Skip to main content


Springer Nature is making Coronavirus research free. View research | View latest news | Sign up for updates

Table 2 State-of-the-art: Data-driven modelling

From: Energy forecasting based on predictive data mining techniques in smart energy grids

Author Forecast model Forecast horizon Performance metrics & forecast error measurement Contribution & Perspective
Liu et al. (Liu et al. 2015) − Back propagation based ANN 24 h ahead MAPE = 7.65% − Aerosol index parameter used as input, which resulted in slightly improved accuracy
Ramsami, Oree (Ramsami and Oree 2015) − Stepwise Regression
24 h ahead RMSE = 2.74% − Stepwise regression: select I/P variables highly correlated to PV power O/P
− GRNN,FFNN,MLR and their hybrid were applied on the I/P
− Hybrid model showed slight improvement
Ordiano et al. (Ordiano et al. 2017) − ANN6
− ANN10 (Data-driven)
24 h ahead MAE = 6.64---7.25%
RMSE = 12.47---13.3%
\( r\left(\rho, \overline{\rho}\right) \)= 85.81---87.65%
− Accuracy of model strictly related to accuracy of historical database (Achieved reasonable accuracy)