Fault location for distribution networks based on pattern recognition technique

Proposed a method to identify a faulted section in d distribution network by voltage sags pattern recognition. The method establish analytical voltage sags database using fault analysis. When a real short circuit fault occurs, the voltage at the monitored node is compared with the analytical voltage sags in the database to find all the possible fault sections. Then, the method applied rank reasoning analysis to prioritize all the possible fault sections, and check the possible fault section according to the rank. Through simulating for an 11kV typical test distribution network, it proves the proposed method´s validity and accuracy.