Electrical Fault Diagnosis Based on Feature Extraction and Support Vector Machine for Permanent Magnet Synchronous Generator

During the operation of the permanent magnet wind generator, electrical faults such as winding short circuit, winding open circuit, and winding asymmetry may occur, which directly affects the regular operation of the wind turbine and adversely affects wind power generation. This paper proposes an electrical fault diagnosis method for permanent magnet generators based on feature extraction and Support Vector Machine. By simulating electrical faults on a permanent magnet generator with a power of 25 kW, the 3-phase current and vibration signals of the generator are collected. Features are extracted from the current signal, and feature value classification is performed through support vector machine to implement pattern recognition of electrical faults and determine the operating status of the generator.