Search NASA⌕ Search

SEARCH · Search NASA

Results for “Edge Based”

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 73 records · Page 4
All sourcesNASA NTRSDOE OSTI
NASA NTRS1993

Development of an unstructured solution adaptive method for the quasi-three-dimensional Euler and Navier-Stokes equations

A general solution adaptive scheme based on a remeshing technique is developed for solving the two-dimensional and quasi-three-dimensional Euler and Favre-averaged Navier-Stokes equations. The numerical scheme is formulated on an unstructured triangular mesh utilizing an edge-based pointer system which defines the edge connectivity of the mesh structure. Jameson's four-stage hybrid Runge-Kutta scheme is used to march the solution in time. The convergence rate is enhanced through the use of local time stepping and implicit residual averaging. As the solution evolves, the mesh is regenerated adaptively using flow field information. Mesh adaptation parameters are evaluated such that an estimated local numerical error is equally distributed over the whole domain. For inviscid flows, the present approach generates a complete unstructured triangular mesh using the advancing front method. For turbulent flows, the approach combines a local highly stretched structured triangular mesh in the boundary layer region with an unstructured mesh in the remaining regions to efficiently resolve the important flow features. One-equation and two-equation turbulence models are incorporated into the present unstructured approach. Results are presented for a wide range of flow problems including two-dimensional multi-element airfoils, two-dimensional cascades, and quasi-three-dimensional cascades. This approach is shown to gain flow resolution in the refined regions while achieving a great reduction in the computational effort and storage requirements since solution points are not wasted in regions where they are not required.

Jiang, Yi-Tsann↗
NASA NTRS1993

Development of an unstructured solution adaptive method for the quasi-three-dimensional Euler and Navier-Stokes equations

A general solution adaptive scheme-based on a remeshing technique is developed for solving the two-dimensional and quasi-three-dimensional Euler and Favre-averaged Navier-Stokes equations. The numerical scheme is formulated on an unstructured triangular mesh utilizing an edge-based pointer system which defines the edge connectivity of the mesh structure. Jameson's four-stage hybrid Runge-Kutta scheme is used to march the solution in time. The convergence rate is enhanced through the use of local time stepping and implicit residual averaging. As the solution evolves, the mesh is regenerated adaptively using flow field information. Mesh adaptation parameters are evaluated such that an estimated local numerical error is equally distributed over the whole domain. For inviscid flows, the present approach generates a complete unstructured triangular mesh using the advancing front method. For turbulent flows, the approach combines a local highly stretched structured triangular mesh in the boundary layer region with an unstructured mesh in the remaining regions to efficiently resolve the important flow features. One-equation and two-equation turbulence models are incorporated into the present unstructured approach. Results are presented for a wide range of flow problems including two-dimensional multi-element airfoils, two-dimensional cascades, and quasi-three-dimensional cascades. This approach is shown to gain flow resolution in the refined regions while achieving a great reduction in the computational effort and storage requirements since solution points are not wasted in regions where they are not required.

Jiang, Yi-Tsann↗
DOE OSTI2024

Radiation-Induced Noise Resilience of Neuromorphic Architectures

Neuromorphic event-based networks use asynchronous time-dependent information to extract features from input data that can allow for edge-based distributed applications such as object recognition. The noise resilience properties of such networks, especially in the context of space applications, are yet to be explored. In this paper, we use the hierarchy of time surfaces (HOTS) algorithm, which is one of the neuromorphic algorithms, to understand the least and most resilient modules in a neuromorphic network. The HOTS algorithm relies on the computing of time surfaces that maps the temporal delays between neighboring pixels into normalized features that involve many computations that are also found in other neuromorphic networks such as exponential decays, distance computations, etcetera. We implemented HOTS on a Digilent PYNQ board with a Xilinx Zynq 7020 system on a chip, and we subjected the boards running the HOTS network inference to neutron radiation at the Los Alamos Neutron Science Center. Furthermore, we used simulation models from our previous similar experiments on the event-based sensor to create a neutron induced noise model to quantify the effect of this noise on the overall performance of the network. This experiment provides the preliminary measurements of the reliability of the HOTS algorithm and proposes methods to create a more reliable HOTS architecture in future spacecraft missions.

