SEARCH · Engineering Papers

Results for “adaptive methods”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,405 records · Page 78

All sources NASA NTRS DOE OSTI

DOE OSTI2024

Transition width of the of the $J$ $π$ = $1$ – two-phonon state of $88$ $Sr$

The ground-state decay width of the two-phonon $J$ $π$ = $1$ – , 4742 keV state of 88 Sr has been determined with the relative self-absorption method combined with a monoenergetic photon beam. This width is important to determine the decay transition strengths into the ground state and the one-phonon $2$$^{+}_{1}$ and $3$$^{–}_{1}$ levels which are required to verify the two-phonon character of the $J$ $π$ = $1$ – state. Here, the experiment was performed at the High Intensity γ-ray Source (HI⁢ γ⁢S) using a novel experimental approach to adapt the relative self-absorption method to monoenergetic photon beams. The result for the ground-state decay width is, thus, independent of any calibration standard and confirms the two-phonon character of the 88 Sr, $J$ $π$ = $1$ – , 4742 keV state within an improved uncertainty.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2023

Improving Deep Neural Networks’ Training for Image Classification With Nonlinear Conjugate Gradient-Style Adaptive Momentum

Momentum is crucial in stochastic gradient-based optimization algorithms for accelerating or improving training deep neural networks (DNNs). In deep learning practice, the momentum is usually weighted by a well-calibrated constant. However, tuning the hyperparameter for momentum can be a significant computational burden. In this article, we propose a novel adaptive momentum for improving DNNs training; this adaptive momentum, with no momentum-related hyperparame- ter required, is motivated by the nonlinear conjugate gradient (NCG) method. Stochastic gradient descent (SGD) with this new adaptive momentum eliminates the need for the momentum hyperparameter calibration, allows using a significantly larger learning rate, accelerates DNN training, and improves the final accuracy and robustness of the trained DNNs. For example, SGD with this adaptive momentum reduces classification errors for training ResNet110 for CIFAR10 and CIFAR100 from 5.25% to 4.64% and 23.75% to 20.03%, respectively. Furthermore, SGD, with the new adaptive momentum, also benefits adversarial training and, hence, improves the adversarial robustness of the trained DNNs.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS2010

Low field SQUID MRI devices, components and methods

Low field SQUID MRI devices, components and methods are disclosed. They include a portable low field (SQUID)-based MRI instrument and a portable low field SQUID-based MRI system to be operated under a bed where a subject is adapted to be located. Also disclosed is a method of distributing wires on an image encoding coil system adapted to be used with an NMR or MRI device for analyzing a sample or subject and a second order superconducting gradiometer adapted to be used with a low field SQUID-based MRI device as a sensing component for an MRI signal related to a subject or sample.

Penanen, Konstantin I.↗

NASA NTRS2011

Low field SQUID MRI devices, components and methods

Penanen, Konstantin I.↗

NASA NTRS2013

Low Field Squid MRI Devices, Components and Methods

Penanen, Konstantin I.↗

NASA NTRS2014

Low Field Squid MRI Devices, Components and Methods

Penanen, Konstantin I.↗

NASA NTRS1993

A direct viable count method for the enumeration of attached bacteria and assessment of biofilm disinfection

This report describes the adaptation of an in situ direct viable count (in situ DVC) method in biofilm disinfection studies. The results obtained with this technique were compared to two other enumeration methods, the plate count (PC) and conventional direct viable count (c-DVC). An environmental isolate (Klebsiella pneumoniae Kp1) was used to form biofilms on stainless steel coupons in a stirred batch reactor. The in situ DVC method was applied to directly assess the viability of bacteria in biofilms without disturbing the integrity of the interfacial community. As additional advantages, the results were observed after 4 h instead of the 24 h incubation time required for colony formation and total cell numbers that remained on the substratum were enumerated. Chlorine and monochloramine were used to determine the susceptibilities of attached and planktonic bacteria to disinfection treatment using this novel analytical approach. The planktonic cells in the reactor showed no significant change in susceptibility to disinfectants during the period of biofilm formation. In addition, the attached cells did not reveal any more resistance to disinfection than planktonic cells. The disinfection studies of young biofilms indicated that 0.25 mg/l free chlorine (at pH 7.2) and 1 mg/l monochloramine (at pH 9.0) have comparable disinfection efficiencies at 25 degrees C. Although being a weaker disinfectant, monochloramine was more effective in removing attached bacteria from the substratum than free chlorine. The in situ DVC method always showed at least one log higher viable cell densities than the PC method, suggesting that the in situ DVC method is more efficient in the enumeration of biofilm bacteria. The results also indicated that the in situ DVC method can provide more accurate information regarding the cell numbers and viability of bacteria within biofilms following disinfection.

