Synchronous motor fault diagnosis method based on MSK-CNN and multi-source electromechanical information fusion
The invention provides a synchronous motor fault diagnosis method based on MSK-CNN and multi-source electromechanical information fusion, and the method comprises the steps: collecting the phase voltage, rotor vibration and stator vibration signals of a synchronous motor in a normal working conditio...
Saved in:
Main Authors | , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
05.08.2022
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention provides a synchronous motor fault diagnosis method based on MSK-CNN and multi-source electromechanical information fusion, and the method comprises the steps: collecting the phase voltage, rotor vibration and stator vibration signals of a synchronous motor in a normal working condition and a working condition of different rotor winding turn-to-turn short circuit in real time, and dividing the signals into a training set and a test set; an optimal MSK-CNN model is determined; respectively inputting original data of phase voltage, rotor vibration and stator vibration into an MSK-CNN model for training, fusing fault features extracted by a multi-scale kernel through a multi-scale feature extraction layer, and taking the fused fault features as output of each MSK-CNN sub-model; and features extracted based on the three kinds of signals are fused, so that the diagnosis purpose is achieved. According to the method, an end-to-end fault diagnosis scheme is provided for the synchronous generator, extra |
---|---|
Bibliography: | Application Number: CN202210393744 |