Search NASA⌕ Search

SEARCH · Search NASA

Results for “Radial basis function”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
NASA NTRS2007

Reconfigurable Control Design with Neural Network Augmentation for a Modified F-15 Aircraft

The viewgraphs present background information about reconfiguration control design, design methods used for paper, control failure survivability results, and results and time histories of tests. Topics examined include control reconfiguration, general information about adaptive controllers, model reference adaptive control (MRAC), the utility of neural networks, radial basis functions (RBF) neural network outputs, neurons, and results of investigations of failures.

Burken, John J.↗
NASA NTRS2011

Development of the Orion Crew Module Static Aerodynamic Database: Hypersonic - Part 1

The Orion aerodynamic database provides force and moment coefficients given the velocity, attitude, configuration, etc. of the Crew Exploration Vehicle (CEV). The database is developed and maintained by the NASA CEV Aerosciences Project team from computational and experimental aerodynamic simulations. The database is used primarily by the Guidance, Navigation, and Control (GNC) team to design vehicle trajectories and assess flight performance. The initial hypersonic re-entry portion of the Crew Module (CM) database was developed in 2006. Updates incorporating additional data and improvements to the database formulation and uncertainty methodologies have been made since then. This paper details the process used to develop the CM database, including nominal values and uncertainties, for Mach numbers greater than 8 and angles of attack between 140deg and 180deg. The primary available data are more than 1000 viscous, reacting gas chemistry computational simulations using both the Laura and Dplr codes, over a range of Mach numbers from 2 to 37 and a range of angles of attack from 147deg to 172deg. Uncertainties were based on grid convergence, laminar-turbulent solution variations, combined altitude and code-to-code variations, and expected heatshield asymmetry. A radial basis function response surface tool, NEAR-RS, was used to fit the coefficient data smoothly in a velocity-angle-of-attack space. The resulting database is presented and includes some data comparisons and a discussion of the predicted variation of trim angle of attack and lift-to-drag ratio. The database provides a variation in trim angle of attack on the order of +/-2deg, and a range in lift-to-drag ratio of +/-0.035 for typical vehicle flight conditions.

Bibb, Karen L.↗
NASA NTRS2012

An Algorithm for Detection of Ground and Canopy Cover in Micropulse Photon-Counting Lidar Altimeter Data in Preparation of the ICESat-2 Mission

The Ice, Cloud and Land Elevation Satellite-II (ICESat-2) mission has been selected by NASA as a Decadal Survey mission, to be launched in 2016. Mission objectives are to measure land ice elevation, sea ice freeboard/ thickness and changes in these variables and to collect measurements over vegetation that will facilitate determination of canopy height, with an accuracy that will allow prediction of future environmental changes and estimation of sea-level rise. The importance of the ICESat-2 project in estimation of biomass and carbon levels has increased substantially, following the recent cancellation of all other planned NASA missions with vegetation-surveying lidars. Two innovative components will characterize the ICESat-2 lidar: (1) Collection of elevation data by a multi-beam system and (2) application of micropulse lidar (photon counting) technology. A micropulse photon-counting altimeter yields clouds of discrete points, which result from returns of individual photons, and hence new data analysis techniques are required for elevation determination and association of returned points to reflectors of interest including canopy and ground in forested areas. The objective of this paper is to derive and validate an algorithm that allows detection of ground under dense canopy and identification of ground and canopy levels in simulated ICESat-2-type data. Data are based on airborne observations with a Sigma Space micropulse lidar and vary with respect to signal strength, noise levels, photon sampling options and other properties. A mathematical algorithm is developed, using spatial statistical and discrete mathematical concepts, including radial basis functions, density measures, geometrical anisotropy, eigenvectors and geostatistical classification parameters and hyperparameters. Validation shows that the algorithm works very well and that ground and canopy elevation, and hence canopy height, can be expected to be observable with a high accuracy during the ICESat-2 mission. A result relevant for instrument design is that even the two weaker beam classes considered can be expected to yield useful results for vegetation measurements (93.01-99.57% correctly selected points for a beam with expected return of 0.93 mean signals per shot (msp9) and 72.85% - 98.68% for 0.48 msp (msp4)). Resampling options affect results more than noise levels. The algorithm derived here is generally applicable for analysis of micropulse lidar altimeter data collected over forested areas as well as other surfaces, including land ice, sea ice and land surfaces.

