Disclosed are a method and a system for online monitoring and fault diagnosis of a synchronous phase modulator. A fault feature vector of a fixed-rotor inter-turn magnetic flux and a fault feature vector of a vibration signal under fault operation of a synchronous phase modulator, form a fault feature vector set. The fault feature vector set is used as data training input for a multi-fault classifier. The multi-fault classifier sets the data range of the vector set corresponding to different faults according to the basic data set and the fault feature vector set. The feature vector set of thesynchronous phase modulator under unknown fault conditions is input into the vector machine classifier, in order to obtain the corresponding fault category. The fault category of the synchronous phasemodulator is diagnosed through the fixed rotor current feature vector. If the diagnosis result is the same as the above, then this kind of fault occurs, and the specific fault mode of the synchronousphase modulator is comprehensively judged. The invention can realize the multi-parameter comprehensive diagnosis of the synchronous phase modulator, solves the problem of simplification of the faultdiagnosis parameter of the existing synchronous phase modulator, and improves the accuracy of fault diagnosis.