Tag
This paper studies scaling laws for physics-aware surrogate models in AC optimal power flow, demonstrating that physics-informed training objectives improve constraint satisfaction at scale with different scaling rates than standard loss functions.
This paper introduces RestoreBench, a benchmark for evaluating LLM agents in diagnosing and resolving non-convergent power flow cases in power systems, covering multiple architectures and test cases.
PowerAtlas is an LLM-agent framework for jointly scheduling electricity and computing in data centers, ensuring grid feasibility and task SLAs. Validated with a real power utility and a new benchmark (ECBench) of 2,000 instances, it shows consistent gains across multiple open-weight LLMs.
FMOPF uses latent flow matching with constraint-aware interaction priors to generate diverse, feasible near-optimal solutions for AC optimal power flow, scaling to hundreds of buses while preserving feasibility.
This paper presents a machine learning surrogate model to predict component criticality in interdependent power and communication networks, achieving high correlation with a high-fidelity simulator while being computationally efficient.
This paper proposes a decision-focused generative framework for correlated scenario generation in distributionally robust optimization for grid dispatch, optimizing scenarios based on downstream operational cost rather than forecast accuracy. It reduces operational cost by 0.80–2.02% compared to accuracy-oriented methods across different generative models.
This paper proposes a Cycle-Space Detector (CSD) for detecting blind false data injection attacks on power systems, where an autoencoder generates stealthy perturbations aligned with the measurement Jacobian null space. The CSD uses topology-derived cycle constraints to improve detection without requiring precise line parameters.
The article introduces Newton's Lantern, a reinforcement learning framework for finetuning warm start models to solve the AC power flow problem more efficiently, particularly near voltage collapse.