Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “neural operator 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 685 records · Page 38

Fault Injection for TensorFlow Applications

As machine learning (ML) has seen increasing adoption in safety-critical domains (e.g., autonomous vehicles), the reliability of ML systems has also grown in importance. While prior studies have proposed techniques to enable efficient error-resilience (e.g., selective instruction duplication), a fundamental requirement for realizing these techniques is a detailed understanding of the application’s resilience. In this work, we present TensorFI 1 and TensorFI 2, high-level fault injection (FI) frameworks for TensorFlow-based applications. TensorFI 1 and 2 are able to inject both hardware and software faults in any general TensorFlow 1 and 2 program respectively. Both are configurable FI tools that are flexible, easy to use, and portable. They can be integrated into existing TensorFlow programs to assess their resilience for different fault types (e.g., bit-flips in particular operations or layers). We use the TensorFI 1 and TensorFI 2 to evaluate the resilience of 12 and 10 ML programs written in TensorFlow, including DNNs used in the autonomous vehicle domain. The results give us insights into why some of the models are more resilient. We also measure the performance overheads of the two injectors, and present 4 case studies, two for each tool, to demonstrate their utility.

97 MATHEMATICS AND COMPUTING↗

Inexact iterative numerical linear algebra for neural network-based spectral estimation and rare-event prediction

Understanding dynamics in complex systems is challenging because there are many degrees of freedom, and those that are most important for describing events of interest are often not obvious. The leading eigenfunctions of the transition operator are useful for visualization, and they can provide an efficient basis for computing statistics, such as the likelihood and average time of events (predictions). Here, we develop inexact iterative linear algebra methods for computing these eigenfunctions (spectral estimation) and making predictions from a dataset of short trajectories sampled at finite intervals. We demonstrate the methods on a low-dimensional model that facilitates visualization and a high-dimensional model of a biomolecular system. Implications for the prediction problem in reinforcement learning are discussed.

Chemistry↗

Device Feasibility Analysis of Multi-level FeFETs for Neuromorphic Computing

As an emerging non-volatile memory device technology, Ferroelectric Field-Effect Transistors (FeFETs) can enable low-power, adaptive intelligent system design. However, device dimension and operating voltage dependent reliability issues of scaled FeFETs can ultimately lead to degraded performance in solving machine learning tasks. In this article, detailed experimental characterization of FeFET devices of different dimensions have been carried out to explicitly evaluate the non-ideal behavior in device conductance programming properties like number of programming states, cycle-to-cycle (C2C) variations, device-to-device (D2D) variations, and state retention. A hardware-aware software simulation approach has been adopted to capture the adversarial effects of the non-idealities on recognition accuracy through algorithm-level performance assessment by including them in NeuroSim, a popular neural network hardware simulator, to execute a neural network model considering all other hardware constraints. With the added non-idealities, significant accuracy degradation has been observed compared to the ideal scenarios where D2D variations play the most critical role. Thereafter, feasibility of a variation-aware training method has been evaluated to tackle the accuracy drop.

42 ENGINEERING↗

Biomimetic, Soft-Material Synapse for Neuromorphic Computing: from Device to Network

Neuromorphic computing refers to a variety of brain-inspired computers, devices, and models inspired by the interconnectivity, performance, and energy efficiency of the human brain. Unlike the ubiquitous von Neumann computer architectures with complex processor cores and sequential computation, biological neurons and synapses operate by storing and processing information simultaneously with the capacity of flexible adaptation resulting in massive computational capability with much less power consumption. The search for a synaptic material which can closely imitate bio-synapse has led to an alamethicin-doped, synthetic biomembrane which can emulate key synaptic functions due to generic memristive property enabling learning and computation. This two-terminal, biomolecular memristor, in contrast to its solid-state counterparts, features similar structure, switching mechanism, and ionic transport modality as biological synapses while consuming considerably lower power. In this paper, we outline a methodology for using this biomolecular synapse to build neural networks capable of solving real-world problems. The physical mechanism underlying its volatile memristance is explored followed by the development of a model of this device for circuit simulation. We outline a circuit design technique to integrate this synapse with solid-state neuron circuit for hardware implementation. Based on these results, we develop a high level simulation framework and use a training scheme called Evolutionary Optimization for Neuromorphic System (EONS) to generate networks for solving two problems, namely iris dataset classification and EEG classification task. The small network size and comparable to state-of-the-art accuracy of these preliminary networks show its potential to enhance synaptic functionality in next generation neuromorphic hardware.

