Fault Analysis In Power System Based On Wavelet Transfer

Fault diagnosis technology is a developing researched field, it's the interactive result of requirement of actuality application and developing of multi-disciplinary theory. From the actuality application, with improving of modernization technology and the greatly increased complexity of all kinds of industrial systems, the reliability and security of these systems has been a key factor to insure economic vantage and social vantage; From the developing of disciplinary theory, fault diagnosis is an interactive disciplinary. These twenty year's developing of modern control theory, signal processing, pattern recognition, artificial intelligence etc has provided powerfully basic theory. Substation is very important in power transmission and transformation system. Substation automation has some functions such as supervisory control and protection for main equipment and transmission or transformation line, as well as communication with control center. So it should be find and remove the fault in time in the power system in order to guarantee the smooth manufacturing and life.Wavelet analysis is a kind of method to process non-stationary signal in these years. Having good time-frequency localization character, and correctly identifying singularity point in fault signal, It's the main analysis method in this paper. This paper is rightly based on this good character, applying it to fault analysis of power system. After studied the basic theory of wavelet transfer and its applying in fault analysis, this paper mainly illustrated samples of apllying wavelet transfer to power system. In this paper, a de-noise example was list to explain how to remove noise from signal, and Lipschitz exponent method was applied to identify the singularity in power system fault signal analysis and a simulation example was given in the last chapter. At last, we studied the application combining the wavelet and BP neural networks to fault detection and simulated system fault diagnose by computer and proved the advantages of using wavelet method in fault detection.