NASA Discipline Number 04-10↗

DOE OSTI2020

A matrix-free approach for finite-strain hyperelastic problems using geometric multigrid

This paper investigates matrix-free algorithms for problems in quasi-static finite-strain hyperelasticity. Iterative solvers with matrix-free operator evaluation have emerged as an attractive alternative to sparse matrices in the fluid dynamics and wave propagation communities because they significantly reduce the memory traffic, the limiting factor in classical finite element solvers. Specifically, we study different matrix-free realizations of the finite element tangent operator and determine whether generalized methods of incorporating complex constitutive behavior might be feasible. Furthermore, in order to improve the convergence behavior of iterative solvers, we also propose a method by which to construct level tangent operators and employ them to define a geometric multigrid preconditioner. Additionally, the performance of the matrix-free operator and the geometric multigrid preconditioner is compared to the matrix-based implementation with an algebraic multigrid (AMG) preconditioner on a single node for a representative numerical example of a heterogeneous hyperelastic material in two and three dimensions. We find that matrix-free methods for finite-strain solid mechanics are very promising, outperforming linear matrix-based schemes by two to five times, and that it is possible to develop numerically efficient implementations that are independent of the hyperelastic constitutive law.

42 ENGINEERING↗

DOE OSTI2020

Whole program Adaptive Error Detection and Mitigation

Our project investigated methods to detect soft errors in computing systems. Such detection schemes must have low overheads and also generate no false positives. Our solutions focused on control-space error detection based on changing addressing schemes so that faults tend to cascade – rather than being masked. Our solutions also addressed data-space error detection by synthesizing error detectors based on rigorous floating-point error analysis.

42 ENGINEERING↗

NASA NTRS1971

Low-Energy Electron Transport with the Method of Discrete Ordinates

One-dimensional code adaptation for electron transport

Bartine, D. E.↗

DOE OSTI2021

Replicated Computational Results (RCR) Report for “Adaptive Precision Block-Jacobi for High Performance Preconditioning in the Ginkgo Linear Algebra Software”

The article by Flegar et al. titled “Adaptive Precision Block-Jacobi for High Performance Preconditioning in the Ginkgo Linear Algebra Software” presents a novel, practical implementation of an adaptive precision block-Jacobi preconditioner. Performance results using state-of-the-art GPU architectures for the block-Jacobi preconditioner generation and application demonstrate the practical usability of the method, compared to a traditional full-precision block-Jacobi preconditioner. A production-ready implementation is provided in the Ginkgo numerical linear algebra library. In this report, the Ginkgo library is reinstalled and performance results are generated to perform a comparison to the original results when using Ginkgo’s Conjugate Gradient solver with either the full or the adaptive precision block-Jacobi preconditioner for a suite of test problems on an NVIDIA GPU accelerator. After completing this process, the published results are deemed reproducible.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1991

Lattice Boltzmann model for simulation of magnetohydrodynamics

A numerical method, based on a discrete Boltzmann equation, is presented for solving the equations of magnetohydrodynamics (MHD). The algorithm provides advantages similar to the cellular automaton method in that it is local and easily adapted to parallel computing environments. Because of much lower noise levels and less stringent requirements on lattice size, the method appears to be more competitive with traditional solution methods. Examples show that the model accurately reproduces both linear and nonlinear MHD phenomena.

Chen, Shiyi↗

DOE OSTI2024

Integrating adaptive learning with post hoc model explanation and symbolic regression to build interpretable surrogate models

Abstract We develop a materials informatics workflow to build an interpretable surrogate model for micromagnetic simulations. Our goal is to predict the energy barrier of a moving isolated skyrmion in rare-earth-free $$\hbox {Mn}_4$$ Mn 4 N. Our approach integrates adaptive learning with post hoc model explanation and symbolic regression methods. We discuss an unexplored acquisition function (information condensing active learning) within the adaptive learning loop and compare it with the known standard deviation function for efficient navigation of the search space. Model-agnostic post hoc explanation techniques then uncover trends learned by the trained model, which we then leverage to constrain the expressions used for symbolic regression. Graphical abstract

Biswas, Ankita↗

DOE OSTI2022

Molecular Alignment of a Meta-Aramid on Carbon Nanotubes by In Situ Interfacial Polymerization

