Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “neural operator”

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 199 records · Page 11

Automation of Some Operations of a Wind Tunnel Using Artificial Neural Networks

Artificial neural networks were used successfully to sequence operations in a small, recently modernized, supersonic wind tunnel at NASA-Lewis Research Center. The neural nets generated correct estimates of shadowgraph patterns, pressure sensor readings and mach numbers for conditions occurring shortly after startup and extending to fully developed flow. Artificial neural networks were trained and tested for estimating: sensor readings from shadowgraph patterns, shadowgraph patterns from shadowgraph patterns and sensor readings from sensor readings. The 3.81 by 10 in. (0.0968 by 0.254 m) tunnel was operated with its mach 2.0 nozzle, and shadowgraph was recorded near the nozzle exit. These results support the thesis that artificial neural networks can be combined with current workstation technology to automate wind tunnel operations.

Decker, Arthur J.↗

Hardware Acceleration of Adaptive Neural Algorithms.

As tradit ional numerical computing has faced challenges, researchers have turned towards alternative computing approaches to reduce power - per - computation metrics and improve algorithm performance. Here, we describe an approach towards non - conventional computing that strengthens the connection between machine learning and neuroscience concepts. The Hardware Acceleration of Adaptive Neural Algorithms (HAANA) project ha s develop ed neural machine learning algorithms and hardware for applications in image processing and cybersecurity. While machine learning methods are effective at extracting relevant features from many types of data, the effectiveness of these algorithms degrades when subjected to real - world conditions. Our team has generated novel neural - inspired approa ches to improve the resiliency and adaptability of machine learning algorithms. In addition, we have also designed and fabricated hardware architectures and microelectronic devices specifically tuned towards the training and inference operations of neural - inspired algorithms. Finally, our multi - scale simulation framework allows us to assess the impact of microelectronic device properties on algorithm performance.

97 MATHEMATICS AND COMPUTING↗

Encoder–decoder neural network for solving the nonlinear Fokker–Planck–Landau collision operator in XGC

An encoder–decoder neural network has been used to examine the possibility for acceleration of a partial integro-differential equation, the Fokker–Planck–Landau collision operator. This is part of the governing equation in the massively parallel particle-in-cell code XGC, which is used to study turbulence in fusion energy devices. The neural network emphasizes physics-inspired learning, where it is taught to respect physical conservation constraints of the collision operator by including them in the training loss, along with the ℓ 2 loss. In particular, network architectures used for the computer vision task of semantic segmentation have been used for training. A penalization method is used to enforce the ‘soft’ constraints of the system and integrate error in the conservation properties into the loss function. During training, quantities representing the particle density, momentum and energy for all species of the system are calculated at each configuration vertex, mirroring the procedure in XGC. This simple training has produced a median relative loss, across configuration space, of the order of 10 –4 , which is low enough if the error is of random nature, but not if it is of drift nature in time steps. The run time for the current Picard iterative solver of the operator is O(n 2 ), where n is the number of plasma species. As the XGC1 code begins to attack problems including a larger number of species, the collision operator will become expensive computationally, making the neural network solver even more important, especially since its training only scales as O(n). Here, a wide enough range of collisionality has been considered in the training data to ensure the full domain of collision physics is captured. An advanced technique to decrease the losses further will be subject of a subsequent report. Eventual work will include expansion of the network to include multiple plasma species.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Terminal attractors for addressable memory in neural networks

A new type of attractors - terminal attractors - for an addressable memory in neural networks operating in continuous time is introduced. These attractors represent singular solutions of the dynamical system. They intersect (or envelope) the families of regular solutions while each regular solution approaches the terminal attractor in a finite time period. It is shown that terminal attractors can be incorporated into neural networks such that any desired set of these attractors with prescribed basins is provided by an appropriate selection of the weight matrix.