Herzfeld, Ute C.↗
NASA NTRS2012

Reduced-Order Modeling for Flutter/LCO Using Recurrent Artificial Neural Network

The present study demonstrates the efficacy of a recurrent artificial neural network to provide a high fidelity time-dependent nonlinear reduced-order model (ROM) for flutter/limit-cycle oscillation (LCO) modeling. An artificial neural network is a relatively straightforward nonlinear method for modeling an input-output relationship from a set of known data, for which we use the radial basis function (RBF) with its parameters determined through a training process. The resulting RBF neural network, however, is only static and is not yet adequate for an application to problems of dynamic nature. The recurrent neural network method [1] is applied to construct a reduced order model resulting from a series of high-fidelity time-dependent data of aero-elastic simulations. Once the RBF neural network ROM is constructed properly, an accurate approximate solution can be obtained at a fraction of the cost of a full-order computation. The method derived during the study has been validated for predicting nonlinear aerodynamic forces in transonic flow and is capable of accurate flutter/LCO simulations. The obtained results indicate that the present recurrent RBF neural network is accurate and efficient for nonlinear aero-elastic system analysis

Yao, Weigang↗
NASA NTRS2014

Algorithm for Detection of Ground and Canopy Cover in Micropulse Photon-Counting Lidar Altimeter Data in Preparation for the ICESat-2 Mission

NASA's Ice, Cloud and Land Elevation Satellite-II (ICESat-2) mission is a decadal survey mission (2016 launch). The mission objectives are to measure land ice elevation, sea ice freeboard, and changes in these variables, as well as to collect measurements over vegetation to facilitate canopy height determination. Two innovative components will characterize the ICESat-2 lidar: 1) collection of elevation data by a multibeam system and 2) application of micropulse lidar (photon-counting) technology. A photon-counting altimeter yields clouds of discrete points, resulting from returns of individual photons, and hence new data analysis techniques are required for elevation determination and association of the returned points to reflectors of interest. The objective of this paper is to derive an algorithm that allows detection of ground under dense canopy and identification of ground and canopy levels in simulated ICESat-2 data, based on airborne observations with a Sigma Space micropulse lidar. The mathematical algorithm uses spatial statistical and discrete mathematical concepts, including radial basis functions, density measures, geometrical anisotropy, eigenvectors, and geostatistical classification parameters and hyperparameters. Validation shows that ground and canopy elevation, and hence canopy height, can be expected to be observable with high accuracy by ICESat-2 for all expected beam energies considered for instrument design (93.01%-99.57% correctly selected points for a beam with expected return of 0.93 mean signals per shot (msp), and 72.85%-98.68% for 0.48 msp). The algorithm derived here is generally applicable for elevation determination from photoncounting lidar altimeter data collected over forested areas, land ice, sea ice, and land surfaces, as well as for cloud detection.

Algorithms↗
NASA NTRS2013

Nonlinear Modeling by Assembling Piecewise Linear Models

To preserve nonlinearity of a full order system over a parameters range of interest, we propose a simple modeling approach by assembling a set of piecewise local solutions, including the first-order Taylor series terms expanded about some sampling states. The work by Rewienski and White inspired our use of piecewise linear local solutions. The assembly of these local approximations is accomplished by assigning nonlinear weights, through radial basis functions in this study. The efficacy of the proposed procedure is validated for a two-dimensional airfoil moving at different Mach numbers and pitching motions, under which the flow exhibits prominent nonlinear behaviors. All results confirm that our nonlinear model is accurate and stable for predicting not only aerodynamic forces but also detailed flowfields. Moreover, the model is robustness-accurate for inputs considerably different from the base trajectory in form and magnitude. This modeling preserves nonlinearity of the problems considered in a rather simple and accurate manner.

