Test Time Training for AC Power Flow Surrogates via Physics and Operational Constraint Refinement

2025-11-27

AI Summary

To address the insufficient physical consistency of machine learning methods in AC power flow computation, this paper proposes a physics-informed test-time training (TTT) framework. During inference, the method performs only a few gradient updates, embedding AC power flow equations and operational constraints as self-supervised signals into the optimization process—enabling online, label-free adaptive correction. Its core techniques include gradient-based lightweight TTT, physics-constraint regularization, and self-supervised output calibration. Evaluated on multiple standard power systems, the approach reduces power flow residuals and constraint violations by one to two orders of magnitude while maintaining millisecond-level solution speed. It thus achieves a unified balance among accuracy, feasibility, and computational efficiency. Moreover, the method significantly enhances model interpretability from a physical perspective and improves engineering reliability.

Technology Category

• Constraint Satisfaction and Optimization: Constraint Learning and Acquisition
• Machine Learning: Calibration & Uncertainty Quantification
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Power Flow (PF) calculation based on machine learning (ML) techniques offer significant computational advantages over traditional numerical methods but often struggle to maintain full physical consistency. This paper introduces a physics-informed test-time training (PI-TTT) framework that enhances the accuracy and feasibility of ML-based PF surrogates by enforcing AC power flow equalities and operational constraints directly at inference time. The proposed method performs a lightweight self-supervised refinement of the surrogate outputs through few gradient-based updates, enabling local adaptation to unseen operating conditions without requiring labeled data. Extensive experiments on the IEEE 14-, 118-, and 300-bus systems and the PEGASE 1354-bus network show that PI-TTT reduces power flow residuals and operational constraint violations by one to two orders of magnitude compared with purely ML-based models, while preserving their computational advantage. The results demonstrate that PI-TTT provides fast, accurate, and physically reliable predictions, representing a promising direction for scalable and physics-consistent learning in power system analysis.

Problem

• Enhancing ML-based power flow surrogate accuracy and feasibility
• Enforcing AC power flow equalities and operational constraints at inference
• Reducing power flow residuals and constraint violations in power systems

Innovation

• Physics-informed test-time training for AC power flow surrogates
• Lightweight self-supervised refinement via few gradient updates
• Enforces AC power flow equalities and operational constraints at inference