Search NASA⌕ Search

SEARCH · Search NASA

Results for “fast 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 217 records · Page 12

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↗

Learning protocols for the fast and efficient control of active matter

Exact analytic calculation shows that optimal control protocols for passive molecular systems often involve rapid variations and discontinuities. However, similar analytic baselines are not generally available for active-matter systems, because it is more difficult to treat active systems exactly. Here we use machine learning to derive efficient control protocols for active-matter systems, and find that they are characterized by sharp features similar to those seen in passive systems. We show that it is possible to learn protocols that effect fast and efficient state-to-state transformations in simulation models of active particles by encoding the protocol in the form of a neural network. We use evolutionary methods to identify protocols that take active particles from one steady state to another, as quickly as possible or with as little energy expended as possible. Our results show that protocols identified by a flexible neural-network ansatz, which allows the optimization of multiple control parameters and the emergence of sharp features, are more efficient than protocols derived recently by constrained analytical methods. Our learning scheme is straightforward to use in experiment, suggesting a way of designing protocols for the efficient manipulation of active matter in the laboratory.

74 ATOMIC AND MOLECULAR PHYSICS↗

Freeform thermoelectrics in single-step manufacturing: additive manufacturing of bismuth-telluride thermoelectrics

The project succeeded in producing crack-free Bismuth Telluride thermoelectric parts with density exceeding 98% through laser powder bed fusion (LPBF) additive manufacturing (AM). This greatly exceeded the highest previously reported density of 88% and is the highest among all semiconducting materials processed by LPBF. The additively manufactured material shows comparable Seebeck coefficient as conventional form and can be made into complex geometries with reduced material loss. On the other hand, measured properties are dramatically sensitive to the AM process parameters used, such that with identical composition, the Seebeck coefficient can be controllably tuned from +120 µV/K to -207 µV/K, which means the material switches between an n-type to a p-type semiconductor depending on processing. These changes are accompanied by significant differences in the as-processed microstructure due to rapid solidification. A machine learning protocol was developed and greatly reduced the experimental burden of the project, reducing the typical process optimization period of 2 years to 6 months. The project was fully successful in the objective of producing defect-free, complex geometry of bismuth-telluride parts through LPBF, but only partially successful in achieving performance goals. First, cost reduction of manufacturing, as measured by material waste, was successfully reduced by up to 70% compared to conventional manufacturing methods. This exceeded the proposed 30% reduction in materials waste needed to reach the 20% cost reduction goal of the project. On the other hand, the device performance, as measured by Seebeck coefficient, failed to reach the 40% improvement in efficiency. Rather, we observe comparable Seebeck coefficient between AM samples and conventionally processed counterparts. The device-level efficiency improvement does exceed 40% for complex geometry samples due to shape-induced increase in temperature gradients, but this was not the originally proposed metric. The machine learning approach developed in this project greatly accelerated the process optimization and can be adopted for fast development of AM processing parameters for other brittle and otherwise difficult-to-print materials. For the public, we deliver an efficient and widely adoptable process for incorporating waste-heat harvesting thermoelectric devices in both industrial and commercial heat exchangers. The geometric flexibility allows the capturing device to conform to the shape of the heat source to improve the system-level conversion efficiency. The technique can be deployed on any commercial LBPF systems with zero modifications, thus poses minimal adoption barrier for any manufacturer that already employed AM technology. Beyond bismuth-telluride, the machine-learning guided optimization protocol can be used in the future to reduce both the time and cost of process development for AM of other energy conversion and harvesting materials.

