The discrimination of the fault type of the transmission line based on combined neural network

Accurate discrimination of the fault type of the transmission line is a key to the fault location.Based on the fault analysis of the high voltage transmission line,the type discriminated network model is created that combines Kohonen self organized feature mapping network and the BP network in order to implement the detection and discrimination of the fault type of the transmission.The theory analysis and the EMTP simulation indicate that this combined network needs less training stylebook and is learnt quicker than the single network.The combined network can achieve the discrimination of the fault type of the transmission line reliably and accurately under any kind of the fault mode,not influenced by the transition resistance,the starting angle of the fault,operation of the power system and the location of fault.