Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Time step adaptivity”

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 289 records · Page 16

Optimizing transmit field inhomogeneity of parallel RF transmit design in 7T MRI using deep learning

Ultrahigh field (UHF) Magnetic Resonance Imaging (MRI) provides a higher signal-to-noise ratio and, thereby, higher spatial resolution. However, UHF MRI introduces challenges such as transmit radiofrequency (RF) field (B+1) inhomogeneities, leading to uneven flip angles and image intensity anomalies. These issues can significantly degrade imaging quality and its medical applications. This study addresses B+1 field homogeneity through a novel deep learning-based strategy. Traditional methods like Magnitude Least Squares (MLS) optimization have been effective but are time-consuming and dependent on the patient’s presence. Recent machine learning approaches, such as RF Shim Prediction by Iteratively Projected Ridge Regression and deep learning frameworks, have shown promise but face limitations like extensive training times and oversimplified architectures. We propose a two-step deep learning strategy. First, we obtain the desired reference RF shimming weights from multi-channel B+1 fields using random-initialized Adaptive Moment Estimation. Then, we employ Residual Networks (ResNets) to train a model that maps B+1 fields to target RF shimming outputs. Our approach does not rely on pre-calculated reference optimizations for the testing process and efficiently learns residual functions. Comparative studies with traditional MLS optimization demonstrate our method’s advantages in terms of speed and accuracy. The proposed strategy achieves a faster and more efficient RF shimming design, significantly improving imaging quality at UHF. This advancement holds potential for broader applications in medical imaging and diagnostics.

Lu, Zhengyi [Vanderbilt University]↗

Adaptive Bayes classifiers for remotely sensed data

An algorithm is developed for a learning, adaptive, statistical pattern classifier for remotely sensed data. The estimation procedure consists of two steps: (1) an optimal stochastic approximation of the parameters of interest, and (2) a projection of the parameters in time and space. The results reported are for Gaussian data in which the mean vector of each class may vary with time or position after the classifier is trained.

Raulston, H. S.↗

On the solution of creep induced buckling in general structure

This paper considers the pre and post buckling behavior of general structures exposed to high temperature fields for long durations wherein creep effects become significant. The solution to this problem is made possible through the use of closed upper bounding constraint surfaces which enable the development of a new time stepping algorithm. This permits the stable and efficient solution of structural problems which exhibit indefinite tangent properties. Due to the manner of constraining/bounding successive iterates, the algorithm developed herein is largely self adaptive, inherently stable, sufficiently flexible to handle geometric material and boundary induced nonlinearity, and can be incorporated into either finite element or difference simulations. To illustrate the capability of the procedure, as well as, the physics of creep induced pre and post buckling behavior, the results of several numerical experiments are included.

Padovan, J.↗

Method and system for training dynamic nonlinear adaptive filters which have embedded memory

Described herein is a method and system for training nonlinear adaptive filters (or neural networks) which have embedded memory. Such memory can arise in a multi-layer finite impulse response (FIR) architecture, or an infinite impulse response (IIR) architecture. We focus on filter architectures with separate linear dynamic components and static nonlinear components. Such filters can be structured so as to restrict their degrees of computational freedom based on a priori knowledge about the dynamic operation to be emulated. The method is detailed for an FIR architecture which consists of linear FIR filters together with nonlinear generalized single layer subnets. For the IIR case, we extend the methodology to a general nonlinear architecture which uses feedback. For these dynamic architectures, we describe how one can apply optimization techniques which make updates closer to the Newton direction than those of a steepest descent method, such as backpropagation. We detail a novel adaptive modified Gauss-Newton optimization technique, which uses an adaptive learning rate to determine both the magnitude and direction of update steps. For a wide range of adaptive filtering applications, the new training algorithm converges faster and to a smaller value of cost than both steepest-descent methods such as backpropagation-through-time, and standard quasi-Newton methods. We apply the algorithm to modeling the inverse of a nonlinear dynamic tracking system 5, as well as a nonlinear amplifier 6.

