SEARCH · Engineering Papers

Results for “Maximum Likelihood”

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 937 records · Page 52

All sources NASA NTRS DOE OSTI

NASA NTRS1993

Constructing An Image From Nonuniform-Scan Data

Algorithm performs maximum-likelihood construction of image of scene scanned nonuniformly by linear array of photodetectors, responses of which unequal. Developed to postprocess data acquired by scanning telescope aboard Infrared Astronomical Satellite (IRAS), also adaptable to generation of high-resolution imagery from data acquired in other nonuniform-scanning applications. Achievable resolution not limited by finite size of detectors, and increases with density of scans. With sufficiently dense scans, one obtains resolution beyond diffraction limit of imaging optics.

Aumann, Hartmut H.↗

NASA NTRS1993

An analytic modeling and system identification study of rotor/fuselage dynamics at hover

A combination of analytic modeling and system identification methods have been used to develop an improved dynamic model describing the response of articulated rotor helicopters to control inputs. A high-order linearized model of coupled rotor/body dynamics including flap and lag degrees of freedom and inflow dynamics with literal coefficients is compared to flight test data from single rotor helicopters in the near hover trim condition. The identification problem was formulated using the maximum likelihood function in the time domain. The dynamic model with literal coefficients was used to generate the model states, and the model was parametrized in terms of physical constants of the aircraft rather than the stability derivatives resulting in a significant reduction in the number of quantities to be identified. The likelihood function was optimized using the genetic algorithm approach. This method proved highly effective in producing an estimated model from flight test data which included coupled fuselage/rotor dynamics. Using this approach it has been shown that blade flexibility is a significant contributing factor to the discrepancies between theory and experiment shown in previous studies. Addition of flexible modes, properly incorporating the constraint due to the lag dampers, results in excellent agreement between flight test and theory, especially in the high frequency range.

Hong, Steven W.↗

NASA NTRS1992

ERS-1 scatterometer calibration and validation activities at ECMWF. B: From radar backscatter characteristics to wind vector solutions

Calibration and validation activities for the ERS-1 scatterometer were carried out at ECMWF (European Center for Medium range Weather Forecast) complementary to the 'Haltenbanken' field campaign off the coast of Norway. At a Numerical Weather Prediction (NWP) center a wealth of verifying data is available both in time and space. This data is used to redefine the wind retrieval procedure given the instrumental characteristics. It was found that a maximum likelihood estimation procedure to obtain the coefficients of a reformulated sigma deg to wind relationship should use radar measurements in logarithmic rather than physical space, and use winds as the wind components rather than wind speed and direction. Doing this, a much more accurate transfer function than the one currently operated by ESA was derived. Sigma deg measurement space shows no signature of a separation in an upwind solution cone and a downwind solution cone. As such signature was anticipated in ESA's wind direction ambiguity removal algorithm, reconsideration of the procedure is necessary. Despite the fact that revisions have to be made in the process of wind retrieval; a grid potential is shown for scatterometry in meteorology and climatology.

Stoffelen, AD↗

NASA NTRS1993

The probabilistic neural network architecture for high speed classification of remotely sensed imagery

In this paper we discuss a neural network architecture (the Probabilistic Neural Net or the PNN) that, to the best of our knowledge, has not previously been applied to remotely sensed data. The PNN is a supervised non-parametric classification algorithm as opposed to the Gaussian maximum likelihood classifier (GMLC). The PNN works by fitting a Gaussian kernel to each training point. The width of the Gaussian is controlled by a tuning parameter called the window width. If very small widths are used, the method is equivalent to the nearest neighbor method. For large windows, the PNN behaves like the GMLC. The basic implementation of the PNN requires no training time at all. In this respect it is far better than the commonly used backpropagation neural network which can be shown to take O(N6) time for training where N is the dimensionality of the input vector. In addition the PNN can be implemented in a feed forward mode in hardware. The disadvantage of the PNN is that it requires all the training data to be stored. Some solutions to this problem are discussed in the paper. Finally, we discuss the accuracy of the PNN with respect to the GMLC and the backpropagation neural network (BPNN). The PNN is shown to be better than GMLC and not as good as the BPNN with regards to classification accuracy.