Yao, Weigang↗
NASA NTRS2016

High Resolution Image Reconstruction from Projection of Low Resolution Images DIffering in Subpixel Shifts

In this paper, we demonstrate a simple algorithm that projects low resolution (LR) images differing in subpixel shifts on a high resolution (HR) also called super resolution (SR) grid. The algorithm is very effective in accuracy as well as time efficiency. A number of spatial interpolation techniques using nearest neighbor, inverse-distance weighted averages, Radial Basis Functions (RBF) etc. used in projection yield comparable results. For best accuracy of reconstructing SR image by a factor of two requires four LR images differing in four independent subpixel shifts. The algorithm has two steps: i) registration of low resolution images and (ii) shifting the low resolution images to align with reference image and projecting them on high resolution grid based on the shifts of each low resolution image using different interpolation techniques. Experiments are conducted by simulating low resolution images by subpixel shifts and subsampling of original high resolution image and the reconstructing the high resolution images from the simulated low resolution images. The results of accuracy of reconstruction are compared by using mean squared error measure between original high resolution image and reconstructed image. The algorithm was tested on remote sensing images and found to outperform previously proposed techniques such as Iterative Back Projection algorithm (IBP), Maximum Likelihood (ML), and Maximum a posterior (MAP) algorithms. The algorithm is robust and is not overly sensitive to the registration inaccuracies.

0↗
NASA NTRS2017

Surface-Height Determination of Crevassed Glaciers-Mathematical Principles of an Autoadaptive Density-Dimension Algorithm and Validation Using ICESat-2 Simulator (SIMPL) Data

Glacial acceleration is a main source of uncertainty in sea-level-change assessment. Measurement of ice-surface heights with a spatial and temporal resolution that not only allows elevation-change calculation, but also captures ice-surface morphology and its changes is required to aid in investigations of the geophysical processes associated with glacial acceleration.The Advanced Topographic Laser Altimeter System aboard NASAs future ICESat-2 Mission (launch 2017) will implement multibeam micropulse photon-counting lidar altimetry aimed at measuring ice-surface heights at 0.7-m along-track spacing. The instrument is designed to resolve spatial and temporal variability of rapidly changing glaciers and ice sheets and the Arctic sea ice. The new technology requires the development of a new mathematical algorithm for the retrieval of height information.We introduce the density-dimension algorithm (DDA) that utilizes the radial basis function to calculate a weighted density as a form of data aggregation in the photon cloud and considers density an additional dimension as an aid in auto-adaptive threshold determination. The auto-adaptive capability of the algorithm is necessary to separate returns from noise and signal photons under changing environmental conditions. The algorithm is evaluated using data collected with an ICESat-2 simulator instrument, the Slope Imaging Multi-polarization Photon-counting Lidar, over the heavily crevassed Giesecke Braer in Northwestern Greenland in summer 2015. Results demonstrate that ICESat-2 may be expected to provide ice-surface height measurements over crevassed glaciers and other complex ice surfaces. The DDA is generally applicable for the analysis of airborne and spaceborne micropulse photon-counting lidar data over complex and simple surfaces.

Herzfeld, Ute C.↗
NASA NTRSDate not supplied

Clutter Assessment for an Autonomous Multi-Agent Search Mission

This paper presents a method for evaluating the amount of clutter in a region where autonomous vehicles in a multi-agent system must operate based on LIDAR point cloud measurements. The point cloud is used to generate an occupancy grid which is then projected onto a 2D plane of vehicle motion, constituting an image. A series of Gaussian radial basis functions (GRBFs) is created, each centered at an occupied pixel in a 2D image, and summed together to form the clutter field. The clutter field is a representation of the density and permeability of the space at each coordinate. The clutter field is then approximated such that iso-clutter contours are simple geometric objects so that intelligent machine assets can easily query the distance between them and any given point in an environment. In this way, agents are able to determine whether to enter into or steer away from areas of interest. Each vehicle has a clutter threshold representing the clutter value of the space in which it can safely maneuver. The iso-clutter contour corresponding to a vehicle’s clutter threshold is treated as the boundary of an obstacle to be avoided. A simulation is presented where a multi-agent system is tasked with persistent observation of a cluttered area. Each vehicle in the simulation has a different clutter threshold. The vehicles use a potential field-based guidance algorithm, and an allocation of vehicles to specific regions of the space emerges.

