Ehv Transmission Line Transient Signal Detection, Classification And Identification Methods

EHV (Extra High Voltage) transmission line is an important part in Chinese transmission grid.In order to guarantee the the transmission capacity and power quality,,it is must to improve the reliability of transmission.To detect and recognize the transmission fault or disturbance signal are very necessary,thus, it ensures the stability of the grid and avoids great loss. To detect timely and recognize disturbance signal accurately provide criterion for protect action.Along with power transmission's expanding and more complex, the transient signal detection and recognization are also becoming more and more important.The paper introduces the resent research of transient signal detection and recognization,and compares the various methods.In addition, it introduces the method of time-frequency analysis, and puts forward S transform in signal analysis compared with other methods of special advantage, especially in the transient signal detection and recognization.500kV transmission line model is Established in EMTDC simulation environment, and get single phase to grobnd, faulty lightning strike,lightning interference and capacitor swithing four high transient signal in transmission lines.To detect the transient signal under different frequencies of the amplitude based on S transform and get the transient signal amplitude of different frequency curves, finally, analyzes the results. Using the S transform transient energy to recognizes the four kinds of transient signal to fault state or the fault state. Then, using S transform transient energy ratio to recognize the four transient signal, and one of the fault distance, time, transition resistance and lightning current amplitude of transient energy ratio are analyzed. Puts forward the criterion to recognize the four kinds of transient signal in different conditions, the criterion has a high accuracy.LIBSVM is constructed using SVM classifier experimental packages. Take the transient energy ratio data on all sorts of transient signal different conditions as SVM training and testing samples, and standardization and not standardized treated. Finally, in two kinds of different situations, the result shows the SVM classification has different effects on the final results of the classification. In addition, using SVM classifier.experimental of four kinds of transient signal classification has a higher accuracy.