Reproduce “ML-OPF-Bench: Benchmarking Machine Learning for Optimal Power Flow” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Xinyi Liu, Xuan He, Danny H.K. Tsang, Yize Chen
ML-OPF-Bench compares methods under shared tests of power dispatch, including shifted, heavily loaded conditions. It evaluates prediction error, constraint violations, cost, and inference time; some methods add feasibility screening and solver-based recovery. In the 118-bus AC case, DNN active-generation MAE rises from 0.306% on typical conditions to 87.063% under API shift. This error does not measure feasibility.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.