Rabinowitz, Matthew↗

The revised FLORIDyn model: implementation of heterogeneous flow and the Gaussian wake

In this paper, a new version of the FLOw Redirection and Induction Dynamics (FLORIDyn) model is presented. The new model uses the three-dimensional parametric Gaussian FLORIS model and can provide dynamic wind farm simulations at a low computational cost under heterogeneous and changing wind conditions. Both FLORIS and FLORIDyn are parametric models which can be used to simulate wind farms, evaluate controller performance and can serve as a control-oriented model. One central element in which they differ is in their representation of flow dynamics: FLORIS neglects these and provides a computationally very cheap approximation of the mean wind farm flow. FLORIDyn defines a framework which utilizes this low computational cost of FLORIS to simulate basic wake dynamics. This is achieved by creating so-called observation points (OPs) at each time step at the rotor plane which inherit the turbine state. In this work, we develop the initial FLORIDyn framework further considering multiple aspects. The underlying FLORIS wake model is replaced by a Gaussian wake model. The distribution and characteristics of the OPs are adapted to account for the new parametric model but also to take complex flow conditions into account. To achieve this, a mathematical approach is developed to combine the parametric model and the changing, heterogeneous world conditions and link them with each OP. We also present a computationally lightweight wind field model to allow for a simulation environment in which heterogeneous flow conditions are possible. FLORIDyn is compared to Simulator for Offshore Wind Farm Applications (SOWFA) simulations in three- and nine-turbine cases under static and changing environmental conditions. The results show a good agreement with the timing of the impact of upstream state changes on downstream turbines. They also show a good agreement in terms of how wakes are displaced by wind direction changes and when the resulting velocity deficit is experienced by downstream turbines. A good fit of the mean generated power is ensured by the underlying FLORIS model. In the three-turbine case, FLORIDyn simulates 4 s simulation time in 24.49 ms computational time. The resulting new FLORIDyn model proves to be a computationally attractive and capable tool for model-based dynamic wind farm control.

17 WIND ENERGY↗

Revisiting Temporal Blocking Stencil Optimizations

Iterative stencils are used widely across the spectrum of High Performance Computing (HPC) applications. Many efforts have been put into optimizing stencil GPU kernels, given the prevalence of GPU-accelerated supercomputers. To improve the data locality, temporal blocking is an optimization that combines a batch of time steps to process them together. Under the observation that GPUs are evolving to resemble CPUs in some aspects, we revisit temporal blocking optimizations for GPUs. We explore how temporal blocking schemes can be adapted to the new features in the recent Nvidia GPUs, including large scratchpad memory, hardware prefetching, and device-wide synchronization. We propose a novel temporal blocking method, EBISU, which champions low device occupancy to drive aggressive deep temporal blocking on large tiles that are executed tile-by-tile. We compare EBISU with state-of-the-art temporal blocking libraries: STENCILGEN and AN5D. We also compare with state-of-the-art stencil auto-tuning tools that are equipped with temporal blocking optimizations: ARTEMIS and DRSTENCIL. Over a wide range of stencil benchmarks, EBISU achieves speedups up to 2.53x and a geometric mean speedup of 1.49x over the best state-of-the-art performance in each stencil benchmark.

Zhang, Lingqi↗

PRMI: A Dataset of Minirhizotron Images for Diverse Plant Root Study

Understanding a plant's root system architecture (RSA) is crucial for a variety of plant science problem domains including sustainability and climate adaptation. Minirhizotron (MR) technology is a widely-used approach for phenotyping RSA non-destructively by capturing root imagery over time. Precisely segmenting roots from the soil in MR imagery is a critical step in studying RSA features. In this paper, we introduce a large-scale dataset of plant root images captured by MR technology. In total, there are over 72K RGB root images across six different species including cotton, papaya, peanut, sesame, sunflower, and switchgrass in the dataset. The images span a variety of conditions including varied root age, root structures, soil types, and depths under the soil surface. All of the images have been annotated with weak image-level labels indicating whether each image contains roots or not. The image-level labels can be used to support weakly supervised learning in plant root segmentation tasks. In addition, 63K images have been manually annotated to generate pixel-level binary masks indicating whether each pixel corresponds to root or not. These pixel-level binary masks can be used as ground truth for supervised learning in semantic segmentation tasks. By introducing this dataset, we aim to facilitate the automatic segmentation of roots and the research of RSA with deep learning and other image analysis algorithms.