Engineering↗
DOE OSTI2019

O3BNN: An Out-Of-Order Architecture for High-Performance Binarized Neural Network Inference with Fine-Grained Pruning

In this work, we demonstrate that the highly-condensed BNN model can be shrunk significantly further by dynamically pruning irregular redundant edges. Based on two new observations on BNN-specific properties, our out-of-order (OoO) architecture – O3BNN, can curtail remaining edge evaluation in cases where the binary output of a neuron can be determined early. Similar to Instruction-Level-Parallelism (ILP), these fine-grained, irregular, runtime pruning opportunities are traditionally presumed to be di cult to exploit. We evaluate our design on an FPGA platform using three well-known networks, including VggNet-16, AlexNet for ImageNet, and a VGG-like network for Cifar-10. Results show that our out-of-order approach can prune 27%, 16%, and 42% of the operations for the three networks respectively, without any accuracy loss, leading to at least 1.7×, 1.5×, and 2.1× speedups over state-of-the-art implementations on FPGA/GPU/CPU BNN implementations. Our approach is inference runtime pruning, so no retrain or fine-tuning is needed. We demonstrate our design on an FPGA platform. However, this is only for showcasing the method; the approach does not rely on any FPGA-specific features and can thus be adopted by other devices as well.

Geng, Tong↗
DOE OSTI2021

O3BNN-R: An Out-Of-Order Architecture for High-Performance and Regularized BNN Inference

Binarized Neural Networks (BNN) have drawn tremendous attention due to significantly reduced computational complexity and memory demand. They have especially shown great potential in cost- and power-restricted domains, such as IoT and smart edge-devices, where reaching a certain accuracy bar is often sufficient, and real-time is highly desired.In this work, we demonstrate that the highly-condensed BNN model can be shrunk significantly further by dynamically pruning irregular redundant edges. Based on two new observations on BNN-specific properties, an out-of-order (OoO) architecture – O3BNN-R, can curtail edge evaluation in cases where the binary output of a neuron can be determined early. Similar to Instruction-Level-Parallelism(ILP), these fine-grained, irregular, runtime pruning opportunities are traditionally presumed to be difficult to exploit. In order to increase the pruning opportunities, we also optimize the training process by adding 2 regularization items in the loss function (1) for pooling pruning and (2) for threshold pruning. We evaluate our design on an FPGA platform using three well-known networks, including VggNet-16, AlexNet for ImageNet, and a VGG-like network for Cifar-10.

Geng, Tong↗
NASA NTRS1998

Capacitively Stabilized Etalon Technology for Spaceborne Wind Lidar Application

Global monitoring by satellites is important for many types of environmental studies. Among these, the wind field is the single most important atmospheric state variable required for understanding atmospheric motion and predicting weather. Studies indicate that a global determination of the tropospheric wind field to an accuracy of 1-5 m/sec is critical for improved numerical weather forecasting. This measurement could be carried out with a spaceborne lidar system sensing the Doppler shift of a laser signal backscattered from the atmosphere. Over the past four years we have developed a ground-based Edge Technique lidar system and demonstrated wind measurements through the troposphere with high accuracy (1-6 m/s) and high spatial resolution. Recently, we began the design of a shuttle based wind measurement system for technology demonstration as part of the Zephyr program. In this paper, we present the characteristics of the high spectral resolution etalon filter technology for lidar wind measurement with the Edge Technique. In conclusion, a spacecraft motion compensation method has been discussed. We demonstrated experimentally that the etalon can be tuned to compensate for any satellite induced Doppler shift to an accuracy of better than 0.6 m/s over a time period of 2.5 ms.

Li, Steven X.↗
NASA NTRS1993

Dynamic mesh adaption for triangular and tetrahedral grids

The following topics are discussed: requirements for dynamic mesh adaption; linked-list data structure; edge-based data structure; adaptive-grid data structure; three types of element subdivision; mesh refinement; mesh coarsening; additional constraints for coarsening; anisotropic error indicator for edges; unstructured-grid Euler solver; inviscid 3-D wing; and mesh quality for solution-adaptive grids. The discussion is presented in viewgraph form.

