A Hybrid Forecasting Model Based on Empirical Mode Decomposition and the Cuckoo Search Algorithm: A Case Study for Power Load

Table 1

Comparison of power load forecasting result by using different methods in Feb. 2.

Time	Actual value (MWh)	BPNN		GABPNN		WNN		CSAWNN		EMD-CSAWNN		ARIMA
		Forecasting	MAPE	Forecasting	MAPE	Forecasting	MAPE	Forecasting	MAPE	Forecasting	MAPE	Forecasting	MAPE
		Value	(%)	Value	(%)	Value	(%)	Value	(%)	Value	(%)	Value	(%)
0:00	8086.20	7267.78	10.12	7471.59	7.60	7828.93	3.18	7708.89	4.67	8033.14	0.66	8146.33	0.74
2:00	7132.07	7381.04	3.49	7193.94	0.87	7303.83	2.41	7150.74	0.26	7013.90	1.66	7298.17	2.33
4:00	7038.50	7023.50	0.21	7010.09	0.40	6470.25	8.07	6938.94	1.41	6938.97	1.41	6899.89	1.97
6:00	8803.29	8561.04	2.75	8611.39	2.18	8159.83	7.31	8612.24	2.17	8668.60	1.53	8619.45	2.09
8:00	10646.56	10778.68	1.24	10416.06	2.16	10890.92	2.30	10474.09	1.62	10707.41	0.57	10692.11	0.43
10:00	11822.68	11877.37	0.46	11811.56	0.09	11975.11	1.29	11880.45	0.49	11787.15	0.30	12002.49	1.52
12:00	12397.22	12329.59	0.55	12453.22	0.45	12328.77	0.55	12432.34	0.28	12326.00	0.57	12597.51	1.62
14:00	12824.82	12638.87	1.45	12918.86	0.73	12708.23	0.91	12929.78	0.82	12599.61	1.76	13091.61	2.08
16:00	13088.14	12810.33	2.12	13065.94	0.17	12903.89	1.41	13171.34	0.64	12830.13	1.97	12949.91	1.06
18:00	12001.48	11857.34	1.20	12090.61	0.74	12148.48	1.22	11500.36	4.18	11987.11	0.12	11776.37	1.88
20:00	10898.97	10827.46	0.66	10896.10	0.03	10967.20	0.63	10679.01	2.02	10778.06	1.11	10705.75	1.77
22:00	9398.37	9516.59	1.26	9665.37	2.84	9415.90	0.19	9706.03	3.27	9451.51	0.57	9446.46	0.51