Research On The Power Quality Analysis Based On Wavelet Transform

In late years, The issues of power quality have captured considerable attention from both power systems and their customers. One of reasons is many precision instruments and devices based on computers are applied, which are sensitive for power quality. The other reason is the nonlinear equipments, including power electronic apparatus, are applied, which have worsened the power quality day by day. For improving power quality, the monitoring and analyzing systems must be established to detect, estimate and classify different disturbances in power system.Firstly, this paper introduces the definition about power quality, classifies power quality phenomena, and explains the cause of power quality. Secondly, three methods, including Fourier transform, short-time Fourier transform and wavelet transform are discussed. This section explains the theories and how to use the theories in power quality analysis, and also indicates their advantages and disadvantages.Because how to detect and classify the non-steady signal is a focus of analysis of power quality, a method is designed, which is based on wavelet transform and Fourier transform. In this method, the non-steady state disturbance is detected by wavelet transform module maximum. If the disturbance is happen, several steps are executed. The time is recorded. The signal is divided by the time of disturbances, and the style of short-duration variations is distinguished by calculating magnitude value. The signal of the disturbance is extracted by reconstructing the signal in high frequency sub-bands by means of the Mallat reconstruction algorithm. When no disturbance is happen, signal is treated as steady signal.This paper also introduces how to analyze the voltage flicker based on wavelet fransform. Compared to conventional methods, this method can detect what time the voltage flicker happen and over. The fluctuating signal can be divided into sub-bands by wavelet transform. The frequency, magnitude of the fluctuating signal can be attained. This method is suitable when the fluctuating signal includes several kind of frequency.