Fault Diagnosis of Gearbox in Multiple Conditions Based on Fine-Grained Classification CNN Algorithm

Table 5

Comparison of fault diagnosis accuracies between traditional CNN models and fine-grained CNN model.

Working condition	Traditional CNN models		Fine-grained CNN model
	Accuracy (%)	Avg (%)	Accuracy (%)	Avg (%)
Single working condition (CNN-S)
Condition 1	95.5	93.6	—	—
Condition 2	88.5		—
Condition 3	94.5		—
Condition 4	95.8		—
Multiple working conditions
Conditions 1–4	79.9 (CNN-M)	98.8 (CNN-FG)