Sakib hasan, Md↗

From neural-based object recognition toward microelectronic eyes

Engineering neural network systems are best known for their abilities to adapt to the changing characteristics of the surrounding environment by adjusting system parameter values during the learning process. Rapid advances in analog current-mode design techniques have made possible the implementation of major neural network functions in custom VLSI chips. An electrically programmable analog synapse cell with large dynamic range can be realized in a compact silicon area. New designs of the synapse cells, neurons, and analog processor are presented. A synapse cell based on Gilbert multiplier structure can perform the linear multiplication for back-propagation networks. A double differential-pair synapse cell can perform the Gaussian function for radial-basis network. The synapse cells can be biased in the strong inversion region for high-speed operation or biased in the subthreshold region for low-power operation. The voltage gain of the sigmoid-function neurons is externally adjustable which greatly facilitates the search of optimal solutions in certain networks. Various building blocks can be intelligently connected to form useful industrial applications. Efficient data communication is a key system-level design issue for large-scale networks. We also present analog neural processors based on perceptron architecture and Hopfield network for communication applications. Biologically inspired neural networks have played an important role towards the creation of powerful intelligent machines. Accuracy, limitations, and prospects of analog current-mode design of the biologically inspired vision processing chips and cellular neural network chips are key design issues.

Sheu, Bing J.↗

nPINNs: nonlocal Physics-Informed Neural Networks for a parametrized nonlocal universal Laplacian operator. Algorithms and Applications

Physics-informed neural networks (PINNs) are effective in solving inverse problems based on differential and integro-differential equations with sparse, noisy, unstructured, and multifidelity data. PINNs incorporate all available information, including governing equations (reflecting physical laws), initial-boundary conditions, and observations of quantities of interest, into a loss function to be minimized, thus recasting the original problem into an optimization problem. In this paper, we extend PINNs to parameter and function inference for integral equations such as nonlocal Poisson and nonlocal turbulence models, and we refer to them as nonlocal PINNs (nPINNs). The contribution of the paper is three-fold. First, we propose a unified nonlocal Laplace operator, which converges to the classical Laplacian as one of the operator parameters, the nonlocal interaction radius δ goes to zero, and to the fractional Laplacian as δ goes to infinity. This universal operator forms a super-set of classical Laplacian and fractional Laplacian operators and, thus, has the potential to fit a broad spectrum of data sets. We provide theoretical convergence rates with respect to δ and verify them via numerical experiments. Second, we use nPINNs to estimate the two parameters, δ and α, characterizing the kernel of the unified operator. The strong non-convexity of the loss function yielding multiple (good) local minima reveals the occurrence of the operator mimicking phenomenon, that is, different pairs of estimated parameters could produce multiple solutions of comparable accuracy. Third, we propose another nonlocal operator with spatially variable order α(γ), which is more suitable for modeling turbulent Couette flow. Our results show that nPINNs can jointly infer this function as well as δ. More importantly, these parameters exhibit a universal behavior with respect to the Reynolds number, a finding that contributes to our understanding of nonlocal interactions in wall-bounded turbulence.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Deep Learning-Based Failure Prognostic Model for PV Inverter Using Field Measurements

Here, this study presents a novel approach for the precise monitoring and prognosis of photovoltaic (PV) inverter status, which is crucial for the proactive maintenance of PV systems. It addresses the gaps in traditional model-based methods, which tend to neglect the overall reliability of inverters, and the limitations of data-driven approaches that largely depend on simulated data. This research presents a robust solution applicable to real-world scenarios. The proposed data-driven model for PV inverter failure prognosis employs actual inverter measurements, integrating various operational and weather-related factors based on domain knowledge. This approach effectively represents inverter stressors and operational status. Utilizing an Enhanced Siamese Convolutional Neural Network (ESCNN), the model merges operational data with domain knowledge features, redefining the prognosis challenge as a classification task. Furthermore, the paper discusses an ESCNN-based real-time inverter failure monitoring method developed on the well-trained model. The proposed models are rigorously trained and tested with real inverter data and a novel filtering method is included to address accidental failures in practical scenarios. The results validate the model's efficacy, and the directions for future research are also outlined.

