Research Article
Prediction Method of TBM Tunneling Parameters Based on Bi-GRU-ATT Model
Table 2
Comparison of evaluation indexes of prediction results among different recurrent neural network models.
| Model | Parameter | | | | |
| Bi-GRU-ATT | Total thrust | 1.01 | 0.925 | 151.52 | 193.62 | Penetration | 1.60 | 0.930 | 0.15 | 0.20 | Cutterhead torque | 1.40 | 0.936 | 31.30 | 44.44 | Cutterhead power | 1.71 | 0.944 | 34.13 | 34.75 |
| Bi-GRU | Total thrust | 1.11 | 0.915 | 146.35 | 184.23 | Penetration | 2.05 | 0.898 | 0.20 | 0.26 | Cutterhead torque | 1.84 | 0.909 | 52.65 | 67.00 | Cutterhead power | 2.13 | 0.884 | 38.30 | 46.02 |
| GRU | Total thrust | 1.21 | 0.859 | 196.68 | 280.99 | Penetration | 2.47 | 0.867 | 0.23 | 0.27 | Cutterhead torque | 2.26 | 0.879 | 60.87 | 73.37 | Cutterhead power | 2.52 | 0.870 | 48.41 | 50.95 |
| LSTM | Total thrust | 1.44 | 0.870 | 205.60 | 252.02 | Penetration | 2.97 | 0.840 | 0.27 | 0.32 | Cutterhead torque | 2.04 | 0.889 | 55.88 | 67.97 | Cutterhead power | 2.7 | 0.860 | 44.61 | 52.89 |
|
|