Research On Detection High Impedance Fault In Distribution Feeder

When the transmission Line occured high resistance earth fault,It will affect the normal operation of the distribution system.Usually,single phase grounding fault can cause both high phase current and zeros equence current. Proper conventional grounding protection set can act reliably and cut off fault. However,the current levels of high impedance fault occured in the distribution feeder are usually lower than the level of a direct short circuit current fault. Since the voltage drops down very little and the voltage among three phases keeps symmetric,this fault increases the difficulty of the check on this fault. Facts have proved that the detection of high impedance can not only use one fault signal to determine the fault,it should considere all the voltage and current changes.First,this paper puts forward a scheme of detection high impedance fault in the transmission line based on Artificial Neural Network. This method can solve the problem of detection high impedance fault.When high impedance fault has occurred, the voltage and the current of the system will have high-frequency components.we may use these characteristics of the failure to research the high impedance fault. on the other hand, high-frequency components also appear in other cases, such as opening and closing of capacitor, transformer tap switch.The article use the wavelet analysis to detect the fault signal,then obtained standard deviation from the series operation of the DWT analysis,in which the currents and voltages are measured in the distribution feeder,then put the standard deviationin into the artificial neural networks to identify failure to solve the HIF in the transmission line.Lots of MATLAB simulation results of HIF with different fault resistance and different kinds of normal switching events show that the two methods have perfect performance.Espatially the Wnn can detect HIF with the high reliability.