Intelligent Method Of Resonance Grounding System Fault Line Selection

When single-phase to earth faults occurs in distribution network, as the increasing hybrid transmission lines, the traditional fault detection methods based on the stable characters have not reached the actual request and the methods based on the transient characters have become a trend.For this situation, a simulation resonant earthed system according with actual situation is established in MATLAB/Simulink environment in this thesis. Based on this simulation system, a large number of researches have been done. In the same time, we put forward a fault line detection method of neural network for hybrid transmission lines of distribution network. In this neural network, We can make use of S-Transform to distill short window energy regarding sample. The main contents of the thesis are listed as follow.(1) Several questions including S-Transform,short window energy,neural network and simulation are deeply analyzed in this thesis.(2) A simulation resonant earthed system is established in MATLAB/Simulink environment. Aiming at different fault lines and situation, a lot of simulation works have been done for future sufficient data.(3) Single phase to ground fault line detection based on improved neural network is established and this method is verified its reliability through great number of situation works.