Dissertation > Excellent graduate degree dissertation topics show
Studies of Numerically Stable Estimation for MultiChannel Systems with Multiplicative Noises
Author: ChenMeng
Tutor: ZuoDongSheng
School: Ocean University of China
Course: Signal and Information Processing
Keywords: Multichannel with multiplicative noise Singular value decomposition (SVD) Linear minimum variance Optimal Estimation
CLC: TB53
Type: Master's thesis
Year: 2004
Downloads: 112
Quote: 13
Read: Download Dissertation
Abstract
State with multiplicative noise system optimal estimation theory has important applications in oil seismic exploration, underwater target detection, speech processing, and many other fields. Made in this area in recent years, a series of new theoretical and applied research results, breaking the multiplicative noise limit of the onedimensional random sequence, the development of the multichannel optimal estimation algorithm with multiplicative noise under, and in turn, the development of a series of research results to a multisensor information fusion technology and twodimensional with multiplicative noise system optimal estimation. The meaning of these algorithms are linear minimum variance is optimal. During the practical application of these algorithms may appear numerical instability in the recursive calculation of long numerical instability will affect the calculation accuracy of the algorithm divergence serious cause the algorithm fails completely. Therefore, these algorithms Numerical stability method is necessary. In this paper, the more universal significance of multichannel state with multiplicative noise system optimal estimation theory, linear minimum variance sense optimal while numerical stability as: First, this article recalled with multiplicative noise systems development and current status of optimal estimation theory, and numerical stability research in this field and describes the development of the status quo. Second, in a multichannel observation models with multiplicative noise system, multiplicative noise is no longer the traditional onedimensional random sequence, but in the form of random matrices, the first is a random diagonal matrix, and then extended to the general random matrix. First observation channel multiplicative noise random diagonal matrix systems, the use of the singular value decomposition (SVD) as a tool for the given channel characteristics not related to the numerical stability of the optimal state filtering algorithm, and the application of this method to multiplicative noise to a general random matrix system, given the same complex numerical stability of the multichannel with multiplicative noise system algorithm. Both filtering algorithms are kept linear minimum variance sense optimality. With multiplicative noise system for multichannel, in the optimal state filtering algorithm based on first derive a the Optimal Fixed case of a diagonal matrix directly smoothing algorithm multiplicative noise is random, and then using the singular value decomposition ( SVD) method to decompose the error covariance matrix of the smoothing algorithm, given its numerical stability of the algorithm, and thus the use of filtering and smoothing optimal estimation results, given the fixed domain deconvolution algorithm. Turn the above algorithm is then extended to the multiplicative noise is generally random array of circumstances, given the complex case of multichannel direct fixed domain smoothing algorithm, deconvolution algorithms and numerical stability of the algorithm. Fourth, simulation examples verify the effectiveness of each algorithm.

Related Dissertations
 The Research of OnOrbit Calibration Algorithm of Star Sensor,V448.2
 Interference Mitigation and Performance Enhancement for Cellular Systems,TN929.5
 A Study of Hybrid Estimation Theory & Applications Based on Particle Filtering,TN911.23
 Pole Feature Extraction and EPulse Target Discrimination for Ultra WideBand Radar,TN958
 Research on Several Robust Digital Watermarking Algorithms,TP309
 Fusion Kalman filter and Wiener filter based on the multisensor observations of modern time series analysis method,TP202.4
 Multi  sensor information fusion filters based on Kalman filtering method,TP202.4
 Multichannel systems with multiplicative noise measurement noise estimates and state fusion estimation algorithm,TN911
 The Study on the Feature of the Random Field of Soil Parameter and the Best Estimation,TU43
 Estimate and Control for Stochastic System with Time Delays,TP13
 Optimal Estimation and Control of Orientation of a Heliostat Tower,TM615
 Research on the Application of Image Processing in Face Recognition,TP391.41
 Study of Numerically Stable Estimation for Singular Systems with Multiplicative Noise,TN911.7
 The Research and Simulation of PVT Information Processing Technology of GPS Receiving System,P228.4
 The New Technology of Wavelet Estimation and Surface Wave Diminution in the Seismic Exploration,P631.41
 Linear model of minimum variance estimation problem,O212.1
 SAR image based on the optimal estimation method to improve the resolution,TN957.52
 Further Studies of Estimation Methods for Systems with Multiplicative Noises,TN911
 The ～*Congruence Class of the Solutions of Matrix Equations and Generalized Inverses of a Normal Matrix,O151.21
 Research on Identity Validation System Based on Face Recognition,TP391.41
CLC: > Industrial Technology > General industrial technology > Acoustic engineering > Vibration, noise and its control
© 2012 www.DissertationTopic.Net Mobile
