PowerModelsGAT-AI: Physics-Informed Graph Attention for Multi-System Power Flow with Continual Learning

TL;DR

PowerModelsGAT-AI introduces a physics-informed graph attention network for real-time power flow prediction, demonstrating high accuracy across multiple systems and effective continual learning with minimal forgetting.

Contribution

The paper presents a novel physics-informed graph attention network that generalizes across multiple power systems and incorporates continual learning strategies to prevent catastrophic forgetting.

Findings

Achieves less than 1% voltage magnitude error on benchmark systems.

Maintains high accuracy during continual learning with less than 2% error increase.

Attention weights correlate with physical parameters, supporting interpretability.

Abstract

Solving the alternating current power flow equations in real time is essential for secure grid operation, yet classical Newton-Raphson solvers can be slow under stressed conditions. Existing graph neural networks for power flow are typically trained on a single system and often degrade on different systems. We present PowerModelsGAT-AI, a physics-informed graph attention network that predicts bus voltages and generator injections. The model uses bus-type-aware masking to handle different bus types and balances multiple loss terms, including a power-mismatch penalty, using learned weights. We evaluate the model on 14 benchmark systems (4 to 6,470 buses) and train a unified model on 13 of these under N-2 (two-branch outage) conditions, achieving an average normalized mean absolute error of 0.89% for voltage magnitudes and R^2 > 0.99 for voltage angles. We also show continual learning: when adapting a base model to a new 1,354-bus system, standard fine-tuning causes severe forgetting with error increases exceeding 1000% on base systems, while our experience replay and elastic weight consolidation strategy keeps error increases below 2% and in some cases improves base-system performance. Interpretability analysis shows that learned attention weights correlate with physical branch parameters (susceptance: r = 0.38; thermal limits: r = 0.22), and feature importance analysis supports that the model captures established power flow relationships.