A Novel Extrapolation-Based Grey Prediction Model for Forecasting China’s Total Electricity Consumption

Table 1

Simulation/prediction values and errors of NEGM(1, 1), SAIGM, and GM(1, 1).

Year	Actual value	NEGM(1, 1)		SAIGM		GM(1, 1)
		2000	1347.2	1347.2	—	1347.2	—	1347.2	—
2001	1463.4	1463.4	—	1348.2	7.8687	1785.5	22.0076
2002	1633.2	1573.9	3.6320	1642.6	0.5776	1955.0	19.7047
2003	1903.2	1893.7	0.5007	1942.5	2.0632	2140.7	12.4777
2004	2197.1	2216.4	0.8757	2247.7	2.3045	2344.0	6.6845
2005	2494.0	2542.1	1.9260	2558.5	2.5873	2466.6	2.9095
2006	2858.9	2870.9	0.4156	2874.9	0.5614	2810.3	1.7001
2007	3271.2	3202.8	2.0968	3197.1	2.2651	3077.2	5.9312
2008	3454.1	3537.8	2.4160	3525.1	2.0553	3369.4	2.4520
2009	3703.2	3875.9	4.6563	3859.0	4.2077	3689.4	0.3732
2010	4193.5	4217.3	0.5576	4199.0	0.1311	4039.7	3.6665
2011	4700.1	4561.8	2.9524	4545.1	3.2971	4423.4	5.8875
2012	4976.3	4909.5	1.3522	4897.5	1.5828	4843.5	2.6696
2013	5420.3	5259.6	2.9583	5256.3	3.0252	5303.4	2.1565
2014	5638.4	5614.9	0.4296	5621.6	0.2978	5807.1	2.9911
2015	5802.0	5972.6	2.9252	5993.5	3.3007	6358.5	9.5919
MRSPE			1.9780		2.4089		6.7469
In-sample (simulation)
2016	6192.7	6333.6	2.2596	6372.1	2.8977	6962.4	12.4292
2017	6482.1	6698.0	3.3149	6757.6	4.2507	7623.5	17.6085
MRPPE			2.8027		3.5742		15.0188
Out-of-sample (prediction)
CMRPE			2.3904		2.9916		10.8829