36 MATERIALS SCIENCE↗

Machine Learning-Based Regression Models for Ironmaking Blast Furnace Automation

Computational fluid dynamics (CFD)-based simulation has been the traditional way to model complex industrial systems and processes. One very large and complex industrial system that has benefited from CFD-based simulations is the steel blast furnace system. The problem with the CFD-based simulation approach is that it tends to be very slow for generating data. The CFD-only approach may not be fast enough for use in real-time decisionmaking. To address this issue, in this work, the authors propose the use of machine learning techniques to train and test models based on data generated via CFD simulation. Regression models based on neural networks are compared with tree-boosting models. In particular, several areas (tuyere, raceway, and shaft) of the blast furnace are modeled using these approaches. The results of the model training and testing are presented and discussed. The obtained R 2 metrics are, in general, very high. The results appear promising and may help to improve the efficiency of operator and process engineer decisionmaking when running a blast furnace.

97 MATHEMATICS AND COMPUTING↗

Measuring the electron temperature and identifying plasma detachment using machine learning and spectroscopy

A machine learning approach has been implemented to measure the electron temperature directly from the emission spectra of a tokamak plasma. This approach utilized a neural network (NN) trained on a dataset of 1865 time slices from operation of the DIII-D tokamak using extreme ultraviolet/vacuum ultraviolet emission spectroscopy matched with high-accuracy divertor Thomson scattering measurements of the electron temperature, T e . This NN is shown to be particularly good at predicting T e at low temperatures (T e < 10 eV) where the NN demonstrated a mean average error of less than 1 eV. Trained to detect plasma detachment in the tokamak divertor, a NN classifier was able to correctly identify detached states (T e < 5 eV) with a 99% accuracy (an F 1 score of 0.96) at an acquisition rate 10× faster than the Thomson scattering measurement. The performance of the model is understood by examining a set of 4800 theoretical spectra generated using collisional radiative modeling that was also used to predict the performance of a low-cost spectrometer viewing nitrogen emission in the visible wavelengths. Furthermore, these results provide a proof-of-principle that low-cost spectrometers leveraged with machine learning can be used to boost the performance of more expensive diagnostics on fusion devices and be used independently as a fast and accurate T e measurement and detachment classifier.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Advanced characterization-informed machine learning framework and quantitative insight to irradiated annular U-10Zr metallic fuels

Abstract U-10Zr Metal fuel is a promising nuclear fuel candidate for next-generation sodium-cooled fast spectrum reactors. Since the Experimental Breeder Reactor-II in the late 1960s, researchers accumulated a considerable amount of experience and knowledge on fuel performance at the engineering scale. However, a mechanistic understanding of fuel microstructure evolution and property degradation during in-reactor irradiation is still missing due to a lack of appropriate tools for rapid fuel microstructure assessment and property prediction based on post irradiation examination. This paper proposed a machine learning enabled workflow, coupled with domain knowledge and large dataset collected from advanced post-irradiation examination microscopies, to provide rapid and quantified assessments of the microstructure in two reactor irradiated prototypical annular metal fuels. Specifically, this paper revealed the distribution of Zr-bearing secondary phases and constitutional redistribution across different radial locations. Additionally, the ratios of seven different microstructures at various locations along the temperature gradient were quantified. Moreover, the distributions of fission gas pores on two types of U-10Zr annular fuels were quantitatively compared.

36 MATERIALS SCIENCE↗

A Machine Learning–Based Approach to Time-series Wave Identification in the Solar Wind

The Wind spacecraft has yielded several decades of high-resolution magnetic field data, a large fraction of which displays small-scale structures. In particular, the solar wind is full of wavelike fluctuations that appear in both the field magnitude and its components. The nature of these fluctuations can be tied to the properties of other structures in the solar wind, such as shocks, that have implications for the time evolution of the solar wind. As such, having a large collection of wave events would facilitate further study of the effects that these fluctuations have on solar wind evolution. Given the large volume of magnetic field data available, machine learning is the most practical approach to classifying the myriad small-scale structures observed. To this end, a subset of Wind data is labeled and used as a training set for a multibranch 1D convolutional neural network aimed at classifying circularly polarized wave modes. Using this algorithm, a preliminary statistical study of 1 yr of data is performed, yielding about 300,000 wave intervals out of about 5,000,000 solar wind intervals. The wave intervals come about more often in the fast solar wind and at higher temperatures, and the number of waves per day is highly periodic. This machine learning–based approach to wave detection has the potential to be a powerful, inexpensive way to catalog waves throughout decades of spacecraft data.