multi-agent↗
NASA NTRSDate not supplied

Machine Learning Explainability and Transferability for Path Navigation

Deep neural networks are powerful tools for machine perception. Unfortunately their decisions are difficult to explain due to the complexity and size of the networks. Previously we have alleviated this issue by using the representational portion of a deep neural network and combining it with a k-nearest neighbor (KNN) classifier. Through inspection of the decisions made by the KNN, we can directly see the training data responsible for the decisions, allowing us to determine the quality of the overall decision and the quality of the representational layer of the deep NN. While the technique worked well, it requires tens of thousands of latent vectors to be stored for classification. In addition, it lacks the ability to show how parts of an image influence the classification decision. Here we address these issues by 1) Using a radial basis function network (RBFN) in place of the KNN allowing far fewer images to be used in deployment and 2) Using an auto encoder network for explainability. In addition to these techniques, we examine the effects of transfer learning to determine that results are robust. All results are tested on a domain where an unmanned aerial vehicle (UAV) navigates a forest trail through a single camera.

explainable machine learning↗
NASA NTRSDate not supplied

Mesh Deformation Boundary Conditions for Three-Dimensional Ablation Solvers

A method for determining boundary node displacements for three-dimensional ablation solvers is developed. The approach is applicable to a wide range of mesh motion algorithms used to determine internal node displacements. The method leverages radial basis functions to analytically define the geometry to which the boundary nodes should be constrained. Constrained optimization, elasticity, and spring analogy solvers are utilized to redistribute the nodes on the evolved geometry surface while maintaining mesh quality. The method is implemented in the CHarring Ablator Response (CHAR) code, and a realistic demonstration case is presented on the Boundary Layer Transition (BOLT) flight experiment configuration. The method is shown to be robust for highly-curved geometries with large domain deformations.

Adam J Amar↗
NASA NTRSDate not supplied

Spacesuit and Mobility Performance Changes

The complex interactions between the human body and spacesuit lead to changes inmovement patterns and mobility performancesof the wearer. In general, factors including the geometric properties, such as shape and size, mechanical properties of the suit, and pressurization of the suit are known to be associated with altered movement patterns as compared to an unsuited human.However, their relative contributions have not been explicitly quantified from the mobility performance perspectives. The goal of this study wasthus to assessthe effects fromthe different types of mobility constraintconditions, namely by wearing either a 3D printed hard upper torso (HUT) assembly orfully pressurized spacesuit. The outcome was also compared against the unsuited motions. For this study, an xEMU (exploration Extravehicular Mobility Unit) suitwas considered, which is the next generation spacesuit developed by NASA. In each test condition, the subject was asked to move the arm and hand as prescribedfor different task types, and the corresponding body segment locations were recordedusing a 3D motion capture system. The following three tasks were performedand analyzed: 1) Outward one-handed reaches:the subject in a standing pose made sweeping motions with the extended right arm from the extreme end-to-end positions, including side-to-side at different elevations and top-to-bottom at different azimuths. 2) Outward two-handed reaches:similarto the previous task, howeverthe subject kept the hands together during the motions in order to assessthe areas that can be reached by both hands. 3) Inward one-handed reaches:the subject made right-hand reach motions to the surface of the HUT.The hand traces collected from each task were modeled by a template shape parametrically deformed with a radial basis function. This process enabled foran abstraction of the hand traces into a smooth surface envelope representing the maximally reachable area of the test subject, of which the shapes and sizes were compared across the different test conditions. The preliminary analysis has shown that theoverall size of reachenvelopes decreases in a pressurized suit compared to 3D printed mockup HUT and unsuited conditions. The specific shape of the envelopes, which were determined by the reachable and unreachable zones, alsovary with the testconditions, and the differences werepronounced with the inward reaches to the HUT surface. The latter observation ispotentially relatedto the increased demandfor shoulder and elbow flexions.Overcomingthe resistance from the pressurizedsoft goods and mechanical constraints of the shoulder assemblywas seen to be associated with the difference in motion patterns between the suited and unsuited conditions. Overall, the information quantified from this study is expected to provide structured metrics for spacesuit mobility, which can improve design optimization and human-system integration.