42 ENGINEERING↗

Non-destructive structural characterization of graphite components using mechanical resonance and deep learning

As compared to conventional nuclear reactors, microreactors have the potential to significantly reduce construction timelines and capital costs, decreasing the barriers for advanced nuclear reactor technologies. However, the lower power output of these microreactors (typically < 20 MWe) creates challenging economics if operation and maintenance costs cannot be sufficiently reduced. The compact size of these designs presents an opportunity for comprehensive in-situ structural health monitoring to provide real-time feedback in order to reduce operational costs associated with maintenance and downtime. Many microreactor concepts use graphite for both in-core neutron moderation and as a structural material, which has typically required some form of periodic and laborious inspection. This report provides a description and assessment of recent work with graphite to couple acoustic-based experimental measurements and characterization with machine learning models to mature structural health monitoring capabilities and generate benefits for the nuclear microreactor industry. With resilient embedded sensors in development in other programs funded by the US Department of Energy’s Office of Nuclear Energy and elsewhere, the work described herein builds upon previously funded efforts to mature non-destructive testing technology that relates measured vibrational signatures to structural changes, using a combination of new experimental measurements and machine learning processing. Building on past successful demonstrations of predictive workflows to identify structural changes in a hexagonal stainless steel test article with excellent acoustic propagation, we first performed baseline characterization on graphite samples with canonical geometries to ensure compatibility and confidence in the applied techniques for a material with distinctly different mechanical properties. In contrast to efforts in previous years, we worked exclusively with unidirectional vibration data that is more comparable to those expected from the existing embedded sensor technologies which are suitable for deployment in a reactor setting. Established acoustic and modern machine-learning-based characterization approaches were applied to the resulting datasets from these simple geometries. Both approaches were found to be highly capable of detecting even small geometric irregularities amongst nominally identical samples. As such, we then moved to testing these approaches for detection of artificial local stress perturbations introduced into a more complex geometry: a hexagonal block with drilled holes. A main outcome of this work is that a generalizable ML workflow can be used to detect and predict the characteristics of small artificial anomalies in a graphite component with a relevant geometry. While this work was performed using surficial vibration data, we expect the approach to be flexible and viable for other monitoring scenarios, such as those with different arrangements or types of sensor arrays. As compared to previously funded efforts, an existing ML workflow based on neural networks was enhanced through the addition of recently developed Fourier neural operators. As applied to previously collected and new vibration datasets, prediction accuracies of anomaly characterizations were greatly improved with minimal added computational cost. As trained on small durations of vibration data (tens of seconds) collected over a realistic number of locations, the model was able to reliably determine the presence of a subtle stress anomaly and begin to provide location estimates. Such an approach is likely to be viable for more relevant reactor damage scenarios for graphite components, such as progressive crack growth or creep.

36 MATERIALS SCIENCE↗

E(n)-Equivariant cartesian tensor message passing interatomic potential

Machine learning potential (MLP) has been a popular topic in recent years for its capability to replace expensive first-principles calculations in some large systems. Meanwhile, message passing networks have gained significant attention due to their remarkable accuracy, and a wave of message passing networks based on Cartesian coordinates has emerged. However, the information of the node in these models is usually limited to scalars, and vectors. In this work, we propose High-order Tensor message Passing interatomic Potential (HotPP), an E(n) equivariant message passing neural network that extends the node embedding and message to an arbitrary order tensor. By performing some basic equivariant operations, high order tensors can be coupled very simply and thus the model can make direct predictions of high-order tensors such as dipole moments and polarizabilities without any modifications. The tests in several datasets show that HotPP not only achieves high accuracy in predicting target properties, but also successfully performs tasks such as calculating phonon spectra, infrared spectra, and Raman spectra, demonstrating its potential as a tool for future research.

