Fault Detection Of Distribution Network Based On Wavelet Analysis And Genetic Algorithm

The quality of citizen’s life relies so much on the electricity, that the security and reliability of distribution network is important. Recently, the studies on the fault detection of single-phase ground fault of small current grounding system emerge endlessly. There are so many methods on fault detection, including zero sequence current mutation method, transfer function method and traveling wave method. Mathematical methods are popular recently, these methods includes wavelet transform, ant Colony Algorithm, Neural Network, Genetic Algorithm(GA) and so on.This paper proposes a method based on wavelet analysis and GA for fault location of distribution network. Wavelet analysis is used to deal with the problem of fault line selection, and the Genetic Algorithms is used to locate the fault. The single-phase ground fault’s of small current grounding system’s model was built by Simulink before line selection, and use the model to get the information of the current and voltage. Analyze the current and voltage by wavelet analysis to select the line. Wavelet singularity principle was used in line selection process, then the wavelet packet analyses the signal in features bands and determines the fault line. This paper selects the zero-sequence voltage as the object in the wavelet analysis. Compared with the zero-sequence current, the mutation of zero-sequence voltage is greater, and the disruption is smaller. So the experiment’s results are more accuracy in this way. GA is used after the line selection. GA used a binary encoding way to detect the fault location, rather than blindly searching all nodes. Genetic Algorithm combined topology for searching, the combination with the network’s topology makes the position more precisely. It’s not only for single power network also applies to dual power network. And the binary encoding way has a larger scale than decimal encoding system. This paper optimizes the GA process, it not only ensures the population’s diversity, but also retains the excellent individuals. This method avoids the local optimum and makes the search results accurately.The simulation experiments verified the reliability and fault tolerance of the algorithm. This paper proves the single-phase ground fault’s location method of small current grounding system is feasible and efficient. The method greatly reduces the operation range, reduces workload, and more realistic for applications.