Fault Diagnosis of Hydropower Unit Based on VMD-WT-CNN and Attention Mechanism
The fault diagnosis of hydropower unit depends on the vibration monitoring signal,but the noise in the sig-nal will interfere with the extraction of effective features and reduce the accuracy of the model.In this paper,a fault diag-nosis method of hydropower unit based on variational mode decomposition and wavelet threshold denoising is proposed.Firstly,the vibration monitoring signal of hydropower unit is decomposed by variational mode to obtain the low,medium and high frequency components.Secondly,wavelet transform is carried out on the high-frequency component and the part whose wavelet coefficient is lower than the set threshold is discarded,and the middle and low frequency component is re-tained.Finally,a multi-channel deep convolutional neural network model based on attention mechanism is established,and the above components are taken as input signals of each channel to realize state recognition of hydropower units.The experimental results show that this method can effectively filter out the noise in the vibration monitoring signals of hydro-power units and improve the recognition accuracy of the diagnostic model.
万方数据
维普
