Motor fault diagnosis method based on graph attention network

Abstract

The invention discloses a motor fault diagnosis method based on a graph attention network. The method comprises the following steps of 1) dividing data samples; 2) respectively carrying out band-passfiltering on the two-phase current in each small sample to obtain denoised state characteristic electric signals; 3) constructing a graph network, namely constructing a current graph network accordingto the extreme points of the state characteristic electric signals, and fusing the n current graph networks of the two-phase current in the large sample to obtain the two-phase current graph networkof the large sample; and 4) constructing a classification model based on the graph attention network, namely respectively constructing the graph attention network based on the two-phase current graphnetwork, and fusing the features extracted by the two networks for classification. The electric signal data adopted by the invention is relatively convenient in data acquisition and low in cost, the method for converting signals into the graph network is provided, and after features are extracted through a graph attention network and a convolutional neural network, motor faults can be effectivelyclassified and diagnosed.

Description

technical field [0001] The invention relates to a motor fault diagnosis method based on a graph attention network, which belongs to the field of motor fault diagnosis. Background technique [0002] Motor bearings are one of the most critical components in an electric motor. Any bearing failure, even a small one, can lead to failure of the entire system. According to statistics, 40-70% of electromechanical drive system and motor failures are caused by rolling bearing damage. Therefore, the detection and diagnosis of motor faults are very important. Early detection of motor faults can prevent the system from shutting down due to accidental bearings, and ensure the continuous operation of the system while ensuring safety, thereby stabilizing work efficiency. [0003] At present, the motor vibration signal can convey the health status information of the motor during operation, so the mainstream motor fault diagnosis method is mainly based on this signal to analyze the motor co...

AI Technical Summary

Problems solved by technology

[0004] Traditional signal processing and feature extraction techniques, such as Fourier transform, wavelet decomposition, empirical mode decomposition, and Silbert transform, are not ideal when applied to motor fault diagnosis based on motor current signals

Images