Short-Term Reactive Power Forecasting Based on Real Power Demand Using Holt-Winters’ Model Ensemble by Global Flower Pollination Algorithm for Microgrid

Table 6

ARIMA model error analysis for the order (2, 1, 2).

Parameter	Coefficient	Std. error			0.025	0.975
Constant	0.2193	1.150	0.191	0.849	-2.035	2.473
Ar.L1.D value	-0.830	0.323	-2.569	0.012	-1.463	-0.197
Ar.L2.D value	0.0338	0.116	0.290	0.772	-0.194	0.262
ma.L1.D value	-0.1873	0.308	-0.609	0.544	-0.790	0.416
ma.L2.D value	-0.8125	0.307	-2.647	0.010	-1.414	-0.211