SEARCH · Search NASA
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.
In this paper filtering and smoothing criteria for maximum-likelihood parameter estimation are described and compared. It is shown that the parameter estimate using the smoothing criterion may be obtained independently of the smoothed state estimate. Monte Carlo test results are presented comparing the efficiency of the estimation procedures.
Optimum linear smoothing is utilized to estimate certain distortions in LANDSAT-D images. Measurements that are processed by the smoother consist of designated control of point locations within the images. Image distortions that are estimated by the smoother are those induced by LANDSAT-D satellite navigation errors and slowly varying attitude and sensor alignment uncertainties. Preliminary results indicate that optimum smoothing produces substantially more accurate distortion estimates than optimum filtering and that optimum smoothing may reduce the number of control points needed to yield desired image correction accuracy.
Two programs smooth and scale arbitrary airfoil coordinates. Airfoil smoothing program (AFSMO) utilizes both least-squares polynomial and leastsquares cubic-spline techniques to smooth iteratively second derivatives of y-axis airfoil coordinates with respect to transformed x-axis system that unwraps airfoil and stretches nose and trailing-edge regions.
Experimental strain measurements have been made at the highly strained regions on notched plate specimens that were made of Hastelloy X. Tests were performed at temperatures up to 1,600 F. Variable load patterns were chosen so as to produce plastic and creep strains. Were appropriate, notch root stresses were experimentally estimated by subjecting a smooth specimen to the measured notch root strains. The results of three analysis techniques are presented and compared to the experimental data. The most accurate results were obtained from an analysis procedure that used a smooth specimen and the Neuber relation to simulate the notch root stress-strain response. When a generalized constitutive relation was used with the Neuber relation, good results were also obtained, however, these results were not as accurate as those obtained when the smooth specimen was used directly. Finally, a general finite element program, ANYSIS, was used which resulted in acceptable solutions, but, these were the least accurate predictions.
Forward and inverse dynamics solved using Kalman filtering and Bryson-Frazier smoothing. Dynamics of serial-link robot arm solved by using recursive techniques from linear filtering and smoothing theory. Solutions of dynamical equations give forces, moments, and accelerations at joints between links, and multilink inertia matrix and its inverse. Theoretical developments lay foundation for use of filtering and smoothing techniques in design of robot controls.
This paper applies linear filtering and smoothing theory to solve recursively the inverse kinematics problem for serial multilink manipulators. This problem is to find a set of joint angles that achieve a prescribed tip position and/or orientation. A widely applicable numerical search solution is presented. The approach finds the minimum of a generalized distance between the desired and the actual manipulator tip position and/or orientation. Both a first-order steepest-descent gradient search and a second-order Newton-Raphson search are developed. The optimal relaxation factor required for the steepest descent method is computed recursively using an outward/inward procedure similar to those used typically for recursive inverse dynamics calculations. The second-order search requires evaluation of a gradient and an approximate Hessian. A Gauss-Markov approach is used to approximate the Hessian matrix in terms of products of first-order derivatives. This matrix is inverted recursively using a two-stage process of inward Kalman filtering followed by outward smoothing. This two-stage process is analogous to that recently developed by the author to solve by means of spatial filtering and smoothing the forward dynamics problem for serial manipulators.
Turbulent flow in tubes with and without longitudinal grooves is examined. The discovery of fine grooves forming a sort of streamline pattern on the body of sharks led to the expectation that the grooves on a surface reduce the momentum change, and thus the drag. To test this thesis, drag law, velocity profile and the profile of the velocity fluctuation were determined. Results show that for moderate Reynolds numbers the drag coefficient for grooved tubes is about 3 percent smaller than that of the smooth tubes. At higher Reynolds numbers, however, the drag coefficient for grooved tubes becomes larger than that for smooth tubes. No significant differences in the velocity profiles between grooved tubes and smooth tubes are found.
