Self-Supervised Learning of Parametric Approximation for Security-Constrained DC-OPF

TL;DR

This paper presents a self-supervised, graph neural network-based method for approximating security-constrained power flow problems, achieving high accuracy and efficiency without labeled data.

Contribution

It introduces a novel self-supervised framework that preserves physical structure and predicts demand-dependent parameters for secure power system optimization.

Findings

High dispatch accuracy in benchmark tests

Low cost approximation error

Outperforms semi-supervised and end-to-end baselines

Abstract

This paper introduces a self-supervised learning framework for approximating the Security-Constrained DC Optimal Power Flow (SC-DCOPF) problem using a parametric linear model. The approach preserves the physical structure of the DC-OPF while incorporating demand-dependent tunable parameters that scale transmission line limits. These parameters are predicted via a Graph Neural Network and optimized through differentiable layers, enabling direct training from contingency costs without requiring labeled data. The framework integrates pre- and post-contingency optimization layers into an implicit loss function. Numerical experiments on benchmark systems demonstrate that the proposed method achieves high dispatch accuracy, low cost approximation error, and strong data efficiency, outperforming semi-supervised and end-to-end baselines. This scalable and interpretable approach offers a promising solution for real-time secure power system operations.