基于VMD-CNN的水电机组故障诊断

为提高水电机组故障诊断精度,减少在振动信号特征选取过程中对专业经验的依赖,提出了一种融合变分模态分解和卷积神经网络的故障诊断方法。首先对水电机组振动信号进行变分模态分解得到若干分量,并利用这些分量构造时间图,然后搭建深度卷积神经网络对时间图进行特征提取和故障识别,建立分量和故障状态的映射关系。以实测水电机组轴向振动信号进行应用检验,并采用多组对比试验,结果表明该方法与其他方法相比故障识别准确率更高。研究成果为水电机组智能故障诊断提供了新思路。

Fault Diagnosis of Hydropower Units Based on VMD-CNN

In order to improve the fault diagnosis accuracy of hydropower units and reduce the dependence on professional experience in the process of vibration signal feature selection, a fault diagnosis method based on the fusion of variational mode decomposition and convolutional neural network is proposed. Firstly, the vibration signal of hydropower unit is decomposed into several components by means of variational mode decomposition, and the time map is constructed by these components. Then, the deep convolution neural network is built to extract the features of the time map and identify the fault, and the mapping relationship between the components and the fault state is established. The application test is carried out with the measured axial vibration signal of the hydropower unit, and the results show that the method is more accurate than other methods. The research results provide new ideas for intelligent fault diagnosis of hydropower units.