关键词: 开关磁阻电机调速系统, 功率变换器, 传感器, k-近邻算法, 极限学习机

Abstract: Aiming at the single-tube short-circuit fault and sensor noise fault of power converter in switched reluctance motor drive (SRD), an adaptive sliding window fault diagnosis method combining k-nearest neighbor (kNN) algorithm and extreme learning machine (ELM) was proposed. The fault was analyzed and the three-phase stator current were collected as the original data. The fast fourier transform and ReliefF algorithm were used for feature extraction and selection. A multi-window adaptive fault diagnosis structure combining kNN algorithm and ELM algorithm was set up. Through offline and online simulation experiments, it is proved that the method has high diagnostic speed and high precision.

Key words: switched reluctance motor drive (SRD),, power converter, sensor, , k-nearest neighbor (kNN),, extreme learning machine (ELM)