Deep Learning Enhanced Solar Energy Forecasting with AI-Driven IoT

Table 2

Predicting MAPE results for one year of data.

		2017.10	2017.11	2017.12	2018.1	2018.2	2018.3	2018.4	2018.5	2018.6	2018.7	2018.8	2018.9	Overall
7.5 min	MLP	22.15	20.07	23.05	25.62	18.10	24.97	26.00	23.19	28.39	25.08	22.18	20.03	23.11
	LSTM	23.87	21.42	20.96	28.76	16.63	28.96	20.99	31.19	40.02	20.21	22.18	21.73	24.61
	ALSTM	21.64	22.44	21.44	28.16	23.52	24.55	18.43	17.82	32.79	25.46	29.05	19.78	23.32
	Prop.	19.95	16.76	20.79	25.01	16.90	19.29	15.84	16.83	17.21	22.20	19.88	18.69	18.82
15 min	MLP	30.19	24.41	26.42	28.58	20.63	37.38	47.40	40.80	53.70	38.39	31.30	27.10	34.17
	LSTM	23.42	21.47	23.74	27.51	17.74	23.62	22.63	21.21	28.95	28.72	19.00	18.64	22.75
	ALSTM	20.30	23.08	22.40	28.38	20.40	29.96	29.69	20.53	68.64	47.60	24.06	22.49	28.82
	Prop.	22.69	20.32	26.51	28.04	21.055	28.17	28.61	25.12	18.23	25.44	23.87	20.89	23.79
30 min	MLP	33.09	35.15	36.98	37.68	28.49	33.44	28.85	28.92	30.17	35.92	32.78	30.65	32.57
	LSTM	37.75	41.62	38.86	38.42	41.45	36.07	36.34	30.06	37.90	39.01	32.59	40.96	37.39
	ALSTM	25.67	29.86	37.76	42.77	33.89	30.41	57.01	28.64	30.09	30.71	37.50	35.79	34.18
	Prop.	32.62	29.59	29.32	32.88	23.50	33.38	34.87	31.34	45.62	30.19	31.84	32.62	32.39