Intelligent Fault Detection and Location Scheme for Low Voltage Microgrids based on Recurrent and Radial Basis Function Neural Networks

Fault location is vital in control, operation and protection of microgrids. In the microgrids due to the existence of inverter-based distributed generations (IBDGs), the fault current is approximately limited to load nominal current and traditional protection schemes are not enable to fault detection in the islanding operation mode of a microgrid. Also in low voltage microgrids, due to the existence of a large number of laterals, fault detection and location are protection challenges. In this paper, an intelligent scheme proposed based on recurrent neural networks (RNN) and radial basis function (RBF). The proposed scheme extracts fault type, phase, and location information. In this scheme, two laterals in the microgrid are selected to evaluate the results of the protection scheme. three-phase fault and sequence components of branch current used to features that these available data are as inputs into neural networks to develope fault information. To show the proposed scheme efficiency, we performed a comprehensive analysis on modified CERTS microgrid with the existence of laterals in DIgSILENT Power Factory and MATLAB softwares.