Chettri, Samir R.↗

NASA NTRS1993

Lunar gravitational field estimation and the effects of mismodeling upon lunar satellite orbit prediction

Lunar spherical harmonic gravity coefficients are estimated from simulated observations of a near-circular low altitude polar orbiter disturbed by lunar mascons. Lunar gravity sensing missions using earth-based nearside observations with and without satellite-based far-side observations are simulated and least squares maximum likelihood estimates are developed for spherical harmonic expansion fit models. Simulations and parameter estimations are performed by a modified version of the Smithsonian Astrophysical Observatory's Planetary Ephemeris Program. Two different lunar spacecraft mission phases are simulated to evaluate the estimated fit models. Results for predicting state covariances one orbit ahead are presented along with the state errors resulting from the mismodeled gravity field. The position errors from planning a lunar landing maneuver with a mismodeled gravity field are also presented. These simulations clearly demonstrate the need to include observations of satellite motion over the far side in estimating the lunar gravity field. The simulations also illustrate that the eighth degree and order expansions used in the simulated fits were unable to adequately model lunar mascons.

Davis, John H.↗

NASA NTRS1993

Combinatorial pulse position modulation for power-efficient free-space laser communications

A new modulation technique called combinatorial pulse position modulation (CPPM) is presented as a power-efficient alternative to quaternary pulse position modulation (QPPM) for direct-detection, free-space laser communications. The special case of 16C4PPM is compared to QPPM in terms of data throughput and bit error rate (BER) performance for similar laser power and pulse duty cycle requirements. The increased throughput from CPPM enables the use of forward error corrective (FEC) encoding for a net decrease in the amount of laser power required for a given data throughput compared to uncoded QPPM. A specific, practical case of coded CPPM is shown to reduce the amount of power required to transmit and receive a given data sequence by at least 4.7 dB. Hardware techniques for maximum likelihood detection and symbol timing recovery are presented.

Budinger, James M.↗

NASA NTRS1991

Distributed parameter estimation for NASA Mini-Mast truss through displacement measurements

Most methods of system identification of large flexible structures by far are based on the lumped parameter approach. Because of the considerable computational burden due to the large number of unknown parameters, distributed parameter approach, which greatly decreases the number of unknowns, has being investigated. In this paper a distributed parameter model for the estimation of modal characteristics of NASA Mini-Mast truss has been formulated. Both Bernoulli-Euler beam and Timoshenko beam equations are used to characterize the lateral bending vibrations of the truss. The measurement of the lateral displacement at the tip of the truss is provided to the maximum likelihood estimator. Closed-form solutions of the partial differential equations and closed-form expressions of the sensitivity functions are derived so that the estimation algorithm is highly efficient. The resulting estimates from test data by using Timoshenko beam model are found to be comparable to those derived from finite element analysis.

Huang, Jen-Kuang↗

NASA NTRS1991

High-speed laser anemometry based on spectrally resolved Rayleigh scattering

Laser anemometry in unseeded flows based on the measurement of the spectrum of Rayleigh scattered laser light is reviewed. The use of molecular scattering avoids the well-known problems (particle lag, biasing effects, seed generation, seed injection) of seeded flows. The fundamental limits on velocity measurement accuracy are determined using maximum likelihood methods. Measurement of the Rayleigh spectrum with scanning Fabry-Perot interferometers is analyzed and accuracy limits are established for both single pass and multi-pass configurations. Multi-pass configurations have much higher selectivity and are needed for measurements where there is a large amount of excess noise caused by stray laser light. It is shown that Rayleigh scattering is particularly useful for supersonic and hypersonic flows. The results of the analysis are compared with measurements obtained with a Rayleigh scattering diagnostic developed for study of the exhaust plume of a small hydrogen-oxygen rocket, where the velocities are in the range 1000 to 5000 m/sec.

