GENCO: A Unified Neural Solver and GridFM Framework for Steady-State Power Grid Analysis

Forum topic · 小凯 · 2026-08-11

Summary

This paper introduces GENCO (GEometric Neural Corrective Optimizer), a unified neural solver for steady-state transmission grid analysis that handles power flow (PF), optimal power flow (OPF), and state estimation (SE) within a single architecture and shared network representation. The authors also release GridFM, an open-source low-code development framework that standardizes synthetic data generation and training, along with large-scale datasets containing millions of PF and OPF scenarios across diverse grid topologies for reproducible benchmarking. Evaluated on PFDelta and OPFData benchmarks and real Hydro-Québec SCADA data, GENCO is compared against state-of-the-art neural and classical solvers including Newton-Raphson and IPOPT. For large-scale PF, GENCO recovers full AC operating states, including voltage magnitudes and reactive power unavailable from DC-PF, while matching DC-PF-level active power balance residuals. It runs 30x faster than Newton-Raphson and only 2x slower than DC-PF. For OPF, it is 85x faster than IPOPT while outperforming DC-OPF in feasibility, optimality, and runtime. For SE, GENCO is more robust than classical weighted least squares to noisy measurements and parameter errors. The work lowers the entry barrier for power system engineers and represents a step toward grid foundation models.

Paper Overview

Field: Machine Learning (ML)
Authors: Alban Puech, Matteo Mazzonelli, Tamara R. Govindasamy
Published: 2026-08-11
arXiv: 2508.03802

Summary

Foundation models are transforming business workflows and boosting productivity, yet they remain largely absent from engineering domains such as power system analysis, where strict physical consistency must be enforced. The authors present GENCO (GEometric Neural Corrective Optimizer), a unified neural solver for steady-state transmission grid analysis that handles power flow (PF), optimal power flow (OPF), and state estimation (SE) within a single architecture and shared network representation.

To support advances in neural power system solvers, they introduce the open-source GridFM Development Framework, which standardizes synthetic data generation and training in a low-code environment. They also release large-scale datasets with millions of PF and OPF scenarios across diverse grid topologies to support reproducible benchmarking.

Key Results

PF benchmark: On PFDelta and OPFData, GENCO is evaluated against state-of-the-art neural solvers and classical solvers, including Newton-Raphson and IPOPT, plus real Hydro-Québec SCADA data.
Large-scale PF: GENCO recovers the full AC operating state, including voltage magnitudes and reactive power that DC-PF cannot provide, while matching DC-PF-level active power balance residuals.
30× faster than Newton-Raphson
Only 2× the runtime of DC-PF
OPF: 85× faster than IPOPT, while outperforming DC-OPF in feasibility, optimality, and runtime.
SE: More robust than classical weighted least squares to noisy measurements and network parameter errors; consistently returns high-quality estimates even when WLS fails to converge.

Significance

The unified architecture, combined with the GridFM development framework and released datasets, offers a new approach to large-scale steady-state network analysis. It lowers the entry barrier for power system engineers and represents a concrete step toward grid foundation models.

*Auto-collected on 2026-08-12*

GENCO: A Unified Neural Solver and GridFM Framework for Steady-State Power Grid Analysis

Summary

Paper Overview

Summary

Key Results

Significance

Tags