Xu, Weihuang↗

Mission definition study for a VLBI station utilizing the Space Shuttle

The uses of the Space Shuttle transportation system for orbiting VeryLong-Baseline Interferometry (OVLBI) were examined, both with respect to technical feasibility and its scientific possibilities. The study consisted of a critical look at the adaptability of current technology to an orbiting environment, the suitability of current data reduction facilities for the new technique, and a review of the new science that is made possible by using the Space Shuttle as a moving platform for a VLBI terminal in space. The conclusions are positive in all respects: no technological deficiencies exist that would need remedy, the data processing problem can be handled easily by straightforward adaptations of existing systems, and there is a significant new research frontier to be explored, with the Space Shuttle providing the first step. The VLBI technique utilizes the great frequency stability of modern atomic time standards, the power of integrated circuitry to perform real-time signal conditioning, and the ability of magnetic tape recorders to provide essentially error-free data recording, all of which combine to permit the realization of radio interferometry at arbitrarily large baselines.

Burke, B. F.↗

An experimental study of an adaptive-wall wind tunnel

A series of adaptive wall ventilated wind tunnel experiments was carried out to demonstrate the feasibility of using the side wall pressure distribution as the flow variable for the assessment of compatibility with free air conditions. Iterative and one step convergence methods were applied using the streamwise velocity component, the side wall pressure distribution and the normal velocity component in order to investigate their relative merits. The advantage of using the side wall pressure as the flow variable is to reduce the data taking time which is one the major contributors to the total testing time. In ventilated adaptive wall wind tunnel testing, side wall pressure measurements require simple instrumentation as opposed to the Laser Doppler Velocimetry used to measure the velocity components. In ventilated adaptive wall tunnel testing, influence coefficients are required to determine the pressure corrections in the plenum compartment. Experiments were carried out to evaluate the influence coefficients from side wall pressure distributions, and from streamwise and normal velocity distributions at two control levels. Velocity measurements were made using a two component Laser Doppler Velocimeter system.

Celik, Zeki↗

Deep Symbolic Optimization for Electric Component Sizing in Fixed Topology Power Converters

Power converters (PC) are a major component in any current electronic hardware device. The development and design is usually guided by expert knowledge and heavily relies on human intuition and experience. The process is a very time consuming and costly activity and it is generally hard to improve upon current designs. As a first step towards autonomous PC design, we are here proposing a new framework for the sizing of components for fixed topology PCs based on given design requirements. To this end, we developed surrogate models for rapid evaluation of new topologies and adapt the deep symbolic optimization (DSO) framework to generate new topologies guided by a reinforcement learning training signal. In an empirical evaluation, we show that our DSO based approach is able to find the optimal configuration for all investigated topologies, while reducing the learning time by at least a factor of 100 compared to popular RL algorithms.

Xue, Lincoln↗

Eulerian formulation of the interacting particle representation model of homogeneous turbulence

The Interacting Particle Representation Model (IPRM) of homogeneous turbulence incorporates information about the morphology of turbulent structures within the con nes of a one-point model. In the original formulation [Kassinos & Reynolds, Center for Turbulence Research: Annual Research Briefs, 31{51, (1996)], the IPRM was developed in a Lagrangian setting by evolving second moments of velocity conditional on a given gradient vector. In the present work, the IPRM is re-formulated in an Eulerian framework and evolution equations are developed for the marginal PDFs. Eulerian methods avoid the issues associated with statistical estimators used by Lagrangian approaches, such as slow convergence. A specific emphasis of this work is to use the IPRM to examine the long time evolution of homogeneous turbulence. We first describe the derivation of the marginal PDF in spherical coordinates, which reduces the number of independent variables and the cost associated with Eulerian simulations of PDF models. Next, a numerical method based on radial basis functions over a spherical domain is adapted to the IPRM. Finally, results obtained with the new Eulerian solution method are thoroughly analyzed. The sensitivity of the Eulerian simulations to parameters of the numerical scheme, such as the size of the time step and the shape parameter of the radial basis functions, is examined. A comparison between Eulerian and Lagrangian simulations is performed to discern the capabilities of each of the methods. Finally, a linear stability analysis based on the eigenvalues of the discrete differential operators is carried out for both the new Eulerian solution method and the original Lagrangian approach.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Direct Adaptive Aircraft Control Using Dynamic Cell Structure Neural Networks

