Pattern Recognition Technique Based Fault Detection in Multi-Microgrid

This paper provides a fault protection scheme for a multi-microgrid system. An Artificial Neural Network (ANN) is intended to identify the fault and its classification. The feed-forward network with a back propagation algorithm is considered for building and training of ANN. With the help of three-phase voltages and currents, the feature extraction is carried out through Discrete wavelet transform and load angle computation is done through Phase Locked Loop (PLL). This proposed method is given results accurate in examining fault occurrence in multi-microgrid systems. The proposed method results are also compared with the Fuzzy logic based fault system.