K Han Kim↗
NASA NTRSDate not supplied

Alteration of Optical Properties in Polymer Materials under Low Earth Orbit (LEO) Space and Simulated Exposure

This paper provides an overview of the 16th Materials International Space Station Experiment (MISSE) project, conducted by a multiorganizational team of researchers. The orbital environment during the mission, including environmental sensors data and unexpected occurrences, are discussed. The research further examines the spectral changes in selected material, Kapton TF, exposed to low Earth orbit conditions during the mission. Optical images of the material taken during the mission were analyzed using a machine learning approach. This process involved the use of a Radial Basis Function network to extract reflectance spectra from RGB/IR images. The findings were then compared with lab-based experiments of identical materials subjected to simulated space weather conditions. This comprehensive study aids in making accurate predictions about the performance of materials in space, thereby improving the reliability and safety of future space missions.

Insulators↗
NASA NTRS2016

Simple Test Functions in Meshless Local Petrov-Galerkin Methods

Two meshless local Petrov-Galerkin (MLPG) methods based on two different trial functions but that use a simple linear test function were developed for beam and column problems. These methods used generalized moving least squares (GMLS) and radial basis (RB) interpolation functions as trial functions. These two methods were tested on various patch test problems. Both methods passed the patch tests successfully. Then the methods were applied to various beam vibration problems and problems involving Euler and Beck's columns. Both methods yielded accurate solutions for all problems studied. The simple linear test function offers considerable savings in computing efforts as the domain integrals involved in the weak form are avoided. The two methods based on this simple linear test function method produced accurate results for frequencies and buckling loads. Of the two methods studied, the method with radial basis trial functions is very attractive as the method is simple, accurate, and robust.

Raju, Ivatury S.↗
NASA NTRS1998

Optimization of Turbine Blade Design for Reusable Launch Vehicles

To facilitate design optimization of turbine blade shape for reusable launching vehicles, appropriate techniques need to be developed to process and estimate the characteristics of the design variables and the response of the output with respect to the variations of the design variables. The purpose of this report is to offer insight into developing appropriate techniques for supporting such design and optimization needs. Neural network and polynomial-based techniques are applied to process aerodynamic data obtained from computational simulations for flows around a two-dimensional airfoil and a generic three- dimensional wing/blade. For the two-dimensional airfoil, a two-layered radial-basis network is designed and trained. The performances of two different design functions for radial-basis networks, one based on the accuracy requirement, whereas the other one based on the limit on the network size. While the number of neurons needed to satisfactorily reproduce the information depends on the size of the data, the neural network technique is shown to be more accurate for large data set (up to 765 simulations have been used) than the polynomial-based response surface method. For the three-dimensional wing/blade case, smaller aerodynamic data sets (between 9 to 25 simulations) are considered, and both the neural network and the polynomial-based response surface techniques improve their performance as the data size increases. It is found while the relative performance of two different network types, a radial-basis network and a back-propagation network, depends on the number of input data, the number of iterations required for radial-basis network is less than that for the back-propagation network.

Shyy, Wei↗
NASA NTRS1994

From neural-based object recognition toward microelectronic eyes

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

Sheu, Bing J.↗
NASA NTRS1992

Negative-energy states in the Dirac-Hartree-Fock problem - The effect of omission of two-electron integrals involving the small component

The effect of omission of two-electron integrals involving basis functions for the small component of the wavefunction on the eigenvalue spectrum in the Dirac-Hartree-Fock problem is studied. From an analysis of the Fock matrix it is shown that omission of these integrals moves the negative-energy states down, not up. Their complete omission does not give rise to intruder states. The appearance of intruder states occurs when only some of the core integrals are omitted, due to the nature of particular contraction schemes used for the core basis functions. Use of radially localized functions rather than atomic functions alleviates the intruder state problem.

Dyall, Kenneth G.↗
NASA NTRS1986

Eigenfunction methods in magnetospheric radial-diffusion theory

Complete sets of orthonormal basis functions constructed according to a generalization of the quantum-mechanical WKB approximation can be used to generate a nearly-diagonal matrix representation of the radial-transport operator for ring-current ions in the presence of radial diffusion and charge exchange. The resulting eigenfunctions (constructed by weighting the basis functions in proportion to the respective components of the eigenvectors of the matrix representation) and eigenvalues provide a spatial and temporal description of the evolving phase-space density during and following a magnetospheric disturbance (e.g., a magnetic storm). A linear superposition of the basis functions can also be used to eliminate any discrepancy between the steady-state solution of the transport equation and the appropriate WKB approximation of this steady-state solution.

Schulz, Michael↗