Table 3 Statistical errors of the best-proposed forecasting models using \(SSO\), \(PSO\), \(CSO\) and NNA
From: Short-term photovoltaic power production forecasting based on novel hybrid data-driven models
| Â | Â | \(SV{R}_{RB}\) | \(SV{R}_{Linear}\) | \(BPN{N}^{1}\) | \(BPN{N}^{2}\) |
|---|---|---|---|---|---|
SSO | \(RMSE \; (\text{kW})\) | 4.4751 | 10.6940 | 4.8460 | 5.0742 |
\(nRMSE \; (\%)\) | 4.3374 | 10.3649 | 4.6969 | 4.9181 | |
\(MAE \; (\text{kW})\) | 2.5699 | 4.8621 | 3.1713 | 3.2365 | |
\(nMAE \; (\%)\) | 2.4909 | 4.7125 | 3.0737 | 3.1369 | |
PSO | \(RMSE \; (\text{kW})\) | 4.487 | 9.1334 | 4.9735 | 4.5564 |
\(nRMSE \; (\%)\) | 4.349 | 8.8524 | 4.8205 | 4.4223 | |
\(MAE \; (\text{kW})\) | 2.5728 | 4.8617 | 3.2410 | 2.8739 | |
\(nMAE \; (\%)\) | 2.4936 | 4.7121 | 3.1413 | 2.7854 | |
CSO | \(RMSE \; (\text{kW})\) | 4.4795 | 10.6932 | 5.0788 | 4.5692 |
\(nRMSE \; (\%)\) | 4.3417 | 10.3642 | 4.9226 | 4.4286 | |
\(MAE \; (\text{kW})\) | 2.5661 | 4.8637 | 3.3053 | 2.9614 | |
\(nMAE \; (\%)\) | 2.4871 | 4.7141 | 3.2036 | 2.8703 | |
NNA | \(RMSE \; (\text{kW})\) | 4.4889 | 10.6732 | 5.0281 | 4.6608 |
\(nRMSE \; (\%)\) | 4.3418 | 10.4944 | 4.8734 | 4.5175 | |
\(MAE \; (\text{kW})\) | 2.5752 | 4.8647 | 3.2662 | 2.9839 | |
\(nMAE \; (\%)\) | 2.4963 | 4.7311 | 3.4168 | 2.8921 |