A Dynamic Cell Structure (DCS) Neural Network was developed which learns topology representing networks (TRNS) of F-15 aircraft aerodynamic stability and control derivatives. The network is integrated into a direct adaptive tracking controller. The combination produces a robust adaptive architecture capable of handling multiple accident and off- nominal flight scenarios. This paper describes the DCS network and modifications to the parameter estimation procedure. The work represents one step towards an integrated real-time reconfiguration control architecture for rapid prototyping of new aircraft designs. Performance was evaluated using three off-line benchmarks and on-line nonlinear Virtual Reality simulation. Flight control was evaluated under scenarios including differential stabilator lock, soft sensor failure, control and stability derivative variations, and air turbulence.

Jorgensen, Charles C.↗

Lessons Learned from Assimilating Altimeter Data into a Coupled General Circulation Model with the GMAO Augmented Ensemble Kalman Filter

Satellite altimetry measurements have provided global, evenly distributed observations of the ocean surface since 1993. However, the difficulties introduced by the presence of model biases and the requirement that data assimilation systems extrapolate the sea surface height (SSH) information to the subsurface in order to estimate the temperature, salinity and currents make it difficult to optimally exploit these measurements. This talk investigates the potential of the altimetry data assimilation once the biases are accounted for with an ad hoc bias estimation scheme. Either steady-state or state-dependent multivariate background-error covariances from an ensemble of model integrations are used to address the problem of extrapolating the information to the sub-surface. The GMAO ocean data assimilation system applied to an ensemble of coupled model instances using the GEOS-5 AGCM coupled to MOM4 is used in the investigation. To model the background error covariances, the system relies on a hybrid ensemble approach in which a small number of dynamically evolved model trajectories is augmented on the one hand with past instances of the state vector along each trajectory and, on the other, with a steady state ensemble of error estimates from a time series of short-term model forecasts. A state-dependent adaptive error-covariance localization and inflation algorithm controls how the SSH information is extrapolated to the sub-surface. A two-step predictor corrector approach is used to assimilate future information. Independent (not-assimilated) temperature and salinity observations from Argo floats are used to validate the assimilation. A two-step projection method in which the system first calculates a SSH increment and then projects this increment vertically onto the temperature, salt and current fields is found to be most effective in reconstructing the sub-surface information. The performance of the system in reconstructing the sub-surface fields is particularly impressive for temperature, but not as satisfactory for salt.

Keppenne, Christian↗

Implicit–explicit multirate infinitesimal stage-restart methods

Implicit–Explicit (IMEX) methods are flexible numerical time integration methods which solve an initial-value problem (IVP) that is split into stiff and nonstiff processes with the goal of lower computational costs than a purely implicit or explicit approach. A complementary form of flexible IVP solvers are multirate infinitesimal methods for problems split into fast- and slow-changing dynamics, that solve a multirate IVP by evolving a sequence of “fast” IVPs using any suitably accurate algorithm. This article introduces a new class of high-order implicit–explicit multirate methods that are designed for multirate IVPs in which the slow-changing dynamics are further split in an IMEX fashion. This new class, which we call implicit–explicit multirate infinitesimal stage-restart (IMEX-MRI-SR), both improves upon the previous implicit–explicit multirate infinitesimal generalized-structure additive Runge Kutta (IMEX-MRI-GARK) methods by allowing for far easier creation of new embedded methods, and extends multirate exponential Runge Kutta (MERK) methods by allowing the fast-changing dynamics to be nonlinear and the methods to be implicit. We leverage GARK theory to derive conditions for orders of accuracy up to four, and we provide second- and third-order accurate example methods, which are the first known embedded MRI methods with IMEX structure. We then perform numerical simulations demonstrating convergence rates and computational performance in both fixed-step and adaptive-step settings.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Constrained Self-adaptive Solutions Procedures for Structure Subject to High Temperature Elastic-plastic Creep Effects

This paper will develop a new solution strategy which can handle elastic-plastic-creep problems in an inherently stable manner. This is achieved by introducing a new constrained time stepping algorithm which will enable the solution of creep initiated pre/postbuckling behavior where indefinite tangent stiffnesses are encountered. Due to the generality of the scheme, both monotone and cyclic loading histories can be handled. The presentation will give a thorough overview of current solution schemes and their short comings, the development of constrained time stepping algorithms as well as illustrate the results of several numerical experiments which benchmark the new procedure.

