Research of Fault Phase Selection based on the Energy Coefficient of Fault Current Component and PNN

Fault-phase selection is an important link on corrective actions of high-voltage transmission lines protection, but also an important prerequisite for accurate fault location. The traditional fault phase selection method is only applicable to simple fault, low sensitivity, and easily affected by system operation mode, fault grounding resistance etc.. On the basis of the analysis for fault component of each phase current, this paper gives a definition of fault component energy by the concept of signal energy and presents a new type of fault phase identification method based on fault component energy coefficient. This method directly uses the fault current to extract fault components and calculate the signal energy of fault component for two cycles after fault inception, and the ratio coefficient that fault component energy of each phase current accounts the total energy can be gotten. Then probabilistic neural network of fault phase selection may be established to select fault phase by fault phase identification feature vector.