip probabilistic inference for charged-particle tracking at the sensor edge

Modern scientific instruments operate under increasingly extreme constraints on bandwidth, latency, and power. Inference at the sensor edge determines experimental data collection efficiency by deciding which information to save for further analysis. Particle tracking detectors at the Large Hadron Collider exemplify this challenge: pixelated silicon sensors generate rich spatiotemporal ionization patterns, yet most of this information is discarded due to data-rate limitations. Concurrently, advancements in co-design tools provide rapid turn-around for incorporating machine learning into application-specific integrated circuits, motivating designs for particle detectors with new integrated technologies. We demonstrate that neural networks embedded in the front-end electronics can infer charged-particle kinematic parameters from a single silicon layer. We regress hit positions and incident angles with calibrated uncertainties, while satisfying stringent constraints on numerical precision, latency, and silicon area. Our results establish a path toward probabilistic inference directly at the edge, opening new opportunities for intelligent sensing in high-rate scientific instruments.

Das, Arghya Ranjan [Purdue U.] (ORCID:000000018451↗

Scaling Field-Theoretic Simulation for Multicomponent Mixtures with Neural Operators

Multicomponent polymer mixtures are ubiquitous in biological self-organization but are notoriously difficult to study computationally. Plagued by both slow single molecule relaxation times and slow equilibration within dense mixtures, molecular dynamics simulations are typically infeasible at the spatial scales required to study the stability of mesophase structure. Polymer field theories offer an attractive alternative, but analytical calculations are only tractable for mean-field theories and nearby perturbations, constraints that become especially problematic for fluctuation-induced effects such as coacervation. Here, we show that a recently developed technique for obtaining numerical solutions to partial differential equations based on operator learning, neural operators, lends itself to a highly scalable training strategy by parallelizing per-species operator maps. We illustrate the efficacy of our approach on six-component mixtures with randomly selected compositions and that it significantly outperforms the state-of-the-art pseudospectral integrators for field-theoretic simulations, especially as polymer lengths become long.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

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↗

Energy-based descriptors to rapidly predict hydrogen storage in metal–organic frameworks

The low volumetric density of hydrogen is a critical limitation to its use as a transportation fuel. Filling a fuel tank with nanoporous materials, such as metal–organic frameworks (MOFs), could greatly improve the deliverable capacity of these tanks if appropriate materials could be found. Yet, since MOFs can be made from many combinations of metal nodes, organic linkers, and functional groups, the design space of possible MOFs is enormous. Experimental characterization of thousands of MOFs is infeasible, and even conventional molecular simulations can be prohibitively expensive for large databases. In this work, we have developed a data-driven approach to accelerate materials screening and learn structure–property relationships. We report new descriptors for gas adsorption in MOFs derived from the energetics of MOF–guest interactions. Using the bins of an energy histogram as features, we trained a sparse regression model to predict gas uptake in multiple MOF databases to an accuracy within 3 g L -1 . The interpretable model parameters indicate that a somewhat weak attraction between hydrogen and the framework is ideal for cryogenic storage and release. Our machine learning method is more than three orders of magnitude faster than conventional molecular simulations, enabling rapid exploration of large numbers of MOFs. As a case study, we applied the method to screen a database of more than 50 000 experimental MOF structures. We experimentally validated one of the top candidates identified from the accelerated screening, MFU-4l. This material demonstrated a hydrogen deliverable capacity of 47 g L -1 (54 g L -1 simulated) when operating at storage conditions of 77 K, 100 bar and delivery at 160 K, 5 bar.

36 MATERIALS SCIENCE↗

Physics-Informed AI for Climate and Weather Risk Prediction (Final Report)

As part of the work Terrafuse developed 1) a model for wildfire risk in California, and 2) a model for downscaled wind fields from Numerical Weather Prediction (NWP) numerical models. The wildfire model is based on 20 years of historical data and captures the dependence of wildfire incidence and spread on climatic, weather and land-use variables by training a machine learning model. Nonlinear relationships between input features are learned and expressed by the model and model transparency allows features to be ranked and interpreted. The fire model is of use for accurately predicting real-time daily and long-term wildfire risk for use cases in energy and insurance. The downscaled wind model is a spatio-temporal deep learning model that emulates the influence of high-resolution variables on wind speed, allowing coarse-resolution operational NWP models to be accurately expressed on fine grids at high resolution, with application to wind energy and weather forecasting.

54 ENVIRONMENTAL SCIENCES↗

Transforming Energy Through Computational Excellence: A Virtual Test-Drive of Energy Scenarios

Researchers at the National Renewable Energy Laboratory (NREL) are conducting groundbreaking research to decarbonize transportation systems and reduce emissions. As they plan for growing populations and related demands on aging infrastructure and energy systems, local transportation departments and city planners are facing significant, complex challenges. New approaches to planning and operations, as well as capitalizing on rapidly advancing technology, can address mobility and energy-centric challenges, including cost and energy efficiency, emissions reduction, decarbonization, and congestion.

