ANN based Faulty Phase Identification and Fault Resistance Estimation in Inverter Interfaced Grid Connected Microgrids

Fault detection is a challenging task in inverter interfaced microgrids as fault current in inverter interfaced distributed generator (IIDG) is limited. The different control strategies adopted by IIDGs make the fault detection more complex. For quick restoration of normal supply, faulty phase identification is essential. The root cause of the fault can be assessed from the estimation of fault resistance. This paper utilizes artificial neural network (ANN) to identify the faulty phase in feeders and sub-feeders in the microgrid acquiring current signals from the substation. Further, fault resistance has been estimated utilizing ANN. Power spectral density (PSD) of the three phase current signals at the substation calculated utilizing discrete wavelet transform (DWT) has been fed to ANN model to classify faults and estimate fault resistance. Only half cycle of the current signal after fault is required and the simulation results establishes the efficacy of the proposed technique.