A New Fault Locating Method for Power Network Based on An Improved Dynamic State Estimation

The increasing diversity of access factors and the growing complexity of the network structure bring new difficulties to the problem fault locating in distribution network. Using regional collection of electrical information, a fault locating method based on an improved dynamic network state estimation is proposed in this paper. At first, a fault network description model is established, in which fault events are depicted as changes in network topology and represented by Kirchhoff's equation. Then, a standardized fault network representation for different fault points is proposed to achieve a unified representation of different fault conditions of components. Further, the distribution network state is estimated with the help of different fault networks representing different component faults. After calculating the matching degree among the state estimation results of the different fault networks, the fault locating function of the distribution network is realized by comparing the matching degree of the state estimation results. Finally, a case simulation considering different fault types and fault scenarios such as transition resistance verifies the effectiveness of the proposed method, and further tests the fault locating performance in the case of data anomalies. The results show that the proposed method also has a good ability of resisting bad data.