Zak, Michail↗

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 $\delta$ goes to zero, and to the fractional Laplacian as $\delta$ 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 also provide theoretical convergence rates with respect to $\delta$ and verify them via numerical experiments. Second, we use nPINNs to estimate the two parameters, $\delta$ and $\alpha$, 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 $\alpha(y)$, which is more suitable for modeling turbulent Couette flow. Our results show that nPINNs can jointly infer this function as well as $\delta$. 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.

97 MATHEMATICS AND COMPUTING↗

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↗

Machine Learning-Assisted High-Temperature Reservoir Thermal Energy Storage Optimization: Numerical Modeling and Machine Learning Input and Output Files

This data set includes the numerical modeling input files and output files used to synthesize data, and the reduced-order machine learning models trained from the synthesized data for reservoir thermal energy storage site identification. In this study, a machine-learning-assisted computational framework is presented to identify High-Temperature Reservoir Thermal Energy Storage (HT-RTES) site with optimal performance metrics by combining physics-based simulation with stochastic hydrogeologic formation and thermal energy storage operation parameters, artificial neural network regression of the simulation data, and genetic algorithm-enabled multi-objective optimization. A doublet well configuration with a layered (aquitard-aquifer-aquitard) generic reservoir is simulated for cases of continuous operation and seasonal-cycle operation scenarios. Neural network-based surrogate models are developed for the two scenarios and applied to generate the Pareto fronts of the HT-RTES performance for four potential HT-RTES sites. The developed Pareto optimal solutions indicate the performance of HT-RTES is operation-scenario (i.e., fluid cycle) and reservoir-site dependent, and the performance metrics have competing effects for a given site and a given fluid cycle. The developed neural network models can be applied to identify suitable sites for HT-RTES, and the proposed framework sheds light on the design of resilient HT-RTES systems. All the simulations and the neural network model were done by Idaho National Laboratory. A detailed description of the work was reported in publication linked below.

15 GEOTHERMAL ENERGY↗

Terminal attractors in neural networks

A new type of attractor (terminal attractors) for content-addressable memory, associative memory, and pattern recognition in artificial neural networks operating in continuous time is introduced. The idea of a terminal attractor is based upon a violation of the Lipschitz condition at a fixed point. As a result, the fixed point becomes a singular solution which envelopes the family of regular solutions, while each regular solution approaches such an attractor in finite time. It will be shown that terminal attractors can be incorporated into neural networks such that any desired set of these attractors with prescribed basins is provided by an appropriate selection of the synaptic weights. The applications of terminal attractors for content-addressable and associative memories, pattern recognition, self-organization, and for dynamical training are illustrated.

Zak, Michail↗

A Feasibility Study for Perioperative Ventricular Tachycardia Prognosis and Detection and Noise Detection Using a Neural Network and Predictive Linear Operators

To locate the accessory pathway(s) in preexicitation syndromes, epicardial and endocardial ventricular mapping is performed during anterograde ventricular activation via accessory pathway(s) from data originally received in signal form. As the number of channels increases, it is pertinent that more automated detection of coherent/incoherent signals is achieved as well as the prediction and prognosis of ventricular tachywardia (VT). Today's computers and computer program algorithms are not good in simple perceptual tasks such as recognizing a pattern or identifying a sound. This discrepancy, among other things, has been a major motivating factor in developing brain-based, massively parallel computing architectures. Neural net paradigms have proven to be effective at pattern recognition tasks. In signal processing, the picking of coherent/incoherent signals represents a pattern recognition task for computer systems. The picking of signals representing the onset ot VT also represents such a computer task. We attacked this problem by defining four signal attributes for each potential first maximal arrival peak and one signal attribute over the entire signal as input to a back propagation neural network. One attribute was the predicted amplitude value after the maximum amplitude over a data window. Then, by using a set of known (user selected) coherent/incoherent signals, and signals representing the onset of VT, we trained the back propagation network to recognize coherent/incoherent signals, and signals indicating the onset of VT. Since our output scheme involves a true or false decision, and since the output unit computes values between 0 and 1, we used a Fuzzy Arithmetic approach to classify data as coherent/incoherent signals. Furthermore, a Mean-Square Error Analysis was used to determine system stability. The neural net based picking coherent/incoherent signal system achieved high accuracy on picking coherent/incoherent signals on different patients. The system also achieved a high accuracy of picking signals which represent the onset of VT, that is, VT immediately followed these signals. A special binary representation of the input and output data allowed the neural network to train very rapidly as compared to another standard decimal or normalized representations of the data.

