Research On Transmission Line Fault Location Based On Extended Prony Algorithm And Two-End Unsynchronized Measurements

ABSTRACT: High-voltage overhead transmission lines are essential to the power system. When transmission lines short-circuit faults occur, the timely and accurate fault location and quickly repairing fault lines are considered as effective measures to ensure the entire power system stability and economic operation. Comparing a variety of high-voltage transmission line fault location algorithms, considering the disadvantage of present algorithms for fault location, this thesis presents a method extracting the fault-phasors from transient sampled data and calculates the fault distance by fault location algorithm based on two-end unsynchronized measurements.This thesis analyzes the factors causing error of algorithms based on phasors; discuss the features of various ranging models. And based on three-phase full transposition and the model of even distributed parameter transmission lines, this thesis derives the equations of fault location using two-end unsynchronized measurements.After faults, transient processes will occur in power system. The relay protection equipment responses quickly and segregates the fault lines from the system, which make fault time shorter and the number of sampled data smaller. The transient sampled data include phasors at 50Hz, attenuating DC, and harmonics. So, how to get accurate phasors used to locate fault affects greatly the accuracy of fault location algorithms. To cope with the problem above, this thesis gives a method using extended Prony to eliminate the attenuating DC components and applying DFT to obtain fault phasors. In this way, the error of fault location was greatly minimized.This thesis selects PSCAD/EMTDC as simulation environment to build models of actual high-voltage transmission lines on various fault conditions. A large number of different fault simulations are conducted. The approach using extended Prony algorithm and Fourier algorithm to extract the phasors components of voltage and current, incorporating fault location algorithm based on two-end unsynchronized measurement, locates the fault. The comparison between the method proposed in this paper and the modified DFT shows that the method proposed achieves smaller errors.But the simulation shows a trend that the location error may be bigger when the faults occur close to the measurement terminals. Since that, to make the error over the whole lines even, ANNs are applied to be trained which makes the error more even and smaller. This thesis uses extended Prony and DFT to extract the phasors and takes advantage of fault location algorithm based on two-end unsynchronized measurements. Finally, ANNs are adopted to cope with location error. Simulations results show that on most fault conditions, this method can locate the fault distance accurately; this method is feasible and valuable.