Power Quality Detection Based On Modulus Maximum Value Of Wavelet Transform

Research on the power quality detection is becoming one of the most important subjects in modern power system. It is very important to find out the influencing factors in order to enhance power quality. However, the electrical signals sampled from the real power system are often corrupted by noises which are harmful to effective handling of the useful information. Therefore, it is necessary to remove the noises before signal processing. When accident or fault occurs in the system, the sampled signals always include short impulses or sudden break components. These components often contain very useful message for detecting the important characteristics of the abnormal moment in time. The wavelet transform is introduced as a new signal-processing tool in power system signal detection. The wavelet transform has good characteristics in both time and frequency domains so that we may distinguish the useful signal from the noisy signal and locate the break point using the modulus maximum value of wavelet transform to realize on-line detection.This thesis focuses on power quality detection based on modulus maximum value of the wavelet transform. The main works are as follows:1. Chapter 1 presents the definition, causes, classification, and evaluation criterion of power quality problems. Three methods of power quality analysis based on transform are compared. An introduction to some software packages for analyzing electromechanical oscillation, including Matlab and PSCAD/ETDC, is made at the end of this chapter.2. Chapter 2 discusses the inherent drawbacks of Fourier analysis and the wavelet transform method is presented. The principles of the continuous wavelet transform, the discrete wavelet transform, the binary wavelet transform, multiresolution analysis and Mallat algorithm are discussed in detail. It is useful for signal denoising and the abnormal point detecting with modulus maximum value of wavelet transform.3. Chapter 3 discuses the main principle of signal denoising and the abnormal point detecting with the modulus maximum value of wavelet transform. Based on the different properties of signals and noises with wavelet transform, the modulus maximum value of wavelet transform is applied to eliminate the noise. This method is put forward to remove the modulus maximum that decreases with the increase of scale. The remainder is reconstructed to the originalsignal which is denoised. The modulus maximum is realized by obtaining modulus maximum according to Adhoc algorithm. At last, the voltage sags and voltage swell emulation examples validate the effectiveness of this method.4. Chapter 4 is a summary of the thesis.