Moebes, T. A.↗

Anomaly detection of turbopump vibration in Space Shuttle Main Engine using statistics and neural networks

The statistical and neural networks methods have been applied to investigate the feasibility in detecting anomalies in turbopump vibration of SSME. The anomalies are detected based on the amplitude of peaks of fundamental and harmonic frequencies in the power spectral density. These data are reduced to the proper format from sensor data measured by strain gauges and accelerometers. Both methods are feasible to detect the vibration anomalies. The statistical method requires sufficient data points to establish a reasonable statistical distribution data bank. This method is applicable for on-line operation. The neural networks method also needs to have enough data basis to train the neural networks. The testing procedure can be utilized at any time so long as the characteristics of components remain unchanged.

Lo, C. F.↗

Resolving extreme jet substructure

We study the effectiveness of theoretically-motivated high-level jet observables in the extreme context of jets with a large number of hard sub-jets (up to N = 8). Previous studies indicate that high-level observables are powerful, interpretable tools to probe jet substructure for N ≤ 3 hard sub-jets, but that deep neural networks trained on low-level jet constituents match or slightly exceed their performance. We extend this work for up to N = 8 hard sub-jets, using deep particle-flow networks (PFNs) and Transformer based networks to estimate a loose upper bound on the classification performance. A fully-connected neural network operating on a standard set of high-level jet observables, 135 N-subjetiness observables and jet mass, reach classification accuracy of 86.90%, but fall short of the PFN and Transformer models, which reach classification accuracies of 89.19% and 91.27% respectively, suggesting that the constituent networks utilize information not captured by the set of high-level observables. We then identify additional high-level observables which are able to narrow this gap, and utilize LASSO regularization for feature selection to identify and rank the most relevant observables and provide further insights into the learning strategies used by the constituent-based neural networks. The final model contains only 31 high-level observables and is able to match the performance of the PFN and approximate the performance of the Transformer model to within 2%.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Multifidelity deep operator networks for data-driven and physics-informed problems

Operator learning for complex nonlinear systems is increasingly common in modeling multi-physics and multi-scale systems. However, training such high-dimensional operators requires a large amount of expensive, high-fidelity data, either from experiments or simulations. In this work, we present a composite Deep Operator Network (DeepONet) for learning using two datasets with different levels of fidelity to accurately learn complex operators when sufficient high-fidelity data is not available. Additionally, we demonstrate that the presence of low-fidelity data can improve the predictions of physics-informed learning with DeepONets. We demonstrate the new multi-fidelity training in diverse examples, including modeling of the ice-sheet dynamics of the Humboldt glacier, Greenland, using two different fidelity models and also using the same physical model at two different resolutions.

