Table 1 Superparameter settings.
Algorithm | Hyperparameter type | Value |
|---|---|---|
BP | Hidden layer nodes | 9 |
Learning rate | 0.001 | |
Network target loss value | 0.0004 | |
Number of learnings | 2000 | |
PSO | Group size | 20 |
Inertia weights \(\omega_{start}\) and \(\omega_{end}\) | 0.9 ;0.4 | |
Learning factor \(c_{1}\) and \(c_{2}\) | 1.49;1.49 | |
Speed extremes | 1.3; − 1.3 | |
Total number of iterations | 100 | |
Optimize the target loss value | 0.0001 |