Molecularly organized nanocomposites of polymers and carbon nanotubes (CNTs) have great promise as high-performance materials; in particular, conformal deposition of polymers can control interfacial properties for mechanical load transfer, electrical or thermal transport, or electro/chemical transduction. However, controllability of polymer-CNT interaction remains a challenge with common processing methods that combine CNTs and polymers in melt or in solution, often leading to nonuniform polymer distribution and CNT aggregation. In this work, we demonstrate CNTs within net-shape sheets can be controllably coated with a conformal coating of meta-aramid by simultaneous capillary infiltration and interfacial polymerization. We determine that π-interaction between the polymer and CNTs results in chain alignment parallel to the CNT outer wall. Subsequent nucleation and growth of the precipitated aramid forms a smooth continuous layered sheath around the CNTs. These findings motivate future investigation of mechanical properties of the resulting composites, and adaptation of the in situ polymerization method to other substrates.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

NASA NTRSDate not supplied

Total Variation Majorization Minimization (TV-MM) Approach to Radiometer Brightness Temperature Gridding and Reconstruction

This paper presents the implementation of an algorithm to enhance the image resolution of the Earth's surface brightness temperature (T B ) data measured by radiometers such as the one onboard of the Soil Moisture Active Passive (SMAP) mission. A key step in radiometer T B processing is the conversion of the swath-based calibrated antenna temperature (T A ) measurements to the Level 3 Earth-centered grid. The simplest algorithm to transform this data from swath to gridded format is called drop-in-the-bucket which simply averages surrounding noisy T A samples to form a T B value at the gridded location. This method reduces noise, however produces low resolution products. To obtain a higher resolution product, SMAP uses other techniques such the Backus-Gilbert (BG) algorithm, which is the conventional method used in microwave radiometry. Although this method performs the required interpolation, it is not effective in denoising and removing blurring effects due to antenna filtering of the radiometer image data. Our motivation for this development is to further improve the resolution through post-processing of the radiometer T B image, a highly cost-effective method of image enhancement. The approach adapted in this work is based on the minimization of the Total Variation (TV) regularized objective function that is used extensively in solving general ill-posed linear inverse problems in image processing. Since the TV-based objective function is convex but not everywhere differentiable, there exists many numerical algorithms that can estimate the solution and the one selected for this work is called Majorization- Minimization (MM). By applying this algorithm, simulation experiments were performed based on synthetic data from the Geophysical model as well as real SMAP data to demonstrate the effectiveness of the technique. Results were then compared against the BG method.

Wing Lee↗

NASA NTRS1990

Solution adaptive local rectangular grid refinement for transonic aerodynamic flow problems

This paper describes the use of solution-adaptive local grid refinement in a numerical method for solving transonic flow problems about complex three-dimensional aircraft configurations. The method is implemented in the TRANAIR code, which has been applied to help solve many practical engineering problems. Attention is focused here on the principal components of the solution-adaptive grid algorithms currently being developed and on two applications that demonstrate the capabilities of the algorithms.

Bieterman, Michael B.↗

NASA NTRS1992

Adaptation of a matrix parameterization of the middle atmospheric radiative cooling for an arbitrary vertical coordinate grid

Matrices corresponding to a revised vertical grid are shown to be derived from the original grid by integrating the radiative cooling rate over model layers. The method is applied to temperature profiles that represent average or extreme atmospheric conditions and a range of vertical resolutions. The example makes use of the 15-micron CO2 band parameterization proposed by Akmaev and Shved (1982) for the LTE region of the middle atmosphere. Cooling rate errors provided by the present method are not more than 0.11 K/day for a vertical grid spacing of less than 5 km. The matrix transformation method is shown to be valuable for adapting effective matrix parameterizations of middle atmosphere cooling to atmospheric models with arbitrary vertical resolutions.

Akmaev, R. A.↗

NASA NTRS2013

Final Report - Regulatory Considerations for Adaptive Systems

This report documents the findings of a preliminary research study into new approaches to the software design assurance of adaptive systems. We suggest a methodology to overcome the software validation and verification difficulties posed by the underlying assumption of non-adaptive software in the requirementsbased- testing verification methods in RTCA/DO-178B and C. An analysis of the relevant RTCA/DO-178B and C objectives is presented showing the reasons for the difficulties that arise in showing satisfaction of the objectives and suggested additional means by which they could be satisfied. We suggest that the software design assurance problem for adaptive systems is principally one of developing correct and complete high level requirements and system level constraints that define the necessary system functional and safety properties to assure the safe use of adaptive systems. We show how analytical techniques such as model based design, mathematical modeling and formal or formal-like methods can be used to both validate the high level functional and safety requirements, establish necessary constraints and provide the verification evidence for the satisfaction of requirements and constraints that supplements conventional testing. Finally the report identifies the follow-on research topics needed to implement this methodology.

Wilkinson, Chris↗