97 MATHEMATICS AND COMPUTING↗

Studies of Vestibular Neurons in Normal, Hyper- and Hypogravity

During the past year, pre-, in- and postflight studies were conducted in association with the Axon project for Bion 10 (Cosmos 2229). Recordings were made during pre- and postflight studies, from 118 horizontal semicircular canal afferents and 27 vestibular nucleus neurons in 7 rhesus monkeys; 137 pulse rotation protocols alone were executed (548 acceleration and deceleration responses were curve fit). Usable data was obtained from 127 horizontal afferents concerning their spontaneous discharge. Curve fits and analysis was made of sinusoidal and sum of sinusoidal responses from 42 and 35 horizontal afferents, respectively. Also recordings were made from neurons inflight from the two flight animals. The mean spontaneous rate varied from 128 spikes/sec. during preflight to 92 spikes/sec during postflight (day 5) - a change of 28%. In direct contrast to the results of Cosmos 2044, the best fitted neural adaptation operator (k) and the gain of the pulse response were decreased during post flight when compared to preflight. Surprisingly, the best fitted gain and k values for the sum of sines were slightly elevated during post flight tests. The gain and phase of single sine responses were compared for pre- and post flight tests and compared to a larger population of afferents. In contrast to Cosmos 2044 results where on the first day of post flight testing the gains of the best fitted sine response were skewed toward the higher values of the Miles and Braitman distribution, the gain of the best fitted sine responses during the first day of post flight testing (day 2) during Cosmos 2229 were exactly on the mode of the Miles and Braitman distribution. Thus, at least for the periodic stimuli, (pulses and sine waves) we found no change in gain and neural adaptation during post flight testing following Cosmos 2229. This conclusion is different from the one derived following the Cosmos 2044 flight. Cosmos flight 2229 differed from Cosmos flight 2044 in several significant ways: For example, during preflight, (1) The animals preflight training was different (less well trained on the gaze task) and (2) the animals were exposed to more experimental manipulations (surgical and rotational). Inflight, (1) the animals were required to make a pointing gesture (motor response) in association with eye movements to obtain reward, (2) the inflight diet was different (more balanced), (3) the feeder for one of the animals clogged following 9 days of flight resulting in evident dehydration and probably less head motion exposure in that monkey and (4) there was limited video taping of the monkeys in space. During postflight, (1) we were unable to test the flight animals until 26 hours postflight as compared to 14.5 hours during Cosmos 2044, (2) the animals received significantly more exposure to motion stimuli during postflight testing than during Cosmos 2044. These differences in the vestibular environment will require analysis of several parameters other than just neural and eye movement responses. For example, computer programs will have to be written and used to recover and quantify the number of head movements made by each animal during flight. This activity is critical to the production of neural adaptation and increased gain.

Correia, Manning J.↗

Deep transfer operator learning for partial differential equations under conditional shift

Transfer learning enables the transfer of knowledge gained while learning to perform one task (source) to a related but different task (target), hence addressing the expense of data acquisition and labelling, potential computational power limitations and dataset distribution mismatches. Here, we propose a new transfer learning framework for task-specific learning (functional regression in partial differential equations) under conditional shift based on the deep operator network (DeepONet). Task-specific operator learning is accomplished by fine-tuning task-specific layers of the target DeepONet using a hybrid loss function that allows for the matching of individual target samples while also preserving the global properties of the conditional distribution of the target data. Inspired by conditional embedding operator theory, we minimize the statistical distance between labelled target data and the surrogate prediction on unlabelled target data by embedding conditional distributions onto a reproducing kernel Hilbert space. We demonstrate the advantages of our approach for various transfer learning scenarios involving nonlinear partial differential equations under diverse conditions due to shifts in the geometric domain and model dynamics. Our transfer learning framework enables fast and efficient learning of heterogeneous tasks despite considerable differences between the source and target domains.

42 ENGINEERING↗

Expediting DECam Multimessenger Counterpart Searches with Convolutional Neural Networks

Searches for counterparts to multimessenger events with optical imagers use difference imaging to detect new transient sources. However, even with existing artifact-detection algorithms, this process simultaneously returns several classes of false positives: false detections from poor-quality image subtractions, false detections from low signal-to-noise images, and detections of preexisting variable sources. Currently, human visual inspection to remove the false positives is a central part of multimessenger follow-up observations, but when next generation gravitational wave and neutrino detectors come online and increase the rate of multimessenger events, the visual inspection process will be prohibitively expensive. We approach this problem with two convolutional neural networks operating on the difference imaging outputs. The first network focuses on removing false detections and demonstrates an accuracy of 92% on our data set. The second network focuses on sorting all real detections by the probability of being a transient source within a host galaxy and distinguishes between various classes of images that previously required additional human inspection. We find the number of images requiring human inspection will decrease by a factor of 1.5 using our approach alone and a factor of 3.6 using our approach in combination with existing algorithms, facilitating rapid multimessenger counterpart identification by the astronomical community.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Separable physics-informed DeepONet: Breaking the curse of dimensionality in physics-informed machine learning