79 ASTRONOMY AND ASTROPHYSICS↗

Fast Fault Injection and Sensitivity Analysis for Collective Communications, In: 2015 IEEE International Conference on Cluster Computing

The collective communication operations, which are widely used in parallel applications for global communication and synchronization are critical for application's performance and scalability. However, how faulty collective communications impact the application and how errors propagate between the application processes is largely unexplored. One of the critical reasons for this situation is the lack of fast evaluation method to investigate the impacts of faulty collective operations. The traditional random fault injection methods relying on a large amount of fault injection tests to ensure statistical significance require a significant amount of resources and time. These methods result in prohibitive evaluation cost when applied to the collectives. In this paper, we introduce a novel tool named Fast Fault Injection and Sensitivity Analysis Tool (FastFIT) to conduct fast fault injection and characterize the application sensitivity to faulty collectives. The tool achieves fast exploration by reducing the exploration space and predicting the application sensitivity using Machine Learning (ML) techniques. A basis for these techniques are implicit correlations between MPI semantics, application context, critical application features, and application responses to faulty collective communications. The experimental results show that our approach reduces the fault injection points and tests by 97% for representative benchmarks (NAS Parallel Benchmarks (NPB)) and a realistic application (Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS)) on a production supercomputer. Further, we statistically generalize the application sensitivity to faulty collective communications for these workloads, and present correlation between application features and the sensitivity.

Computer Science↗

Accelerated Machine Learning as a Service for Particle Physics Computing

Accelerated Machine Learning as a Service for Particle Physics Computing: • Amount and complexity of high-energy physics data increases dramatically from 2025 onward • Traditional algorithms will require too much CPU time • Machine learning can solve combinatorially-scaling problems in constant time, but must be fast enough

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Automated Dynamic Demand Response Implementation on a Micro-grid

In this paper, we describe a system for real-time automated Dynamic and Sustainable Demand Response with sparse data consumption prediction implemented on the University of Southern California campus microgrid. Supply side approaches to resolving energy supply-load imbalance do not work at high levels of renewable energy penetration. Dynamic Demand Response (D 2 R) is a widely used demand-side technique to dynamically adjust electricity consumption during peak load periods. Our D 2 R system consists of accurate machine learning based energy consumption forecasting models that work with sparse data coupled with fast and sustainable load curtailment optimization algorithms that provide the ability to dynamically adapt to changing supply-load imbalances in near real-time. Our Sustainable DR (SDR) algorithms attempt to distribute customer curtailment evenly across sub-intervals during a DR event and avoid expensive demand peaks during a few sub-intervals. It also ensures that each customer is penalized fairly in order to achieve the targeted curtailment. We develop near linear-time constant-factor approximation algorithms along with Polynomial Time Approximation Schemes (PTAS) for SDR curtailment that minimizes the curtailment error defined as the difference between the target and achieved curtailment values. Our SDR curtailment problem is formulated as an Integer Linear Program that optimally matches customers to curtailment strategies during a DR event while also explicitly accounting for customer strategy switching overhead as a constraint. We demonstrate the results of our D 2 R system using real data from experiments performed on the USC smartgrid and show that 1) our prediction algorithms can very accurately predict energy consumption even with noisy or missing data and 2) our curtailment algorithms deliver DR with extremely low curtailment errors in the 0.01-0.05 kWh range.

