Synchrophasors Covariance Index-based Fault Section Location for Active Distribution Networks

We intend to introduce the statistic inference into to the fault diagnosis in active distribution networks. In this respected, we propose a real-time fault detection and fault section identification by computing the comprehensive covariance index and confidence intervals of monitored buses. In the method, the historical data set is chosen from historical operation synchronization data that have representative operational states and thus avoid priori knowledge of networks. Generalized eigenvalue analysis is then performed to find the directions with maximum variance deviation in the monitored period versus the historical operation period. For the fault section location purpose, a comprehensive covariance index and confidence intervals are proposed to evaluate overall monitored nodes' state. The validity and practicability of the proposed method are verified by the simulation results of IEEE 34 modified system.