Scholarlink

<p>The integration of Artificial Intelligence (AI) into the management and control of smart grids is advancing rapidly, with Large Language Models (LLMs) emerging as a cutting-edge solution for Industry 4.0 applications. However, the trustworthiness of LLMs remains a critical concern in managing these systems. This paper investigates the behavior of various locally hosted LLMs in addressing unauthorized operations within smart grids. To emphasize the benefits of local hosting, we compare the performance of these models with online LLMs. Our results demonstrate that locally hosted LLMs outperform their online counterparts in mitigating vulnerabilities. This study introduces an educational platform designed to train future smart grid workforce and outlines essential guidelines for incorporating LLM vulnerability awareness into smart grid management curriculum. Additionally, the findings offer valuable insights for shaping standardization, legislation, and regulatory frameworks for AI applications in smart grids.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

SmartGridAgent: An Educational Framework for Reliable Digital Twin-Based Smart Grid Workforce Training with Locally Hosted LLMs

  • 期刊论文

  • 发表日期：

    2025年5月30日

查看全文
  • Ali Imanifard,
  • Babak Majidi,
  • Abdolah Shamisa

摘要

The integration of Artificial Intelligence (AI) into the management and control of smart grids is advancing rapidly, with Large Language Models (LLMs) emerging as a cutting-edge solution for Industry 4.0 applications. However, the trustworthiness of LLMs remains a critical concern in managing these systems. This paper investigates the behavior of various locally hosted LLMs in addressing unauthorized operations within smart grids. To emphasize the benefits of local hosting, we compare the performance of these models with online LLMs. Our results demonstrate that locally hosted LLMs outperform their online counterparts in mitigating vulnerabilities. This study introduces an educational platform designed to train future smart grid workforce and outlines essential guidelines for incorporating LLM vulnerability awareness into smart grid management curriculum. Additionally, the findings offer valuable insights for shaping standardization, legislation, and regulatory frameworks for AI applications in smart grids.