Table 6 CRPS values in multivariate case for the three models (Q-CNN-GRU, Q-LSTM, Q-GRU) across all four seasons (summer, winter, fall, spring) and forecasting horizons (1 h, 6 h, 12 h, 24 h) for hebei, China.
From: Seasonal quantile forecasting of solar photovoltaic power using Q-CNN-GRU
Time | Model | Summer | Spring | Fall | Winter |
|---|---|---|---|---|---|
1 h | Q-CNN-GRU | 0.0301 | 0.0385 | 0.0395 | 0.0410 |
Q-LSTM | 0.0430 | 0.0445 | 0.0460 | 0.0475 | |
Q-GRU | 0.0345 | 0.0420 | 0.0425 | 0.0460 | |
6 h | Q-CNN-GRU | 0.0685 | 0.0740 | 0.0775 | 0.1065 |
Q-LSTM | 0.0785 | 0.0835 | 0.0875 | 0.1115 | |
Q-GRU | 0.0715 | 0.0780 | 0.0810 | 0.1045 | |
12 h | Q-CNN-GRU | 0.0885 | 0.0920 | 0.0950 | 0.1075 |
Q-LSTM | 0.1015 | 0.1035 | 0.1065 | 0.1165 | |
Q-GRU | 0.0965 | 0.0985 | 0.1015 | 0.1110 | |
24 h | Q-CNN-GRU | 0.1005 | 0.1010 | 0.1065 | 0.1155 |
Q-LSTM | 0.1125 | 0.1175 | 0.1215 | 0.1335 | |
Q-GRU | 0.1015 | 0.1085 | 0.1135 | 0.1325 |
