Fault synthetic recognition for an EHV transmission line using a group of neural networks with a time-space property

Fault diagnosis of an extra high voltage (EHV) transmission line is of great importance to the restoration decision system of power systems. At present, the research of neural networks (NNs) in this problem area still has some limitations. This paper details the development and building of an intelligent system of NN groups with a time-space property for EHV transmission line fault synthetic recognition and performance analysis. The structure of each NN model is divided according to the principle of dynamic time interval on the basis of analysing the interrelation and indefinite operating sequences of all apparatus in the case of faults occurring in the transmission line. Simulation results on the system show that this system can perform fault synthetic recognition exactly and has a forecast fault function