Page 73 - 《爆炸与冲击》2026年第11期

−20% error bound
−40% error bound
−40% error bound

第 46 卷 王帅帅，等： 数据驱动的箍筋加密钢筋混凝土柱爆炸损伤预测与评估模型 −20% error bound 第 11 期

1.0 1.0
+20% error bound
+20% error bound
2
2
Training set, R =0.999 Training set, R =0.997
2
Testing set, R =0.986 +40% error bound Testing set, R =0.984 +40% error bound
2
0.8 0.8
Prediction value 0.6 −20% error bound Prediction value 0.6 −20% error bound
−40% error bound
−40% error bound
0.4
0.4
0.2 0.2
0 0
0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0
Actual value Actual value
(c) XGBoost (d) LightGBM
1.0 1.0
+20% error bound
+20% error bound
Training set, R =0.984 Training set, R =0.997
2
2
2
2
Testing set, R =0.958 +40% error bound Testing set, R =0.989 +40% error bound
0.8 0.8
Prediction value 0.6 −20% error bound Prediction value 0.6 −20% error bound
−40% error bound
−40% error bound
0.4
0.4
0.2 0.2
0 0
0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0
Actual value Actual value
(e) MLP (f) TabPFN
图 11 6 种算法的预测值与真实值的比较
Fig. 11 Comparison of predicted values and actual values for six algorithms

表 7 每个机器学习模型的性能指标
Table 7 Performance metrics for each machine learning model
模型 数据集 R 2 e MAE e RMSE e MedAE
Training set 0.998 0.007 0.014 0.005
DT
Testing set 0.970 0.029 0.050 0.015
Training set 0.996 0.010 0.017 0.006
RF
Testing set 0.981 0.026 0.040 0.018
Training set 0.999 0.006 0.008 0.004
XGBoost
Testing set 0.986 0.024 0.034 0.018
Training set 0.997 0.011 0.016 0.008
LightGBM
Testing set 0.984 0.024 0.036 0.018
Training set 0.984 0.026 0.036 0.019
MLP
Testing set 0.958 0.039 0.059 0.024
Training set 0.997 0.008 0.015 0.005
TabPFN
Testing set 0.989 0.018 0.030 0.011
注：字体加粗表示模型在该方面表现最佳。

2
e MAE =0.006，e RMSE =0.008，e MedAE =0.004 和测试集的 R =0.986，e MAE =0.024，e RMSE =0.034，e MedAE =0.018；基于
2
2
TabPFN 训练集的 R =0.997，e MAE =0.008，e RMSE =0.015，e MedAE =0.005 和测试集的 R =0.989，e MAE =0.018，

111504-14