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

<div>The comparison between the traditional CNN fault diagnosis method and the fine-grained fault diagnosis algorithm. (a, b) The framework of the traditional CNN methods. “fc” denotes the fully connected layer. “C1”∼“C4” denote 1∼4 gear position, respectively. (c) The fine-grained fault diagnosis algorithm with two-level hierarchy label structure.</div>

Shock and Vibration

fig1

Figure 1

Figure 1: Fault Diagnosis of Gearbox in Multiple Conditions Based on Fine-Grained Classification CNN Algorithm 