Generator rotor magnetic flux and inter-turn short circuit fault identification method and system

Abstract

The application relates to the technical field of motor fault diagnosis, in particular to a generator rotor magnetic flux and inter-turn short circuit fault identification method and system, which adopts a key phase signal as a synchronous reference to collect multiple-channel original magnetic flux waveforms in parallel; the waveforms are subjected to working condition normalization processing; the waveform segments of each magnetic pole are intercepted based on the key phase signal, and a differential magnetic flux sequence is obtained by point-by-point difference with a healthy reference; three features are extracted to form a feature vector, a pre-trained fuzzy logic classifier is used to accurately distinguish inter-turn short circuit, rotor eccentricity and a healthy state, and the fault magnetic pole number or the eccentricity direction and degree are output; the recess depth and width are further calculated to evaluate the short circuit severity level, and the slot number range where the short circuit occurs is inversely deduced according to the waveform recess boundary. The application effectively suppresses working condition interference, realizes reliable differentiation of inter-turn short circuit and eccentricity faults, can accurately locate to a magnetic pole or even a slot area, and provides high-sensitivity early warning and decision basis for generator state maintenance.

Images