Network Construction for Bearing Fault Diagnosis Based on Double Attention Mechanism

Table 5

Comparison of accuracy rates of different models under different loads.

Method	0 hp (%)	1 hp (%)	2 hp (%)	3 hp (%)	Average accuracy (%)
Attention convolutional	74.27	60.34	87.27	77.71	74.9
Attention LSTM	82.5	87.42	84.28	84.72	84.73
1D-CNN-LSTM	96.08	98.57	98.73	97.68	97.69
Double attention lacks EMD	99	99.24	98.94	99.48	99.17
SSA-SVM	69.97	70.58	91.16	79.68	77.85
PSO-RF	80.36	79.63	93.16	93.16	86.58
AVSSA-KELM	89.18	88.01	82.49	91.46	87.79
Paper method	99.64	99.78	99.46	99.68	99.64