Research on the Grey Verhulst Model Based on Particle Swarm Optimization and Markov Chain to Predict the Settlement of High Fill Subgrade in Xiangli Expressway
Table 5
The appraise indices of different prognostic models to predict the settlement of high fill subgrade in Xiangli Expressway.
Year
Month
GM(1,1)
PSOMGM(1,1)
ARIMA
GVM
MGVM
PSOMGVM
APE (%)
APE (%)
APE (%)
APE (%)
APE (%)
APE (%)
2017
Jul.
22.90
/
/
/
/
/
/
Aug.
25.50
4.76
2.76
/
6.42
3.89
3.95
Sept.
21.50
2.57
0.52
2.77
3.50
0.90
0.96
Oct.
18.80
0.67
4.21
4.89
0.67
1.29
0.39
Nov.
16.00
6.55
1.38
7.12
7.72
0.64
1.33
Dec.
14.40
6.25
1.09
2.08
8.77
1.61
2.31
2018
Jan.
13.50
1.27
3.64
6.03
5.11
1.81
1.13
Feb.
15.50
0.05
2.15
7.16
5.40
1.54
0.86
Mar.
12.60
1.82
0.24
3.78
5.10
1.82
1.14
Apr.
14.30
10.79
3.73
2.07
2.86
0.24
0.30
MAPE (%)
3.86
2.19
4.49
5.06
1.52
1.37
RMSE
0.83
0.44
0.77
0.95
0.38
0.38
2018
May
12.40
7.04
0.31
19.50
2.17
0.43
0.41
Jun.
11.50
13.03
6.15
33.72
6.65
4.99
4.18
MAPE (%)
10.04
3.23
26.61
4.41
2.71
2.30
RMSE
1.23
0.50
3.23
0.57
0.41
0.34
Note: the optimal indices of the six competing models are in italic.