A Novel Transmission System Fault Identification Method Based on K-Nearest Neighbour Algorithm

With the large-scale grid integration of wind turbines, the “double high” characteristics of the system emerge, and the existing power system fault identification methods are difficult to meet the needs of safe and stable operation of modern power grids. A power system fault identification method based on K-nearest neighbour (KNN) algorithm and stand-alone energy function is proposed to address this problem. Firstly, the stand-alone energy function based on short-circuit current is derived and its feasibility as a criterion for fault type identification is analysed. Secondly, considering the variation characteristics of wind turbine output energy, the partition boundary voltage values are used to initially screen the fault areas, and then a fault identification strategy for power systems containing new energy units is proposed by combining the KNN algorithm. Finally, experimental validation is carried out on a ten-unit system containing wind farms using Power System Analysis Software Package(PSASP) software. The simulation results show that the method proposed in this paper has strong applicability and stability, and can maintain high fault type identification accuracy under different working conditions.