33 ADVANCED PROPULSION SYSTEMS↗

Computational Tools and Workflows for Quantitative Risk Assessment and Decision Support for Geologic Carbon Storage Sites: Progress and Insights from the U.S. DOE’s National Risk Assessment Partnership

The 2005 Intergovernmental Panel on Climate Change (IPCC) Special Report on CCS raised the profile of CO2 capture and storage (CCS) as an important technology for reducing greenhouse gas (GHG) emissions. CCS is now recognized as a key component of most climate change mitigation scenarios. Since publication of that report the international research, development, and deployment (RD&D) community has advanced key technical aspects, clarified regulatory requirements, explored value chain and infrastructure solutions, and developed incentive paradigms to enable and promote large-scale deployment of CCS. These efforts have included research to better characterize geologic storage resources, to improve injection performance and storage efficiency, to assess and manage subsurface environmental risks, and to advance monitoring technologies to assure system conformance. These efforts have helped to build confidence in the viability of geologic carbon storage (GCS), but stakeholder concerns about long-term risks and liability associated with GCS remain a hurdle to broad acceptance and large-scale deployment of CCS. Since 2010, the U.S. DOE’s National Risk Assessment Partnership (NRAP) – a research collaboration between five contributing national laboratories – has worked to establish and demonstrate methods and tools to quantify and manage the subsurface environmental risks associated with GCS, amidst uncertainty. This work supports the Office of Fossil Energy and Carbon Management Carbon Transport and Storage Program’s goal of advancing safe and secure commercial-scale GCS deployment. To address the technical challenge of simulating the physical response of the GCS site to large-scale CO2 injection, NRAP has adopted an approach that relies on coupling computationally efficient reduced-order and/or data-driven proxy models of important system components (i.e., storage reservoir, sealing caprock, leakage pathways, intermediate formations, overlying groundwater aquifers, and the atmosphere) in integrated assessment framework. That integrated model of the physical system is complemented with fit-for purpose functionality to support site characterization and risk-related decisions. The recently released NRAP Phase II toolset includes the Open-Source Integrated Assessment Model (NRAP-Open-IAM) for evaluation of trends in leakage risk and potential impact, tools to support monitoring design optimization (Designs for Risk Evaluation and Management – DREAM v3.0 and Passive Seismic Monitoring Tool - PSMT), and tools for state of stress evaluation (State-of-Stress Analysis Tool - SOSAT) and forecasting induced seismicity risk. The NRAP team has also released a pair of reports describing conceptual workflows to incorporate physics-based, quantitative risk assessment into many of the design, planning, operation, and closure decisions for GCS projects. An online catalogue highlights published studies where these tools and methods are demonstrated. In this presentation, the utility of these products to assess risks and address key stakeholder questions will be highlighted through examples, and related insights about the safety and security of geologic carbon storage in qualified storage sites will be discussed. The prospect of rapid, large-scale deployment of GCS technology to aggressively reduce anthropogenic CO2 emissions requires careful consideration of interference between multiple commercial-scale storage projects within a basin. Going forward, NRAP is expanding and adapting site-scale risk quantification tools and methods to enable assessment of risks and inform management decisions for basin-scale deployment. Increasingly, this work will leverage next-generation approaches for surrogate modelling, fast prediction, and advanced visualization enabled by machine learning and artificial intelligence to promote virtual learning, scenario evaluation, and augment risk-based decision making.

quantitative risk assessment, geologic carbon stor↗

Real-Time Edge AI for Distributed Systems (READS): Progress on Beam Loss De-Blending for the Fermilab Main Injector and Recycler

The Fermilab Main Injector enclosure houses two accelerators, the Main Injector and Recycler. During normal operation, high intensity proton beams exist simultaneously in both. The two accelerators share the same beam loss monitors (BLM) and monitoring system. Beam losses in the Main Injector enclosure are monitored for tuning the accelerators and machine protection. Losses are currently attributed to a specific machine based on timing. However, this method alone is insufficient and often inaccurate, resulting in more difficult machine tuning and unnecessary machine downtime. Machine experts can often distinguish the correct source of beam loss. This suggests a machine learning (ML) model may be producible to help de-blend losses between machines. Work is underway as part of the Fermilab Real-time Edge AI for Distributed Systems Project (READS) to develop a ML empowered system that collects streamed BLM data and additional machine readings to infer in real-time, which machine generated beam loss.

43 PARTICLE ACCELERATORS↗

Refine your search to explore more results.