HH-MPNN achieves under 1% optimality gap on default topologies from 14 to 2000 buses, zero-shot N-1 generalization under 3% gap, and improved size generalization via pre-training on small grids.
Accelerating optimal power flow with GPUs: SIMD abstraction of nonlinear programs and condensed- space interior-point methods
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Towards Generalization of Graph Neural Networks for AC Optimal Power Flow
HH-MPNN achieves under 1% optimality gap on default topologies from 14 to 2000 buses, zero-shot N-1 generalization under 3% gap, and improved size generalization via pre-training on small grids.