Learning-based AC-OPF Solvers on Realistic Network and Realistic Loads

TL;DR

This paper introduces TAS-97, a realistic dataset for AC-OPF problems, and develops deep learning models that significantly improve solution speed and accuracy on real-world power network data.

Contribution

The creation of TAS-97 dataset with realistic network and load data, and the development of feasibility-optimized deep neural network AC-OPF solvers trained on this data.

Findings

Achieved 0.13% cost optimality gap

99.73% feasibility rate

38.62 times speedup over PYPOWER

Abstract

Deep learning approaches for the Alternating Current-Optimal Power Flow (AC-OPF) problem are under active research in recent years. A common shortcoming in this area of research is the lack of a dataset that includes both a realistic power network topology and the corresponding realistic loads. To address this issue, we construct an AC-OPF formulation-ready dataset called TAS-97 that contains realistic network information and realistic bus loads from Tasmania's electricity network. We found that the realistic loads in Tasmania are correlated between buses and they show signs of an underlying multivariate normal distribution. Feasibility-optimized end-to-end deep neural network models are trained and tested on the constructed dataset. Trained on samples with bus loads generated from a fitted multivariate normal distribution, our learning-based AC-OPF solver achieves 0.13% cost optimality gap, 99.73% feasibility rate, and 38.62 times of speedup on realistic testing samples when compared to PYPOWER.