97 MATHEMATICS AND COMPUTING↗

A Length Adaptive Algorithm-Hardware Co-design of Transformer on FPGA Through Sparse Attention and Dynamic Pipelining

Transformers are considered one of the most important deep learning models since 2018, in part because it establishes state-of-the-art (SOTA) records and could potentially replace existing Deep Neural Networks (DNNs). Despite the remarkable triumphs, the prolonged turnaround time of Transformer models is a widely recognized roadblock. The variety of sequence lengths imposes additional computing overhead where inputs need to be zero-padded to the maximum sentence length in the batch to accommodate the parallel computing platforms. This paper targets the field-programmable gate array (FPGA) and proposes a coherent sequence length adaptive algorithm–hardware co-design for Transformer acceleration. Particularly, we develop a hardware-friendly sparse attention operator and a length-aware hardware resource scheduling algorithm. The proposed sparse attention operator brings the complexity of attention-based models down to linear complexity and alleviates the off-chip memory traffic. The proposed length-aware resource hardware scheduling algorithm dynamically allocates the hardware resources to fill up the pipeline slots and eliminates bubbles for NLP tasks. Experiments show that our design has very small accuracy loss and has 80.2 × and 2.6 × speedup compared to CPU and GPU implementation, and 4 × higher energy efficiency than state-of-the-art GPU accelerator optimized via CUBLAS GEMM.

Peng, Hongwu↗

A Low-Power High-Speed Smart Sensor Design for Space Exploration Missions

A low-power high-speed smart sensor system based on a large format active pixel sensor (APS) integrated with a programmable neural processor for space exploration missions is presented. The concept of building an advanced smart sensing system is demonstrated by a system-level microchip design that is composed with an APS sensor, a programmable neural processor, and an embedded microprocessor in a SOI CMOS technology. This ultra-fast smart sensor system-on-a-chip design mimics what is inherent in biological vision systems. Moreover, it is programmable and capable of performing ultra-fast machine vision processing in all levels such as image acquisition, image fusion, image analysis, scene interpretation, and control functions. The system provides about one tera-operation-per-second computing power which is a two order-of-magnitude increase over that of state-of-the-art microcomputers. Its high performance is due to massively parallel computing structures, high data throughput rates, fast learning capabilities, and advanced VLSI system-on-a-chip implementation.

Fang, Wai-Chi↗

Trustworthiness modeling and evaluation for a nearly autonomous management and control system

The Nearly Autonomous Management and Control (NAMAC) system supports the advanced reactor operation by recommending control actions to operators based on real-time measurements and digital twins (DTs) learning from the knowledge base. To enable the safe and reliable use of autonomous technologies, NAMAC and its recommendations should be trustworthy to operators and regulators at both the design and operation stages. This study proposes a NAMAC trustworthiness modeling and evaluation framework supported by trustworthiness ontologies and evidence-based approaches. The development-time and run-time ontologies are separately constructed and then converted to Bayesian networks to quantitatively evaluate the NAMAC trustworthiness. This evaluation is demonstrated by collecting and characterizing evidence from NAMAC practices, such as the development and assessment of the NAMAC system, data coverage assessment, and the training and optimizations of neural-network-based DTs. Our proposed approach can aggregate various trustworthiness attributes of complex artificial-intelligence-supported systems for safety-critical applications. It also considers the interaction between different DTs and extends beyond the trustworthiness evaluation of a single DT. In conclusion, the evidence-based method enhances the transparency of the trustworthiness modeling and evaluation processes and helps identify uncertainties and subjectivity involved in the processes.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Digital holography for real-time non-invasive monitoring of larval fish at power plant intakes

Effective evaluation of technological and operational approaches to reduce entrainment of marine organisms at cooling water intake structures (CWIS) requires accurate organism-sensing systems. Current detection methods lead to large temporal data gaps, require tedious manual analysis, and are fatal to organisms. In this report we describe integrating deep learning with a non-lethal, non-intrusive imaging method—digital holography—to rapidly detect fish larvae. Laboratory experiments demonstrated that the instrument could successfully image fish larvae at flow rates exceeding ranges seen in CWIS. Holograms of two fish larvae species, in the presence of bubbles and detritus, were collected to build a large database for training a lightweight convolutional neural network. The model achieves 97% extraction accuracy in quantifying larvae, and distinguishing them from other particles, including detritus and bubbles, when applied to a dataset of manually classified images, exceeding previous metrics for non-lethal, accurate, and real-time detection. These results demonstrate the potential of in situ holographic imaging for monitoring endangered larval fish species at power plant intake structures, and for high-fidelity, real-time applications in monitoring aquatic ichthyoplankton.