Seasholtz, R. G.↗

NASA NTRS1991

Signal processing considerations for low signal to noise ratio laser Doppler and phase Doppler signals

The relative performance of current methods used for estimating the phase and the frequency in LDV and phase Doppler applications in low signal to noise ratio conditions is analyzed. These methods include the Fourier analysis and the correlation techniques. Three methods that use the correlation function for frequency and phase estimations are evaluated in terms of accuracy and speed of processing. These methods include: (1) the frequency estimation using zero crossings counting of the auto-correlation function, (2) the Blackman-Tukey method, and (3) the AutoRegressive method (AR). The relative performance of these methods is evaluated and compared with the Fourier analysis method which provides the optimum performance in terms of the Maximum Likelihood (ML) criteria.

Ibrahim, K. M.↗

NASA NTRS1992

X-ray emission at low-mass end of the MS - Results from an extensive Einstein Observatory survey

The 0.16-4.0 keV X-ray emission from K and M stars of luminosity classes IV, V, and VI within 25 parsec of the sun was measured using all available Einstein Observatory IPC data and a critical compilation of catalogued optical data. Fifty four of 88 stars were detected, 70 of 138 M stars with Mv less than 13.4 (corresponding to M6) and 15 or 31 fainter M stars. The surveyed stars were grouped, on the basis of U, V, W space velocity components, into old-disk, young-disk, and halo population stars. Then, a subsample was selected which is statistically representative of the population of K and M stars in the solar neighborhood, on the basis of which unbiased Maximum Likelihood X-ray luminosity functions were constructed for K, early M, and late M stars. The investigation revealed a decrease of X-ray luminosity with increasing stellar age in the range of ages of disk population stars.

Barbera, M.↗

NASA NTRS1993

Optimal thresholds for the estimation of area rain-rate moments by the threshold method

Optimization of the threshold method, achieved by determination of the threshold that maximizes the correlation between an area-average rain-rate moment and the area coverage of rain rates exceeding the threshold, is demonstrated empirically and theoretically. Empirical results for a sequence of GATE radar snapshots show optimal thresholds of 5 and 27 mm/h for the first and second moments, respectively. Theoretical optimization of the threshold method by the maximum-likelihood approach of Kedem and Pavlopoulos (1991) predicts optimal thresholds near 5 and 26 mm/h for lognormally distributed rain rates with GATE-like parameters. The agreement between theory and observations suggests that the optimal threshold can be understood as arising due to sampling variations, from snapshot to snapshot, of a parent rain-rate distribution. Optimal thresholds for gamma and inverse Gaussian distributions are also derived and compared.

Short, David A.↗

NASA NTRS1993

Timoshenko beam modeling for parameter estimation of NASA Mini-Mast truss

In this paper a distributed parameter model for the estimation of modal characteristics of NASA Mini-Mast truss is proposed. A closed-form solution of the Timoshenko beam equation, for a uniform cantilevered beam with two concentrated masses, is derived so that the procedure and the computational effort for the estimation of modal characteristics are improved. A maximum likelihood estimator for the Timoshenko beam model is also developed. The resulting estimates from test data by using Timoshenko beam model are found to be comparable to those derived from other approaches.

Shen, Ji Y.↗

NASA NTRS1993

Nonlinear smoothing identification algorithm with application to data consistency checks

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.

Idan, M.↗

NASA NTRS1991

Neuro-classification of multi-type Landsat Thematic Mapper data

Neural networks have been successful in image classification and have shown potential for classifying remotely sensed data. This paper presents classifications of multitype Landsat Thematic Mapper (TM) data using neural networks. The Landsat TM Image for March 23, 1987 with accompanying ground observation data for a study area In Miami County, Indiana, U.S.A. was utilized to assess recognition of crop residues. Principal components and spectral ratio transformations were performed on the TM data. In addition, a layer of the geographic information system (GIS) for the study site was incorporated to generate GIS-enhanced TM data. This paper discusses (1) the performance of neuro-classification on each type of data, (2) how neural networks recognized each type of data as a new image and (3) comparisons of the results for each type of data obtained using neural networks, maximum likelihood, and minimum distance classifiers.

