SafePowerGraph-LLM: Novel Power Grid Graph Embedding and Optimization with Large Language Models

2025-01-13

AI Summary

Solving optimal power flow (OPF) in large-scale, complex power systems faces challenges in scalability, robustness, and accurate modeling of physical constraints. Method: This paper introduces large language models (LLMs) to power system optimization for the first time, proposing a hybrid encoding framework that jointly leverages graph-structured and tabular representations. The framework integrates power grid topology embeddings, an OPF-specific in-context learning protocol, and domain-adaptive fine-tuning to enforce real-world component-level physical constraints. Contribution/Results: The resulting LLM-driven graph optimization framework is the first to natively support detailed physical constraints in OPF. Experiments on standard benchmark systems demonstrate high solution accuracy and real-time inference capability, broad compatibility with mainstream commercial LLMs, and significant improvements in computational efficiency, generalization across network topologies, and adaptability to operational uncertainties.

Technology Category

• Search and Optimization: Learning to Search
• Constraint Satisfaction and Optimization: Constraint Optimization
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Efficiently solving Optimal Power Flow (OPF) problems in power systems is crucial for operational planning and grid management. There is a growing need for scalable algorithms capable of handling the increasing variability, constraints, and uncertainties in modern power networks while providing accurate and fast solutions. To address this, machine learning techniques, particularly Graph Neural Networks (GNNs) have emerged as promising approaches. This letter introduces SafePowerGraph-LLM, the first framework explicitly designed for solving OPF problems using Large Language Models (LLM)s. The proposed approach combines graph and tabular representations of power grids to effectively query LLMs, capturing the complex relationships and constraints in power systems. A new implementation of in-context learning and fine-tuning protocols for LLMs is introduced, tailored specifically for the OPF problem. SafePowerGraph-LLM demonstrates reliable performances using off-the-shelf LLM. Our study reveals the impact of LLM architecture, size, and fine-tuning and demonstrates our framework's ability to handle realistic grid components and constraints.

Problem

• Large Language Models
• Optimization of Power Networks
• Optimal Power Flow (OPF)

Innovation

• Large Language Models
• Optimal Power Flow
• Grid Graph Integration