When a fault occurs, the current of the GIS isolating switch will change greatly compared with the normal operating state of the angle of rotation. The fault diagnosis method based on BP neural network is adopted. First, collect the current and rotation angle signals. Then extract relevant features for network training. Finally, the diagnosis model is used to diagnose the fault of the GIS isolating switch. Simulation and experiment show that this method can effectively improve the fault recognition rate.

BP neural network， GIS isolation switch， current， rotation angle， fault diagnosis