Fault Location Identification in Meshed AC Microgrids Using Interharmonic Injection

This paper presents a support vector machine (SVM)-based fault detection algorithm for islanded meshed microgrids (MGs). Considering the low level of fault current in islanded MGs, the fault detection is carried out based on the magnitude of the voltage rather than the current. Accordingly, the rms voltages of the DERs are assigned as the feature vector for fault detection SVM. Upon the occurrence of a fault, the fault location is identified by means of an interharmonic injection scheme. To identify the faulty zone, the DER which is electrically closest to the fault injects interharmonic currents into the grid and the zone with highest current is considered as the faulty zone. Then, a second DER located in the same zone inject interharmonic currents at different frequencies. The resulting interharmonic voltages are measured by each DER to form a feature vector. A multi class SVM uses this feature vector to identify the faulty line within that zone. The proposed algorithm is tested on a nine bus meshed MG. Test results show that the proposed method is capable of accurate detection of fault location even with high fault impedances.