Hybrid Time-Series Prediction Method Based on Entropy Fusion Feature

Table 6

Improved comparative experiment with tuning parameters.

Index	Model	Power consumption (kW h)	Temperature (°C)	Wind speed (M/s)	Humidity (%)
RMSE	EEMD-ELSTM	501.79	0.51	0.0040	1.08
	LSTM	1140.19	0.72	0.0072	2.78
	ARIMA	1217.73	0.72	0.0054	4.98
	ELSTM	1108.35	0.68	0.0060	2.52
	LSTM-TCN	680.20	0.43	0.0052	2.08
MAE	EEMD-ELSTM	399.93	0.24	0.0028	0.87
	LSTM	758.65	0.45	0.0042	2.05
	ARIMA	928.45	0.49	0.0041	2.62
	ELSTM	727.27	0.43	0.0036	1.86
	LSTM-TCN	513.91	0.34	0.0032	1.51
MAPE (%)	EEMD-ELSTM	1.27	1.61	3.58	1.16
	LSTM	2.41	4.57	5.88	2.99
	ARIMA	2.42	4.92	5.04	3.89
	ELSTM	2.39	4.48	5.75	2.70
	LSTM-TCN	1.71	3.02	4.62	2.55