Research On Fault Location Analysis Of Transmission Line Based On PMU

The characteristics of China’s electric system are wide transmission range,long distance,large transmission energy,high voltage level and complex transmission network.It’s very important and necessary to find out the reason and location of failure and repair it timely and accurately.The reason is that if the electric system fails lightly,it may cause inconvenience to our life.However,if it fails heavily,the electrical equipment will be destoryed or people’s life will be in danger.Based on GPS clock timing,synchronous phasor measurement unit(PMU)can realize realtime measurement and extraction of voltage,current and phase angle.PMU is also widely used in state estimation,fault diagnosis and equipment state assessment of transmission lines.Therefore,according to the data characteristics of PMU and the electrical characteristics of transmission line after failure,this thesis carries out specific research on fault line selection,fault location and fault classification.Firstly,in the aspect of fault line selection,in view of the difficult problem of determining fault lines under the condition of single-phase grounding fault,combined with the defects of many methods,the high frequency transient energy in zero-sequence current is extracted as the main feature.The main feature is extracted by variational mode decomposition(VMD),and the average current amplitude difference is used as the auxiliary feature to define the fault reliability.By PSCAD the results of the simulation test of the fault line with the software,it can be seen that the accuracy of the proposed line selection method is effectively verified in all kinds of fault situations.Secondly,in the aspect of fault location,in order to solve the problem of large fault location error of transmission lines,on the basis of PMU data,this thesis takes the algorithm of VMDHilbert by Dual-terminal positioning to extract the traveling waves,and at the same time,the frequency mutation point is taken as the time difference.After simulation and analysis,the traveling wave head can be effectively extracted by VMD-Hilbert traveling wave detection method.And it is not necessary to determine the process of traveling wave velocity according to the time of receiving the reflected traveling wave head and failure,the accuracy of fault location is greatly improved.Finally,in the aspect of fault classification,in order to distinguish the fault types of short circuit effectively,combined with the PSO-RF algorithm to classify the fault lines,three kinds of feature quantities are constructed to identify ten fault types,which consist of correlation coefficient,energy coefficient and zero sequence current.By PSO-RF algorithm to identify,train,test,classify and optimize the fault types,the simulation results show that the classifier established in this thesis can significantly improve the classification accuracy.