endangered species↗

Surrogate Model of Flexible Research Platform EnergyPlus Models to Enable Sensitivity Analysis

This letter report describes the surrogate models developed from the EnergyPlus model of Oak Ridge National Laboratory’s Flexible Research Platform. Two data-driven black-box models were developed, and the outputs of the surrogate models were compared with the EnergyPlus model. The two models developed are a multilayer perceptron deep learning model, and a long short-term memory (LSTM) neural network model. The three factors for selecting the black-box models are scalability, computation time, and accuracy. A total of 107 input variables were the dominant variables in determining the outputs of building energy consumptions and thermal comfort. A total of 54 output variables were identified as the prediction targets, including the system- and zone-level outputs. The large set of the simulation cases were generated by integrating sensor errors into an emulator based on EnergyPlus and Python EMS, which includes advanced control sequences from ASHRAE Guideline 36-2018: High-Performance Sequences of Operation. The surrogate models were developed based on a set of large-scale simulation runs (i.e., 4,000 runs) on a cloud platform. The comparison analysis shows that the two black-box models had good accuracy for predicting new outputs for sensitivity analysis using the root mean square error metric. As a next step, the developed surrogate models will be used to perform sensitivity analysis for different sensor impacts (e.g., sensor types, sensor locations).

42 ENGINEERING↗

Evaluating large scale aqueous organic redox flow battery performance with a hybrid numerical and machine learning framework

Aqueous organic redox flow battery (AORFB) is a promising cost-competitive technology for large-scale energy storage. Among existing work, the dihydroxyphenazine (DHP)-based AORFB has demonstrated high energy density and low-capacity degradation in 10 cm$^2$ cells during lab tests. However, its commercial-scale performance in more complex environments remains unknown, posing a barrier to commercialization. To address this gap, this work presents a comprehensive performance evaluation of a 780 cm$^2$ DHP-based AORFB by combining a physics-based numerical model, machine learning (ML)-based surrogate models, and ML-derived sensitivity quantification. Specifically, we first select 12 key battery parameters that include 10 physicochemical and 2 operation quantities, then select 6 performance metrics that include energy efficiency (EE), discharging capacity, charging energy, and power losses due to concentration, activation, and ohmic over-potentials. With such selection, 12800 combinations of the 12 parameters are subsequently generated using the Latin Hypercube Sampling method. Such combinations, together with 38 pre-defined State of Charge, are then integrated to a validated AORFB model developed in COMSOL to compute the performance metrics. With both input parameters and performance metrics, 60 deep neural network (DNN) surrogate models are then trained to approximate the relationship between the 10 physicochemical quantities and 6 performance metrics at each flow rate and current density. Sensitivity scores are then calculated based on the DNN models. Two additional sensitivity analysis tools, i.e., MARS, and SHAP, are also used to cross-validate the sensitivity scores from the DNN. The results demonstrate that 1) the standard potential ranks first in controlling EE and charging energy, 2) the membrane conductivity is most critical for power loss and EE, and 3) specific area and reaction rate control activation power loss.

25 ENERGY STORAGE↗

Heavy particle irradiation, neurochemistry and behavior: thresholds, dose-response curves and recovery of function

Exposure to heavy particles can affect the functioning of the central nervous system (CNS), particularly the dopaminergic system. In turn, the radiation-induced disruption of dopaminergic function affects a variety of behaviors that are dependent upon the integrity of this system, including motor behavior (upper body strength), amphetamine (dopamine)-mediated taste aversion learning, and operant conditioning (fixed-ratio bar pressing). Although the relationships between heavy particle irradiation and the effects of exposure depend, to some extent, upon the specific behavioral or neurochemical endpoint under consideration, a review of the available research leads to the hypothesis that the endpoints mediated by the CNS have certain characteristics in common. These include: (1) a threshold, below which there is no apparent effect; (2) the lack of a dose-response relationship, or an extremely steep dose-response curve, depending on the particular endpoint; and (3) the absence of recovery of function, such that the heavy particle-induced behavioral and neural changes are present when tested up to one year following exposure. The current report reviews the data relevant to the degree to which these characteristics are common to neurochemical and behavioral endpoints that are mediated by the effects of exposure to heavy particles on CNS activity. c2004 COSPAR. Published by Elsevier Ltd. All rights reserved.

