LLMs and Agentic AI Systems for Smart Grids: A Tutorial on Architectures and Applications

2026-07-20

AI Summary

This work addresses the challenge that large language model (LLM) outputs in smart grid applications often violate physical constraints and lack a unified design and evaluation framework. To ensure reliability, the authors propose the “solver-anchored” principle, which clearly delineates the responsibilities between LLMs and optimization solvers such as CVXPY. They develop a trustworthy system encompassing prompt engineering, agent architecture, tool invocation, and result verification. A four-dimensional evaluation framework is introduced to guarantee the physical correctness and safety of generated solutions. Experimental results demonstrate that EVAgent reproduces CVXPY’s optimal solutions and reduces unmet energy demand by 7.5–9.5×, while GridDebugAgent successfully resolves 17 out of 39 fault cases, achieving a 52.3% reduction in total constraint violations.

Abstract

Large language models (LLMs) and agentic AI systems have evolved from natural language tasks to using external tools to plan, retrieve, and act in technical domains. In smart grids, recent work applies agentic schemes to forecasting, optimization, and control, wrapping trusted solvers behind language interfaces and orchestrating multi-step workflows. The literature lacks a unified approach to designing and evaluating such systems. LLMs can produce numerically plausible yet physically infeasible outputs, evaluation protocols vary across tasks, and the boundary between what the model should and should not compute is implicit. This paper presents a solver-grounded design principle: a numerical result is reported only when it originates from a trusted tool and passes explicit verification. We review the building blocks of LLM and agentic AI systems for power systems: prompting strategies and agentic architectures. We instantiate the principle in four case studies: wind power forecasting, EV charging scheduling, power flow analysis, and contingency diagnosis, each comparing an LLM-only baseline against its solver-grounded counterpart on identical data and metrics. EVAgent reproduces the CVXPY optimum while reducing LLM-only unmet energy by 7.5-9.5x, and GridDebugAgent repairs 17/39 contingency cases while reducing total violations by 52.3%. We propose a four-group evaluation framework spanning task utility, solver-grounded correctness, faithfulness and safe failure, and cost and latency. A consistent division of labor emerges: the agentic system reliably orchestrates, retrieves, and explains, while trusted tools compute and a verification gate decides what is reported.

Problem

• Large Language Models
• Agentic AI Systems
• Smart Grids
• Solver-grounded Design
• Evaluation Framework

Innovation

• solver-grounded design
• agentic AI systems
• large language models
• smart grids
• verification gate