The napthalene sublimation technique was employed to study the detailed mass transfer distributions around the sharp 180 deg turns in a two-pass, square, smooth channel and in an identical channel with two rib-roughened opposite walls. Experiments conducted for Reynolds numbers of 15,000, 30,000, and 60,000 indicate that the Sherwood numbers on the top, outer, and inner walls around the turn in the rib-roughened channel are higher than the corresponding Sherwood numbers around the turn in the smooth channel. Sherwood numbers after the sharp turn are found to be higher than those before the turn for both the smooth and the ribbed channels.
Analyses of several carbonaceous chondrites (including the Orgueil samples 626, 629, and 630; and Ivuna samples 627, 628, and 631) for elements from Fe through Ru were performed using proton-induced X-rays, in order to test the smoothness of the elemental abundances in these chondrites. In general, the results confirm previously published abundance curves. The abundances were exceptionally smooth and strongly decreasing in the mass region 60-75. From the mass number 75 to 101, a smooth curve could be drawn, within limits of intersample variability, except for Y peak. Over the whole periodic table, a large number of peaks of probable nucleosynthetic origin could be identified.
A smooth three-dimensional mass distribution is approximated by a model with multiple thin screens, with surface mass density varying smoothly on each screen. It is found that 16 screens are sufficient for a good approximation of the three-dimensional distribution of matter. It is also found that in this multiscreen model the distribution of amplifications of single images is dominated by the convergence due to matter within the beam. The shear caused by matter outside the beam has no significant effect. This finding considerably simplifies the modeling of lensing by a smooth three-dimensional mass distribution by effectively reducing the problem to one dimension, as it is sufficient to know the mass distribution along a straight light ray.
The Planetary Radio Astronomy (PRA) experiment Warwick et al., (1977) on the spacecraft Voyager 2 observed three types of smooth emission: (1) numerous recurrent episodes are modeled by filled emission cones from midlatitude conjugate sources; (2) an 'equatorial' feature seen soon after closest approach includes electron cyclotron harmonic emission above the upper hybrid resonance, as well as smooth recurrent emission, its strange appearance is a result of rapid change in Voyager's magnetic latitude and; (3) unique 'high-latitude' emission is seen near closest approach during Voyager's single excursion to high north (+) magnetic latitude when fc, the electron cyclotron frequency at the spacecraft, lies in the observable range. The stronger component covers a broad band of frequencies above 2fc; its sensitivity to magnetic field identifies it as extraordinary (X) mode. The weaker component extends smoothly through f = fc and is identified as ordinary (O) mode. At each frequency f the observed sense of polarization reverses when f = 2fc.
The effects of deficient labyrinthine function on smooth visual tracking with the eyes and head were investigated, using ten patients with bilateral peripheral vestibular disease and ten normal controls. Active, combined eye-head tracking (EHT) was significantly better in patients than smooth pursuit with the eyes alone, whereas normal subjects pursued equally well in both cases. Compensatory eye movements during active head rotation in darkness were always less in patients than in normal subjects. These data were used to examine current hypotheses that postulate central cancellation of the vestibulo-ocular reflex (VOR) during EHT. A model that proposes summation of an integral smooth pursuit command and VOR/compensatory eye movements is consistent with the findings. Observation of passive EHT (visual fixation of a head-fixed target during en bloc rotation) appears to indicate that in this mode parametric gain changes contribute to modulation of the VOR.
In improved method for generating trajectory of robotic manipulator, each straight-line segment of trajectory composed of constant-velocity main portion sandwiched between smooth acceleration at start and smooth deceleration at finish. Algorithm implementing method computes velocity in each accelerating portion as sinusoidal function of position along line. This motion chosen for two reasons: closely approximates motion of human hand along straight-line trajectory, and provides very smooth transitions between constant-velocity portion and accelerated and decelerational end portions.