Rupak Biswas↗
NASA NTRSDate not supplied

An Automated Marching Scheme for Overset Structured Surface Mesh Generation

Starting with a Boundary Representation (BRep) of the geometry of an aerospace vehicle, an automated marching scheme is presented for generation of structured overset surface meshes. First, a pre-processing step automatically generates discrete representations of the BRep faces and BRep edges by tessellating in parameter space. Topological connectivity between the discretized BRep edges is then established, followed by automatic grid point distribution on these edges based on local turning angle, proximity to sharp geometric features, and prescribed maximum stretching ratio and grid spacing. A set of initial curves for algebraic or hyperbolic marching on a surface is then derived from the redistributed edge curves. A spatially-variable marching distance together with a grid point distribution in the marching direction are automatically determined for each initial curve. A set of overset surface meshes that covers the entire geometry is then obtained by combining the surface meshes around the BRep edges, and the structured meshes derived from the discretized BRep faces

Shishir A Pandya↗
NASA NTRS1991

Modes of Visual Recognition and Perceptually Relevant Sketch-based Coding for Images

A review of visual recognition studies is used to define two levels of information requirements. These two levels are related to two primary subdivisions of the spatial frequency domain of images and reflect two distinct different physical properties of arbitrary scenes. In particular, pathologies in recognition due to cerebral dysfunction point to a more complete split into two major types of processing: high spatial frequency edge based recognition vs. low spatial frequency lightness (and color) based recognition. The former is more central and general while the latter is more specific and is necessary for certain special tasks. The two modes of recognition can also be distinguished on the basis of physical scene properties: the highly localized edges associated with reflectance and sharp topographic transitions vs. smooth topographic undulation. The extreme case of heavily abstracted images is pursued to gain an understanding of the minimal information required to support both modes of recognition. Here the intention is to define the semantic core of transmission. This central core of processing can then be fleshed out with additional image information and coding and rendering techniques.

Jobson, Daniel J.↗
DOE OSTI2018

Discontinuous Finite Element Quasidiffusion Methods

Here in this paper, two-level methods for solving transport problems in one-dimensional slab geometry based on the quasi-diffusion (QD) method are developed. A linear discontinuous finite element method (LDFEM) is derived for the spatial discretization of the low-order QD (LOQD) equations. It involves special interface conditions at the cell edges based on the idea of QD boundary conditions (BCs). We consider different kinds of QD BCs to formulate the necessary cell-interface conditions. We develop two-level methods with independent discretization of the high-order transport equation and LOQD equations, where the transport equation is discretized using the method of characteristics and the LDFEM is applied to the LOQD equations. We also formulate closures that lead to the discretization consistent with a LDFEM discretization of the transport equation. The proposed methods are studied by means of test problems formulated with the method of manufactured solutions. Numerical experiments are presented demonstrating the performance of the proposed methods. Lastly, we also show that the method with independent discretization has the asymptotic diffusion limit.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2019

Area-Efficient, 600V 4H-SiC JBS Diode-Integrated MOSFETs (JBSFETs) for Power Converter Applications

This article reports the demonstration of a low-voltage (<; 600V) monolithically integrated 4H-silicon carbide (SiC) MOSFET and JBS diode (JBSFET). A singlemetal and thermal treatment process were implemented to form ohmic contacts on the n+ and p+ source regions while forming the Schottky contact on the N- SiC epitaxial layer. Different layout methodologies are discussed for fabricating an energy-efficient low-voltage JBSFET by intermittently placing the JBS diode portion in the orthogonal direction to minimize the device area, hence improving the specific ON-resistance and reducing the overall chip size by 46%. A junction termination extension (JTE)-based edge termination structure (the Hybrid-JTE) was implemented to achieve a high breakdown voltage with a very low leakage current. In addition, it was investigated that the forward characteristic of the JBSFET can be further improved by adopting Ti-based metal as the Schottky contact for the JBS diode. Device design, layout approach, fabrication, electrical characterization, and future prospects of the 4H-SiC JBSFETs are discussed in this article.

42 ENGINEERING↗
DOE OSTI2018

In situ edge engineering in two-dimensional transition metal dichalcogenides

Exerting synthetic control over the edge structure and chemistry of two-dimensional (2D) materials is of critical importance to direct the magnetic, optical, electrical, and catalytic properties for specific applications. Here in this study, we directly image the edge evolution of pores in Mo 1-x W x Se 2 monolayers via atomic-resolution in situ scanning transmission electron microscopy (STEM) and demonstrate that these edges can be structurally transformed to theoretically predicted metastable atomic configurations by thermal and chemical driving forces. Density functional theory calculations and ab initio molecular dynamics simulations explain the observed thermally induced structural evolution and exceptional stability of the four most commonly observed edges based on changing chemical potential during thermal annealing. Finally, the coupling of modeling and in situ STEM imaging in changing chemical environments demonstrated here provides a pathway for the predictive and controlled atomic scale manipulation of matter for the directed synthesis of edge configurations in Mo 1-x W x Se 2 to achieve desired functionality.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Fast Adaptive Neural Control of Resonant Extraction at Fermilab

