Research Article
A New Hybrid Model Based on Data Preprocessing and an Intelligent Optimization Algorithm for Electrical Power System Forecasting
Table 4
Comparison and contrast of E-MFA-BP proposed with other traditional time series forecasting models.
| ā | AR | ARIMA | BPNN | GM | SVM | E-MFA-BP |
| Short-term wind speed time series | MAPE (%) | 9.83% | 8.25% | 7.38% | 4.82% | 4.03% | 3.86% | MAE | 0.779 | 0.782 | 0.631 | 0.595 | 0.367 | 0.468 | MSE | 0.734 | 0.644 | 0.568 | 0.428 | 0.422 | 0.389 |
| Electrical load time series | MAPE (%) | 4.87% | 2.85% | 3.22% | 2.51% | 1.33% | 1.23% | MAE | 316.782 | 258.966 | 230.090 | 101.438 | 92.781 | 95.919 | MSE (104) | 3.013 | 2.665 | 2.874 | 1.593 | 1.688 | 1.488 |
| Electricity price time series | MAPE (%) | 8.62% | 7.33% | 6.37% | 5.36% | 5.02% | 5.56% | MAE | 5.007 | 4.283 | 3.426 | 4.225 | 3.189 | 2.505 | MSE | 139.428 | 100.690 | 58.259 | 89.517 | 37.621 | 11.788 |
|
|