Kuppannagari, Sanmukh R.↗

Predicting resistive wall mode stability in NSTX through balanced random forests and counterfactual explanations

Abstract Recent progress in the disruption event characterization and forecasting framework has shown that machine learning guided by physics theory can be easily implemented as a supporting tool for fast computations of ideal stability properties of spherical tokamak plasmas. In order to extend that idea, a customized random forest (RF) classifier that takes into account imbalances in the training data is hereby employed to predict resistive wall mode (RWM) stability for a set of high beta discharges from the NSTX spherical tokamak. More specifically, with this approach each tree in the forest is trained on samples that are balanced via a user-defined over/under-sampler. The proposed approach outperforms classical cost-sensitive methods for the problem at hand, in particular when used in conjunction with a random under-sampler, while also resulting in a threefold reduction in the training time. In order to further understand the model’s decisions, a diverse set of counterfactual explanations based on determinantal point processes (DPP) is generated and evaluated. Via the use of DPP, the underlying RF model infers that the presence of hypothetical magnetohydrodynamic activity would have prevented the RWM from concurrently going unstable, which is a counterfactual that is indeed expected by prior physics knowledge. Given that this result emerges from the data-driven RF classifier and the use of counterfactuals without hand-crafted embedding of prior physics intuition, it motivates the usage of counterfactuals to simulate real-time control by generating the β N levels that would have kept the RWM stable for a set of unstable discharges.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Reactive Transport Modeling with Physics-Informed Machine Learning for Critical Minerals Applications

This study presents a physics-informed neural network (PINN) framework for reactive transport modeling for simulating fast bimolecular reactions in porous media. Accurate characterization of cAhemical interactions and product formation in surface and subsurface environments is essential for advancing critical mineral extraction and related geoscience applications. The proposed methodology sequentially addresses the flow and diffusion–reaction subproblems. The flow field is computed using a mixed formulation, while the diffusion–reaction system is modeled via two uncoupled tensorial diffusion equations reformulated in terms of chemical invariants. PINNs are employed to solve the governing equations, enabling data-efficient, mesh-free prediction of chemical concentration fields. The framework is validated through a series of benchmark problems involving flow in heterogeneous porous media. Initial verification is conducted using patch tests for the flow field, followed by validation of the transport problem with emphasis on preserving non-negativity of concentrations. The complete fast bimolecular reaction scenario is then solved, yielding spatial distributions of reactants and product species. Results demonstrate that the PINNs-based approach effectively captures sharp, mixing-limited reaction fronts and dispersive mixing behavior, offering reliable predictions of reactive plume evolution. These capabilities are crucial for evaluating long-term subsurface behavior in applications such as fluid storage, energy extraction, and efficient extraction of critical minerals.

42 ENGINEERING↗

A Machine Learning-Based Approach to Time-Series Wave Identification in the Solar Wind

The Wind spacecraft has yielded several decades of high-resolution magnetic field data, a large fraction of which displays small-scale structures. In particular, the solar wind is full of wavelike fluctuations that appear in both the field magnitude and its components. The nature of these fluctuations can be tied to the properties of other structures in the solar wind, such as shocks, that have implications for the time evolution of the solar wind. As such, having a large collection of wave events would facilitate further study of the effects that these fluctuations have on solar wind evolution. Given the large volume of magnetic field data available, machine learning is the most practical approach to classifying the myriad small-scale structures observed. To this end, a subset of Wind data is labeled and used as a training set for a multi-branch 1D convolutional neural network aimed at classifying circularly polarized wave modes. Using this algorithm, a preliminary statistical study of one year of data is performed, yielding about 300,000 wave intervals out of about 5,000,000 solar wind intervals. The wave intervals come about more often in the fast solar wind and at higher temperatures, and the number of waves per day is highly periodic. This machine learning-based approach to wave detection has the potential to be a powerful, inexpensive way to catalog waves throughout decades of spacecraft data.

