Optimal Power Flow (ACOPF) on ACOPF118 (test)
0Optimality GapIPOPT
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| IPOPT2025.02 | 0 | 100 | 13.11 | 0.97 | 0 | 0 | 0 | — | — | |
| MSE + LG-ND (L)Hidden Width=50 (2x), Post-processing=Clipped2026.06 | 0.0691 | — | — | — | 0 | — | — | 0.0399 | — | |
| LG-NDHidden Width=50 (2×), Parameters=31.9K2026.06 | 0.0691 | — | — | — | 0 | — | — | 0.0399 | 1 | |
| Naïve MSE ||·||2Hidden Width=472 (2x), Post-processing=Clipped, Number of training samples (Ntrain)=200002026.06 | 0.7118 | — | — | — | 0 | — | — | 0.2062 | — | |
| Naïve MSEHidden Width=472 (2×), Parameters=498K2026.06 | 0.7118 | — | — | — | 0 | — | — | 0.2062 | 15.58 | |
| MAE + PenaltyHidden Width=472 (2x), Post-processing=Clipped, Number of training samples (Ntrain)=200002026.06 | 0.8055 | — | — | — | 0 | — | — | 0.2051 | — | |
| Penalty MAEHidden Width=472 (2×), Parameters=498K2026.06 | 0.8055 | — | — | — | 0 | — | — | 0.2051 | 15.58 | |
| MSE + PenaltyHidden Width=472 (2x), Post-processing=Clipped, Number of training samples (Ntrain)=200002026.06 | 0.919 | — | — | — | 0 | — | — | 0.2199 | — | |
| Penalty MSEHidden Width=472 (2×), Parameters=498K2026.06 | 0.919 | — | — | — | 0 | — | — | 0.2199 | 15.58 | |
| Naïve MAE ||·||1Hidden Width=472 (2x), Post-processing=Clipped, Number of training samples (Ntrain)=200002026.06 | 1.1598 | — | — | — | 0 | — | — | 0.1253 | — | |
| Naïve MAEHidden Width=472 (2×), Parameters=498K2026.06 | 1.1598 | — | — | — | 0 | — | — | 0.1253 | 15.58 | |
| DiOptKe=642025.02 | 2.26 | 84.33 | 13.41 | 0.07 | 0 | 0.01 | 0.2 | — | — | |
| DC3correction steps=2002025.02 | 2.49 | 43 | 13.44 | 0.06 | 0 | 0.02 | 0.73 | — | — | |
| RectFlow2025.02 | 2.9 | 0 | 13.49 | 0.07 | 0 | 0.38 | 9.1 | — | — | |
| MBD2025.02 | 37.39 | 0 | 17.86 | 5.34 | 0.03 | 4.5 | 23.16 | — | — |