Fault Identification Of High-voltage Transmission Lines Based On SVM

High-voltage transmission network is the hub of the transmitting end and electricity consumption end, which plays a very important connective role. Since the high-voltage transmission lines running environment is more severe, all kinds of malfunctions occur frequently. Due to the single-phase instantaneous earth fault constitute a large proportion, the fault identification method for single-phase instantaneous earth fault of high-voltage transmission lines is proposed.The article firstly deals with the fault recorder data processed by the fault recorder. Aiming at time synchronization problem, it puts forward to the corresponding solutions. From the point of view of wavelet transform, the signal energy is divided in the scale domain. By theoretical analysis and real fault data, it concludes that the bior3.5 is the better choice for compression and reconstruction. At the same time, the characteristics of the signal feature extraction are introduced, such as energy index, variance index and entropy index. This characteristics are substituted into the support vector machine classification model, you can enable predictive classifier training and prediction.Secondly, a detailed analysis of different causes of power line fault is given. The causes are divided into six categories, namely fire, ice, lightning, discharge of trees, foreign body, external damage. On the basis of the existing fault recording data, SVM classification method is used to identify the fault characteristics. The semblance of nuclear matrix and nuclear calibration are proposed to select the kernel function. And the genetic algorithm is used to estimate the penalty factor C and the nuclear parameters. Simulation results validate that the final predictive results of SVM classification model are more accurate, the model has some practical value.At last, the paper introduces the structure frame and related technology of the intelligent support platform of relay protection. The method based on fault recording data is put forward to evaluate the function of relay protection. The application of advanced functions like fault identification of high-voltage transmission, are realized in the actual platform.