The deep operator network (DeepONet) has shown remarkable potential in solving partial differential equations (PDEs) by mapping between infinite-dimensional function spaces using labeled datasets. However, in scenarios lacking labeled data, the physics-informed DeepONet (PI-DeepONet) approach, which utilizes the residual loss of the governing PDE to optimize the network parameters, faces significant computational challenges, particularly due to the curse of dimensionality. This limitation has hindered its application to high-dimensional problems, making even standard 3D spatial with 1D temporal problems computationally prohibitive. Additionally, the computational requirement increases exponentially with the discretization density of the domain. Here, to address these challenges and enhance scalability for high-dimensional PDEs, we introduce the Separable physics-informed DeepONet (Sep-PI-DeepONet). This framework employs a factorization technique, utilizing sub-networks for individual one-dimensional coordinates, thereby reducing the number of forward passes and the size of the Jacobian matrix required for gradient computations. By incorporating forward-mode automatic differentiation (AD), we further optimize computational efficiency, achieving linear scaling of computational cost with discretization density and dimensionality, making our approach highly suitable for high-dimensional PDEs. We demonstrate the effectiveness of Sep-PI-DeepONet through three benchmark PDE models: the viscous Burgers’ equation, Biot’s consolidation theory, and a parameterized heat equation. Our framework maintains accuracy comparable to the conventional PI-DeepONet while reducing training time by two orders of magnitude. Notably, for the heat equation solved as a 4D problem, the conventional PI-DeepONet was computationally infeasible (estimated 289.35 h), while the Sep-PI-DeepONet completed training in just 2.5 h. These results underscore the potential of Sep-PI-DeepONet in efficiently solving complex, high-dimensional PDEs, marking a significant advancement in physics-informed machine learning.

Neural operator↗

Baroreflex regulation of blood pressure during dynamic exercise

From the work of Potts et al. Papelier et al. and Shi et al. it is readily apparent that the arterial (aortic and carotid) baroreflexes are reset to function at the prevailing ABP of exercise. The blood pressure of exercise is the result of the hemodynamic (cardiac output and TPR) responses, which appear to be regulated by two redundant neural control systems, "Central Command" and the "exercise pressor reflex". Central Command is a feed-forward neural control system that operates in parallel with the neural regulation of the locomotor system and appears to establish the hemodynamic response to exercise. Within the central nervous system it appears that the HLR may be the operational site for Central Command. Specific neural sites within the HLR have been demonstrated in animals to be active during exercise. With the advent of positron emission tomography (PET) and single-photon emission computed tomography (SPECT), the anatomical areas of the human brain related to Central Command are being mapped. It also appears that the Nucleus Tractus Solitarius and the ventrolateral medulla may serve as an integrating site as they receive neural information from the working muscles via the group III/IV muscle afferents as well as from higher brain centers. This anatomical site within the CNS is now the focus of many investigations in which arterial baroreflex function, Central Command and the "exercise pressor reflex" appear to demonstrate inhibitory or facilitatory interaction. The concept of whether Central Command is the prime mover in the resetting of the arterial baroreceptors to function at the exercising ABP or whether the resetting is an integration of the "exercise pressor reflex" information with that of Central Command is now under intense investigation. However, it would be justified to conclude, from the data of Bevegard and Shepherd, Dicarlo and Bishop, Potts et al., and Papelier et al. that the act of exercise results in the resetting of the arterial baroreflex. In addition, if, as we have proposed, the cardiopulmonary baroreceptors primarily monitors and reflexly regulates cardiac filling volume, it would seem from the data of Mack et al. and Potts et al. that the cardiopulmonary baroreceptor is also reset at the beginning of exercise. Therefore, investigations of the neural mechanisms of regulation involving Central Command and cardiopulmonary afferents, similar to those being undertaken for the arterial baroreflex, need to be established.

NASA Discipline Cardiopulmonary↗