A Hybrid Forecasting Model Based on EMD-GASVM-RBFNN for Power Grid Investment Demand
Table 6
Comparison of forecasting results of different forecasting models from 2000 to 2017.
Year
AV
EMD Model
GA-SVM
BPNN
FV
APE/
FV
APE/
FV
APE/
2000
-0.50
-0.57
12.78
-0.64
25.70
-0.61
21.27
2001
-0.56
-0.47
20.14
-0.55
3.05
-0.56
0.23
2002
-0.46
-0.58
22.00
-0.55
15.92
-0.54
14.46
2003
-0.55
-0.57
5.24
-0.46
19.51
-0.39
34.78
2004
-0.54
-0.43
24.54
-0.52
5.65
-0.38
33.72
2005
-0.45
-0.44
2.42
-0.42
5.17
-0.36
16.55
2006
-0.24
-0.29
5.82
-0.28
4.72
-0.26
2.48
2007
-0.11
-0.10
0.64
-0.09
1.66
-0.10
0.79
2008
0.06
0.11
5.03
0.06
0.38
0.10
4.06
2009
0.43
0.32
7.79
0.14
20.36
0.15
19.31
2010
0.26
0.30
2.76
0.43
13.31
0.40
10.38
2011
0.35
0.32
2.25
0.26
7.01
0.26
6.82
2012
0.34
0.33
0.60
0.31
2.40
0.33
0.61
2013
0.42
0.42
0.39
0.40
1.07
0.44
1.55
2014
0.51
0.50
0.93
0.59
5.09
0.69
11.43
2015
0.71
0.66
2.52
0.83
6.99
0.73
1.27
2016
1.00
1.03
1.47
0.83
8.54
0.74
12.69
2017
0.96
0.90
3.12
1.12
8.33
1.07
5.79
MAPE/
6.69
8.60
11.01
RMSE
167.08
311.71
345.77
AV and FV represent the actual value and the forecasting value of the power grid investment, respectively. APE, MPAE, and RMSE represent the absolute percentage error, mean absolute percentage error, and root mean square error, respectively.