Rotor winding inter-turn short circuit fault diaginosis system based on artificial neural network

Abstract

The diagnosis system of generator rotor winding inter-turn short circuit fault based on artificial neural network (ANN) is developed. First, adapting for the need of faster speed and higher diagnosis precision of fault diagnosis, a new faster back- propagation (BP) with the error contracting gradually algorithm is selected, and the diagnosis model of rotor winding inter-turn short circuit fault based on ANN is proposed. Then the diagnosis system is developed by Visual Basic and SQL Server database, which uses the RS485 digit communication port of the exciting current meter and power meter of the generator to collect the data of exciting current, active power, inactive power, and uses the NPort to transfer them to the SQL Server database of the server. Finally, the diagnosis system is successfully applied to power plant. © 2007 IEEE.