Research On The Fault Detection In Power Systems Based On Wavelet Theory

After exploring the basic theory of wavelet alternation, the paper goes on to study how to use it to fix fault time quickly and precisely among fault signals in power systems. In addition, the paper explores fault identification methods and analyses the harmonic in power systems. The main idea is as follows:1. From the angle of engineering application, the paper studies signals division and reconstruction, signals strangeness detection, data condensation and reconstruction, and signals noise reduction. It discusses further the fault features in power systems, bringing forward a data condensation method based on the best optional wavelet packet. As to signals noise reduction, it puts forward a threshold value estimation method so as to greatly increase the efficiency of data condensation and signals noise reduction.2. Starting from the circuit vibration equation, the paper explains the basic spread theory of traveling wave. It describes the traveling wave mathematical equation of single phase line without loss and three phase transposed line without loss. Also, according to the theory of the refraction and reflection of traveling wave, the paper analyses the transmission process of fault traveling wave with the help of the grid way, deducing the equation of traveling wave voltage and current.3. By comparing various methods of signal analyses, together with taking into account the signal features of grounding fault, the paper puts forward an idea of faulted feeder identification, as well as a new algorithm for faulted feeder identification of transient information, both based on the analyses of wavelet packet. As to the design of the new algorithm and the concrete problems in practice, such as choosing wavelet functions, coping with the alternative boundaries of wavelet, deciding the size of data windows, the paper gives qualitative and quantitative analyses, so as to provide a theoretical guide for better applying the new algorithm and solving practical problems.4. Starting from signal strangeness, the paper analyzes the basic theory -modulus maxima theory in signal strangeness detection and applies the theory of signal strangeness detection to the traveling wave test of temporary current in power systems. By detecting signal strangeness and reducing signal noise, the paper arrives at the maximum value of the detected signal, thus the fault location of the traveling wave signal is achieved.5. Considering that wavelet transform is linear orthogonal transform, the paper analyses the harmonic of active power filter with the help of self-adaptive wavelet algorithm, and applies the lifting wavelet algorithm to the high harmonics test of active power filter.The operation of the programs on the emulator and experimental results of real system has proved the correctness and validity of the analyses and syntheses methods in this paper, which indicates that the research has its practical application.