Padovan, J.↗

Spatio-temporal adaptation algorithm for two-dimensional reacting flows

A spatio-temporal adaptive algorithm for solving the unsteady Euler equations with chemical source terms is presented. Quadrilateral cells are used in two spatial dimensions which allow for embedded meshes tracking moving flow features with spatially varying time-steps which are multiples of global minimum time-steps. Blast wave interactions corresponding to a perfect gas (frozen) and a Lighthill dissociating gas (nonequilibrium) are considered for circular arc cascade and 90 degree bend duct geometries.

Pervaiz, Mehtab M.↗

Torsional Eye Movements Evoked by Unilateral Labyrinthine Galvanic Polarizations in the Squirrel Monkey

Electrical stimulation of vestibular-nerve afferents innervating the semicircular canals has been used to identify the extraocular muscles receiving activation or inhibition by individual ampullary nerves. This technique was originally developed by Szentagothai (1950) and led to the description of three neuron reflex arcs that connect each semicircular canal through an interneuron traversing in the region of the medial longitudinal fasciculus to one ipsilateral and one contralateral eye muscle. Selective ampullary nerve stimulation was subsequently used by Cohen and colleagues (Cohen and Suzuki, 1963; Cohen et al., 1964; Suzuki et al., 1964; Cohen et al., 1966) to study movements of the eyes and activation of individual extraocular muscles in response to stimulation of combinations of ampullary nerves. This work led to a description of the now familiar relationships between activation of a semicircular canal ampullary nerves and the anticipated movement in each eye. Disconjugacy of eye movements induced by individual vertical canal stimulation and dependence of the pulling direction of vertical recti and oblique muscles on eye position were also defined in these experiments. Subsequent studies have defined the mechanisms by which externally applied galvanic currents result in a change in vestibular-nerve afferent discharge. The currents appear to act at the spike trigger site. Perilymphatic cathodal currents depolarize the trigger site and lead to excitation whereas anodal currents hyperpolarize and result in inhibition. Afferents innervating all five vestibular endorgans appear to be affected equally by the currents (Goldberg et al., 1984). Irregularly discharging afferents are about 5-10 times more sensitive than regularly discharging ones because of the steeper slope of the former's faster postspike recovery of excitability in encoder sensitivity (Smith and Goldberg, 1986). Response adaptation similar to that noted during acceleration steps is apparent for longer periods of current administration. This adaptation is manifested as a perstimulus return toward resting discharge and poststimulus after-response in the opposite direction (Goldberg et al., 1984; Minor and Goldberg, l991). Cathodal currents (with respect to the perilymphatic space of the vestibule) are excitatory whereas anodal currents are inhibitory. Horizontal eye movements evoked by unilateral galvanic polarizations administered through chronically implanted labyrinthine stimulating electrodes have been studied in alert squirrel monkeys (Minor and Goldberg, 1991). We sought to extend this analysis by recording three-dimensional eye movements during galvanic stimulation. As predicted based upon roughly equal stimulation of ampullary nerves innervating the vertical canals, a substantial torsional component to the nystagmus is noted. The trajectory of torsional slow phases and nystagmus profile after the polarization provide insight into the central mechanisms that influence these responses.

Minor, Lloyd B.↗

Computation of propulsion-related flowfields using unstructured adaptive meshes

In this paper, we describe a computational fluid dynamics (CFD) technique based on unstructured triangular/tetrahedral meshes. A finite-volume scheme is used in conjunction with a multi-stage Runge-Kutta algorithm. Convergence enhancements in the form of dual time-stepping and time-derivative preconditioning are used to overcome the limitations of conventional multi-stage schemes. The method is applied to propulsion-related flows and shown to perform satisfactorily.

Weiss, Jonathan M.↗