Adaptive Fisher-Based Deep Convolutional Neural Network and Its Application to Recognition of Rolling Element Bearing Fault Patterns and Sizes

<div>The output layer residuals corresponding to the evolution algebra and diagnostic classification of PSO-AFDCNN (diagnosis accuracy 100%).</div>

Mathematical Problems in Engineering

fig7

Figure 7

Figure 7: Adaptive Fisher-Based Deep Convolutional Neural Network and Its Application to Recognition of Rolling Element Bearing Fault Patterns and Sizes 

Figure 7 | Adaptive Fisher-Based Deep Convolutional Neural Network and Its Application to Recognition of Rolling Element Bearing Fault Patterns and Sizes 