Scholarlink

<p>DC microgrids are being increasingly adopted for their effectiveness in integrating renewable energy sources and energy storage technologies. However, their dependence on communication and control systems makes them susceptible to cyber-attacks. This research introduces a detailed architecture for detecting cyber-attacks in DC microgrids utilizing Long Short-Term Memory (LSTM) networks, augmented by an Unknown Input Observer (UIO) system. The study explores the past applications of LSTM in DC microgrids, outlines the components and signals of interconnected DC microgrids, and introduces a mathematical model for input–output interactions. The proposed system is evaluated on four interrelated DC microgrids, demonstrating high accuracy in identifying attacks. The results are analyzed, leading to conclusions regarding the approach’s effectiveness.</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

Cyber Attack Detection on DC Microgrids with the Use of an Unknown Input Observer (UIO) and Long Short-Term Memory (LSTM)

  • 期刊论文

  • 发表日期：

    2025年10月17日

查看全文
  • Hamza Mokrane,
  • Ahcene Abed,
  • Abdelghafour Herizi,
  • Mohamed Ould Zmirli

摘要

DC microgrids are being increasingly adopted for their effectiveness in integrating renewable energy sources and energy storage technologies. However, their dependence on communication and control systems makes them susceptible to cyber-attacks. This research introduces a detailed architecture for detecting cyber-attacks in DC microgrids utilizing Long Short-Term Memory (LSTM) networks, augmented by an Unknown Input Observer (UIO) system. The study explores the past applications of LSTM in DC microgrids, outlines the components and signals of interconnected DC microgrids, and introduces a mathematical model for input–output interactions. The proposed system is evaluated on four interrelated DC microgrids, demonstrating high accuracy in identifying attacks. The results are analyzed, leading to conclusions regarding the approach’s effectiveness.