Review↗

Predicting departure from nucleate boiling with advanced data- and physics-driven approaches

The subcooled and low-quality saturated flow boiling critical heat flux (CHF) corresponding to the departure from nucleate boiling (DNB) is one of the major limiting factors in the design and operation of pressurized water reactors (PWRs). Due to lack of agreement on its physical mechanisms, development of an accurate DNB-type CHF model has been elusive. The various steady-state (SS) predictive tools that have been proposed can be grouped into two categories: (1) data-driven methods—namely best-fit empirical correlations and look-up tables that result in relatively good agreement with specific experimental datasets but often fail to extend beyond their ranges of validity, and (2) the physics-based mechanistic models that rely on reasonable yet limited understanding of the underlying physics and apply constitutive relations to close the conservation equations. On one hand, with recent advances in computational capabilities and optimization techniques, machine learning (ML)-based methods constitute an alternative and a more advanced, data-driven approach. With enough data, they can be particularly useful in engineering fields where the physical phenomena are complex and challenging to model. On the other hand, a state-of-the-art physics-driven approach would not only provide more accurate prediction of SS CHF, but it would also help better understand and model transient applications. The data-driven regression models use deep feed-forward neural networks (NNs) and random forests to cross-validate with 1,865 CHF test cases, covering a wide range of flow conditions and channel geometries. The best-estimate ML-based predictors compare favorably with the widely used look-up table for annulus and plate, and sensitivity analysis has confirmed their effectiveness. The key advantage of ML-based methods is their online extensibility of applicability domain. The proposed physics-driven approach combines key assumptions and parameters in the relatively well-accepted theories of liquid sublayer dryout and near-wall bubble crowding. A more realistic understanding of local mechanisms has been modeled. The new model has been optimized and validated against 1,439 tube data, and it shows considerably improved performance when compared to a recent mechanistic model, demonstrating unbiased close agreement with measurements over a wide range of operating conditions. Future work will extend the improved physics-driven model to non-tube geometry applications and will couple the model with ML via a hybrid approach. The mechanistic SS work will also serve to improve understanding and to model transient CHF scenarios.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Time-lapse VSP integration and calibration of subsurface stress field utilizing machine learning approaches: A case study of the morrow B formation, FWU

This study aims to develop a methodology for calibrating subsurface stress changes through time-lapse vertical seismic profiling (VSP) integration. The selected study site is a region around the injector well located within Farnsworth field unit (FWU), where there is an ongoing CO 2 -enhanced oil recovery (EOR) operation. In our study, a site-specific rock physics model was created from extensive geological, geophysical, and geomechanical characterization through 3D seismic data, well logs, and core assessed as part of the 1D MEM conducted on the characterization well within the study area. Here, the Biot-Gassmann workflow was utilized to combine the rock physics and reservoir simulation outputs to determine the seismic velocity change due to fluid substitution. Modeled seismic velocities attributed to mean effective stress were determined from the geomechanical simulation outputs, and the stress-velocity relationship developed from ultrasonic seismic velocity measurements. A machine learning-assisted workflow comprised of an artificial neural network and a particle swarm optimizer (PSO) was utilized to minimize a penalty function created between the modeled seismic velocities and the observed time-lapse VSP dataset. The successful execution of this workflow has affirmed the suitability of acoustic time-lapse measurements for 4D-VSP geomechanical stress calibration pending measurable stress sensitivities within the anticipated effective stress changes and the availability of suitable and reliable datasets for petroelastic modeling. © 2023 Society of Chemical Industry and John Wiley & Sons, Ltd.

58 GEOSCIENCES↗