基于空洞卷积和增强型多尺度特征自适应融合的滚动轴承故障诊断

韩康,战洪飞,余军合,王瑞

Rolling bearing fault diagnosis based on dilated convolution and enhanced multi-scale feature adaptive fusion

Kang HAN,Hongfei ZHAN,Junhe YU,Rui WANG

表 6 变工况下8种模型的准确率

Tab.6 Accuracy of 8 models under variable working conditions

模 型 A A1 A2 A3 A4 MSCNN-LSTM[14] 0.7459 0.9592 0.9185 0.7799 WDCNN[15] 0.8119 0.8988 0.8602 0.8242 MSCNN[16] 0.8947 0.9280 0.9581 0.8508 DC 0.8771 0.9150 0.8535 0.8062 Resnet[17] 0.9338 0.9842 0.8970 0.9102 DRSN-CS[18] 0.9982 0.9928 0.9318 0.9027 MA1DCNN[19] 0.8953 0.9106 0.9459 0.7685 DC-MAFFM 0.9977 0.9999 0.9983 0.9447