Zhuang, Xin↗

NASA NTRS1992

On the use of stochastic process-based methods for the analysis of hyperspectral data

Further development in remote sensing technology requires refinement of information system design aspects, i.e., the ability to specify precisely the data to collect and the means to extract increasing amounts of information from the increasingly rich and complex data stream created. One of the principal directions of advance is that data from much larger numbers of spectral bands can be collected, but with significantly increased signal-to-noise ratio. The theory of stochastic or random processes may be applied to the modeling of second-order variations. A multispectral data set with a large number of spectral bands is analyzed using standard pattern recognition techniques. The data were classified using first a single spectral feature, then two, and continuing on with greater and greater numbers of features. Three different classification schemes are used: a standard maximum likelihood Gaussian scheme; the same approach with the mean values of all classes adjusted to be the same; and the use of a minimum distance to means scheme such that mean differences are used.

Landgrebe, David A.↗

NASA NTRS1992

Absolute classification with unsupervised clustering

An absolute classification algorithm is proposed in which the class definition through training samples or otherwise is required only for a particular class of interest. The absolute classification is considered as a problem of unsupervised clustering when one cluster is known initially. The definitions and statistics of the other classes are automatically developed through the weighted unsupervised clustering procedure, which is developed to keep the cluster corresponding to the class of interest from losing its identity as the class of interest. Once all the classes are developed, a conventional relative classifier such as the maximum-likelihood classifier is used in the classification.

Jeon, Byeungwoo↗

NASA NTRS1993

Cosmic background radiation anisotropy at degree angular scales - Further results from the South Pole

We report further results from the University of California at Santa Barbara program to measure anisotropy in the cosmic background radiation at angular scales near 1 deg, an angular range corresponding to the largest scales where structure is observed. A 30 GHz high electron mobility transistor amplifier-based detector was coupled to the Advanced Cosmic Microwave Explorer, a 1 m off-axis Gregorian telescope. We present data that represent 64 of the total of 500 hr acquired with this system during the 1990-1991 season. The data have a statistical error of 13.5/micro-K/pixel. These are the smallest error bars of any data set of this type published to date. The data contain a significant signal with a maximum likelihood Delta T/T roughly 1 x 10 exp -5. The spectrum of the signal seen in slightly less than 2 sigma away from the thermal spectrum expected of primordial fluctuations in the cosmic background radiation. If the source of the fluctuations is primordial, then the data are consistent with cold dark matter scenarios when normalized to the large-scale anisotropy observed by COBE, while if the origin of the signal is foreground emission or another form of contaminant then the data are marginally inconsistent with standard cold dark matter models. In either case, the data are sufficiently sensitive to provide a crucial test of many models.

Schuster, Jeffrey↗

NASA NTRS1993

Identification of integrated airframe-propulsion effects on an F-15 aircraft for application to drag minimization

The application of an adaptive real-time measurement-based performance optimization technique is being explored for a future flight research program. The key technical challenge of the approach is parameter identification, which uses a perturbation-search technique to identify changes in performance caused by forced oscillations of the controls. The controls on the NASA F-15 highly integrated digital electronic control (HIDEC) aircraft were perturbed using inlet cowl rotation steps at various subsonic and supersonic flight conditions to determine the effect on aircraft performance. The feasibility of the perturbation-search technique for identifying integrated airframe-propulsion system performance effects was successfully shown through flight experiments and postflight data analysis. Aircraft response and control data were analyzed postflight to identify gradients and to determine the minimum drag point. Changes in longitudinal acceleration as small as 0.004 g were measured, and absolute resolution was estimated to be 0.002 g or approximately 50 lbf of drag. Two techniques for identifying performance gradients were compared: a least-squares estimation algorithm and a modified maximum likelihood estimator algorithm. A complementary filter algorithm was used with the least squares estimator.

Schkolnik, Gerald S.↗