Samuel Fordin↗

Fast Adaptive Neural Control of Resonant Extraction at Fermilab

We present progress on the development of a machine learning (ML) regulation system for third-order resonant extraction of the beam delivered to the Mu2e experiment at Fermilab. We consider classical and ML-based controllers optimized on semi-analytic simulations and provide performance comparisons for several models. Additionally, we discuss the efficiency of each model in training, which has implications for future work on adaptive control. We also discuss progress on developing optimized implementations of ML models for edge-based inference.

Whitbeck, A. [Fermilab] (ORCID:0000000342245164)↗

Fast Adaptive Neural Control of Resonant Extraction at Fermilab

We present progress on the development of a machine learning (ML) regulation system for third-order resonant extraction of the beam delivered to the Mu2e experiment at Fermilab. We consider classical and ML-based controllers optimized on semi-analytic simulations and provide perfor- mance comparisons for several models. Additionally, we discuss the efficiency of each model in training, which has implications for future work on adaptive control. We also discuss progress on developing optimized implementations of ML models for edge-based inference.

Whitbeck, A. [Fermilab]↗

Upgrading the Fast Calorimeter Simulation in ATLAS

The tremendous need for simulated samples now and even more so in the future, encourages the development of fast simulation techniques. The Fast Calorimeter Simulation is a faster though less accurate alternative to the full calorimeter simulation with GEANT4. It is based on parametrizing the longitudunal and lateral energy deposits of single particles in the ATLAS calorimeter. Principal component analysis and machine learning techniques are used to improve the performance and decrease the memory need compared to the current version of the ATLAS Fast Calorimeter Simulation. The parametrizations are expanded to cover very high energies and very forward detector regions, to increase the applicability of the tool. A prototype of this upgraded Fast Calorimeter Simulation has been developed and first validations with single particles show substantial improvements over the previous version.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Towards accelerating particle-resolved direct numerical simulation with neural operators

In this paper, we present our ongoing work aimed at accelerating a particle-resolved direct numerical simulation model designed to study aerosol–cloud–turbulence interactions. The dynamical model consists of two main components—a set of fluid dynamics equations for air velocity, temperature, and humidity, coupled with a set of equations for particle (i.e., cloud droplet) tracing. Rather than attempting to replace the original numerical solution method in its entirety with a machine learning (ML) method, we consider developing a hybrid approach. We exploit the potential of neural operator learning to yield fast and accurate surrogate models and, in this study, develop such surrogates for the velocity and vorticity fields. We discuss results from numerical experiments designed to assess the performance of ML architectures under consideration as well as their suitability for capturing the behavior of relevant dynamical systems.

54 ENVIRONMENTAL SCIENCES↗

Challenging Practices of Algebraic Battery Life Models through Statistical Validation and Model Identification via Machine-Learning

Various modeling techniques are used to predict the capacity fade of Li-ion batteries. Algebraic reduced-order models, which are inherently interpretable and computationally fast, are ideal for use in battery controllers, technoeconomic models, and multi-objective optimizations. For Li-ion batteries with graphite anodes, solid-electrolyte-interphase (SEI) growth on the graphite surface dominates fade. This fade is often modeled using physically informed equations, such as square-root of time for predicting solvent-diffusion limited SEI growth, and Arrhenius and Tafel-like equations predicting the temperature and state-of-charge rate dependencies. In some cases, completely empirical relationships are proposed. However, statistical validation is rarely conducted to evaluate model optimality, and only a handful of possible models are usually investigated. This article demonstrates a novel procedure for automatically identifying reduced-order degradation models from millions of algorithmically generated equations via bi-level optimization and symbolic regression. Identified models are statistically validated using cross-validation, sensitivity analysis, and uncertainty quantification via bootstrapping. On a LiFePO 4 /Graphite cell calendar aging data set, automatically identified models utilizing square-root, power law, stretched exponential, and sigmoidal functions result in greater accuracy and lower uncertainty than models identified by human experts, and demonstrate that previously known physical relationships can be empirically "rediscovered" using machine learning.

25 ENERGY STORAGE↗