Intelligent Signal Processing Using Wavelet And Neural Network

Lee, Ian Wen Chun (2003) Intelligent Signal Processing Using Wavelet And Neural Network. Masters thesis, Multimedia University.

Full text not available from this repository.
Official URL: http://vlib.mmu.edu.my/diglib/login/dlusr/login.ph...

Abstract

This thesis presents a novel approach for the detection, classification and data mining of non-stationary signals in power networks by combining the S-Transform and neural networks. The S-Transform provides frequency dependent resolution that simultaneuously localizes the real and imaginary spectra. The S-Transform is similar to the Wavelet Transform but with a phase correction. This property is used to obtain useful features of the non-stationary signals that make the pattern recognition much simpler in comparison to the wavelet multiresolution analysis. Two neural network configurations are trained with features from the S-Transform for recognizing the waveform class. The classification accuracy for a variety of power network disturbance signals for both types of neural networks is shown and is found to be a significant improvement over multiresolution wavelet analysis with multiple neural networks. After the patterns are classified a knowledge discovery approach is used to extract further information of the non-stationary time series of the power signal data. This was done by integrating an expert system and Fuzzy MLP to generate important rules that reflect the relationship between the feature vectors and the signals. The efficacy of the rules was tested on a fuzzy inference system with good results. The proposed procedure is able to quantify relevant parameters of the signals with very high accuracy. The entire procedure completes the data mining of non-stationary time series data of power network signals. A quantitative comparison is done between the proposed method and learning wavelet based methods. Neural networks are used to test the classification accuracy of the proposed method and the leading wavelet based methods. The classification results of the proposed method are better than the wavelet based methods in several aspects.

Item Type: Thesis (Masters)
Subjects: L Education > LB Theory and practice of education > LB2300 Higher Education
Divisions: Faculty of Engineering (FOE)
Depositing User: Mr Shaharom Nizam Mohamed
Date Deposited: 03 Dec 2009 07:40
Last Modified: 04 Dec 2009 11:00
URI: http://shdl.mmu.edu.my/id/eprint/60

Actions (login required)

View Item View Item