Provided is a bearing fault diagnosis method based on instantaneous frequency optimization VMD. The method comprises: S1, acquiring a bearing test signal; S2, decomposing the bearing test signal on the basis of an instantaneous frequency optimization VMD algorithm, and by allowing candidate K values to be traversed within a preset range, calculating a mean value of instantaneous frequencies under different candidate K values; S3, determining an optimal K value by using a Lagrange polynomial on the basis of the mean value of the instantaneous frequencies under different candidate K values; and S4, inputting a signal component corresponding to the optimal K value into a fault diagnosis model to obtain a bearing fault diagnosis result. By means of the method, a fault feature of a vibration signal can be better reflected and applied to bearing fault diagnosis, thereby effectively improving the classification precision of a fault diagnosis model.
展开▼