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

Whitbeck, A. [Fermilab] (ORCID:0000000342245164)↗
DOE OSTI2025

Fast Adaptive Neural Control of Resonant Extraction at Fermilab

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

Whitbeck, A. [Fermilab]↗
DOE OSTI2026

Towards Secure Autonomous Vehicles: An Integrated Edge and Multi-Modal Machine Learning Framework for Intrusion Detection

Autonomous vehicles (AVs) are vulnerable to cyberattacks targeting both internal communication networks and external perception sensors. While edge-based intrusion de- tection for Controller Area Network (CAN) buses offers real-time protection, it cannot detect cross-modal threats. Conversely, multi-modal fusion approaches improve coverage but often lack efficiency for in-vehicle deployment. This thesis integrates two complemen- tary solutions: (1) a lightweight, edge-deployable machine learning framework for CAN bus intrusion detection, and (2) a late-fusion system combining CAN FD and LiDAR data. Together, they form a hierarchical defense capable of handling single-modality and coordi- nated attacks. Simulations show that CAN-only models reach 93% accuracy on simulated DoS, spoofing, replay, and fuzzy attacks, while the fusion system achieves 0.87 AUC and 0.82 F1-score at 2 ms latency. This unified framework establishes a scalable, explainable, and field-ready strategy for AV cybersecurity.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1996

Parallel Implementation of an Adaptive Scheme for 3D Unstructured Grids on the SP2

Dynamic mesh adaption on unstructured grids is a powerful tool for computing unsteady flows that require local grid modifications to efficiently resolve solution features. For this work, we consider an edge-based adaption scheme that has shown good single-processor performance on the C90. We report on our experience parallelizing this code for the SP2. Results show a 47.OX speedup on 64 processors when 10% of the mesh is randomly refined. Performance deteriorates to 7.7X when the same number of edges are refined in a highly-localized region. This is because almost all mesh adaption is confined to a single processor. However, this problem can be remedied by repartitioning the mesh immediately after targeting edges for refinement but before the actual adaption takes place. With this change, the speedup improves dramatically to 43.6X.

Oliker, Leonid↗
NASA NTRS1996

Parallel implementation of an adaptive scheme for 3D unstructured grids on the SP2

Dynamic mesh adaption on unstructured grids is a powerful tool for computing unsteady flows that require local grid modifications to efficiently resolve solution features. For this work, we consider an edge-based adaption scheme that has shown good single-processor performance on the C90. We report on our experience parallelizing this code for the SP2. Results show a 47.0X speedup on 64 processors when 10 percent of the mesh is randomly refined. Performance deteriorates to 7.7X when the same number of edges are refined in a highly-localized region. This is because almost all the mesh adaption is confined to a single processor. However, this problem can be remedied by repartitioning the mesh immediately after targeting edges for refinement but before the actual adaption takes place. With this change, the speedup improves dramatically to 43.6X.

Strawn, Roger C.↗
NASA NTRS1999

Parallel Tetrahedral Mesh Adaptation with Dynamic Load Balancing

The ability to dynamically adapt an unstructured grid is a powerful tool for efficiently solving computational problems with evolving physical features. In this paper, we report on our experience parallelizing an edge-based adaptation scheme, called 3D_TAG. using message passing. Results show excellent speedup when a realistic helicopter rotor mesh is randomly refined. However. performance deteriorates when the mesh is refined using a solution-based error indicator since mesh adaptation for practical problems occurs in a localized region., creating a severe load imbalance. To address this problem, we have developed PLUM, a global dynamic load balancing framework for adaptive numerical computations. Even though PLUM primarily balances processor workloads for the solution phase, it reduces the load imbalance problem within mesh adaptation by repartitioning the mesh after targeting edges for refinement but before the actual subdivision. This dramatically improves the performance of parallel 3D_TAG since refinement occurs in a more load balanced fashion. We also present optimal and heuristic algorithms that, when applied to the default mapping of a parallel repartitioner, significantly reduce the data redistribution overhead. Finally, portability is examined by comparing performance on three state-of-the-art parallel machines.

Oliker, Leonid↗