A parameter identification algorithm for nonlinear systems is presented. It is based on smoothing test data with successively improved sets of model parameters. The smoothing, which is iterative, provides all of the information needed to compute the gradients of the smoothing performance measure with respect to the parameters. The parameters are updated using a quasi-Newton procedure, until convergence is achieved. The advantage of this algorithm over standard maximum likelihood identification algorithms is the computational savings in calculating the gradient. This algorithm was used for flight-test data consistency checks based on a nonlinear model of aircraft kinematics. Measurement biases and scale factors were identified. The advantages of the presented algorithm and model are discussed.
Studying the topology of large-scale structure as a way to better understand initial conditions has become more widespread in recent years. Studying topology of simulations (which have periodic boundary conditions) in redshift space produces results compatible with the real topological characteristics of the simulation. Thus we expect we can extract useful information from redshift surveys. However, with nonperiodic boundary conditions, the use of smoothing must result in the loss of information at survey boundaries. In this paper, we test different methods of smoothing samples with nonperiodic boundary conditions to see which most efficiently preserves the topological features of the real distribution. We find that a smoothing method which (unlike most previous published analysis) sums only over cells inside the survey volume produces the best results among the schemes tested.
We present a new method of three-dimensional smoothed particle hydrodynamics (SPH) designed to model systems dominated by deformation along a preferential axis. These systems cause severe problems for SPH codes using spherical kernels, which are best suited for modeling systems which retain rough spherical symmetry. Our method allows the smoothing length in the direction of the deformation to evolve independently of the smoothing length in the perpendicular plane, resulting in a kernel with a spheroidal shape. As a result the spatial resolution in the direction of deformation is significantly improved. As a test case we present the one-dimensional homologous collapse of a zero-temperature, uniform-density cloud, which serves to demonstrate the advantages of spheroidal kernels. We also present new results on the problem of the tidal disruption of a star by a massive black hole.
The project is based upon research of Tessler et al. (1994) on an improved variational formulation for post-processing stress predictions in Finite Element Analysis. The methodology, called Smoothing Element Analysis (SEA), employs a three-node smoothing finite element. The present effort focused on verifying the basic constant strain criterion for the three-node smoothing element subject to a set of internal penalty constraints. The convergence characteristics of the element are assessed by first deriving the constrained form of the assumed element stress and stress gradient fields, and then by verifying the validity of the constant strain criterion once the element penalty constraints are explicitly imposed. The analytical investigation is carried out with the use of the symbolic manipulation code 'Mathematica'.
During the rise from sunspot minimum to maximum, the observed value of smoothed monthly mean sunspot number at maximum RM is found to correlate with increasing strength against the current value of smoothed monthly mean sunspot number R(t), where t is the elapsed time in months from minimum. On the basis of the modern era sunspot cycles (i.e., cycles 10-22), the inferred linear correlation is found to be statistically important (i.e., at the 95-percent level of confidence) from about 11 mo past minimum and statistically very important (i.e.. at the 99-percent level of confidence) from about 15 mo past minimum; ignoring cycle 19, the largest cycle of the modern era, the inferred linear correlation is found to be statistically important from cycle onset. On the basis of R(t), estimates of RM can be gauged usually to within about +/- 30 percent during the first 2 yr and to within about +/- 20 percent (or better) after the first 2 yr of a cycle's onset. For cycle 23, because controversy exists regarding the placement of its minimum (i.e., its onset), being either May 1996 or perhaps August 1996 (or shortly thereafter), estimates of its RM are divergent, being lower (more like a mean size cycle) when using the earlier epoch of minimum and higher (above average in size) when using the later-occurring minimum. For smoothed monthly mean sunspot number through October 1997 (t = 17 or 14 mo, respectively), having a provisional value of 32.0. the earlier minimum date projects an RM of 110.3 +/- 33.1, while the later minimum date projects one of 137.2 +/- 41.2. The projection is slowly decreasing in size using the earlier onset date, while it is slowly increasing in size using the later onset date.