【摘要】 为提高轴承故障诊断模型的跨工况检测能力,提出了一种基于信号奇偶分离(OE)、卷积块注意力模块(CBAM)、卷积神经网络(CNN)和双向门控单元(BiGRU)相结合的OE-ACNNBiGRU故障诊断方法,以优化诊断性能。首先,对输入样本信号进行奇偶分离,分别进行卷积运算;其次,用CBAM注意力模块对奇偶信号,分别施加空间注意力和通道注意机制;再次,把经过处理的奇偶信号重新进行特征融合;最后,用双向门控循环单元对融合信号提取时序特征,通过全连接层和Softmax后输出检测分类结果。实验表明:检测精度达到了99.66%以上,跨工况效果相对其他消融模型,平均检测精度提高了3个点以上达到94.36%。更多还原

【Abstract】 To improve the cross-working condition detection capability of the bearing fault diagnosis model,an OE-ACNN-BiGRU fault diagnosis method is proposed to optimize diagnosis performance by combining signal parity separation(OE),convolutional block attention module(CBAM),convolutional neural network(CNN)and bidirectional gating unit(BiGRU).First,the input sample signals are separated into odd and even ones,and convolution operations are performed respectively.Second,the CBAM attention module is used to apply spatial attention and channel attention mechanisms to the odd and even signals respectively.In addition,the processed odd and even signals are re-fused with features.Finally,a bidirectional gated recurrent unit is used to extract temporal features from the fused signal,and the detection and classification results are output after passing through the fully connected layer and softmax.The experiment results show that the detection accuracy reaches more than 99.66%.Compared with other ablation models,the average detection accuracy of the cross-working condition increases by more than 3 points to 94.36%.更多还原