A novel fault classification method based on neural network in smart grid

This paper presents a back propagation (BP) neural network method to identify fault types and phases in the smart grid with renewable sources. A fault classification method based on the BP neural network is used to identify patterns of voltage and current measured with phasor measurement unit (PMU). The voltage and current values are decomposed, and then they are used as the input matrix of the BP neural network, which identifies the fault type and fault phasor. All the common short-circuit faults occurred are analyzed, BP neural network method is proposed for fault classification in a power active distribution network. Simulation results show feasibility and usefulness of the BP neural network method, which provides good classification performances.