Scholarlink

<p>The integration of distributed energy resources (DERs) and flexible loads is transforming modern power systems into complex, interactive networks. While improving sustainability, this trend introduces significant challenges in real-time scheduling and exposes systems to cyber-physical risks. This paper proposes a distributed smart grid (DSG) collaborative scheduling framework based on reflective reasoning and distributed mitigation (DLT). The framework equips agents with reflective reasoning to assess information credibility, detect anomalies, and mitigate attacks autonomously while optimizing operations. A hierarchical edge-cloud architecture and a global–local reward design ensure coordination, efficiency, and security. Simulations on IEEE and real-world test systems demonstrate that our method outperforms benchmarks in latency, energy cost, voltage stability, control rate, and power loss under diverse conditions, validating its robustness and scalability for secure, real-time smart grid operation.</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

A reflective reasoning and distributed mitigation-based collaborative scheduling framework for secure and resilient smart grids

  • 期刊论文

  • 发表日期：

    2025年12月23日

查看全文
  • Dikai Deng,
  • Hushuang Zeng

摘要

The integration of distributed energy resources (DERs) and flexible loads is transforming modern power systems into complex, interactive networks. While improving sustainability, this trend introduces significant challenges in real-time scheduling and exposes systems to cyber-physical risks. This paper proposes a distributed smart grid (DSG) collaborative scheduling framework based on reflective reasoning and distributed mitigation (DLT). The framework equips agents with reflective reasoning to assess information credibility, detect anomalies, and mitigate attacks autonomously while optimizing operations. A hierarchical edge-cloud architecture and a global–local reward design ensure coordination, efficiency, and security. Simulations on IEEE and real-world test systems demonstrate that our method outperforms benchmarks in latency, energy cost, voltage stability, control rate, and power loss under diverse conditions, validating its robustness and scalability for secure, real-time smart grid operation.