Special Session Ⅲ

Special Session Ⅲ

Resilience Enhancement of Grid Federated Learning and Collaborative Control of Distributed Self-Healing Microgrids

电网联邦学习弹性增强与分布式自愈微电网协同控制

Chair:	Co-chair:
Tao Mao	Jie Zhang
Wuhan Donghu University, China	Wuhan Donghu University, China

Keywords: 

• Grid Federated Learning

  (电网联邦学习)

• Resilience Enhancement

  (弹性增强)

• Distributed Self-Healing Microgrid

  (分布式自愈微电网)

• Collaborative Control

  (协同控制)

• Edge Computing

  (边缘计算)

• Privacy Preservation

  (隐私保护)

Topic:

• Federated Learning Incentive Mechanisms and Knowledge Distillation Techniques for Multi-Microgrid Collaboration

  (面向多微电网协同的联邦学习激励机制与知识蒸馏技术)

Summary: 

• This special session focuses on resilience enhancement techniques for grid federated learning (FL) and collaborative control strategies for distributed self-healing microgrids. Addressing the needs for data privacy preservation in power grids and robustness of distributed energy systems, it investigates model aggregation optimization, anti-interference communication protocols, and adaptive edge device scheduling within FL frameworks to improve system resilience against extreme weather, cyberattacks, and other disruptions. Meanwhile, it explores mechanisms for distributed microgrids to achieve fault self-detection, self-isolation, and self-reconfiguration through multi-agent collaboration, combined with FL-enabled knowledge sharing and global optimization across microgrids. The goal is to construct a "decentralized-self-healing-intelligent" next-generation power grid control system. The session covers theoretical modeling, algorithm design, hardware-in-the-loop simulation, and real-world engineering validation, promoting deep integration of artificial intelligence with energy power systems.

  
  

• 本专题聚焦于电网联邦学习（Federated Learning）的弹性增强技术及其与分布式自愈微电网的协同控制策略。针对电网数据隐私保护与分布式能源系统鲁棒性需求，研究联邦学习框架下的模型聚合优化、抗干扰通信协议及边缘设备自适应调度方法，提升系统在极端天气、网络攻击等场景下的弹性恢复能力。同时，探索分布式微电网通过多智能体协同实现故障自检测、自隔离与自重构的机制，结合联邦学习实现跨微电网的知识共享与全局优化，构建“去中心化-自愈-智能”的新一代电网控制体系。专题涵盖理论建模、算法设计、硬件在环仿真及实际工程验证，推动人工智能与能源电力系统的深度融合。

DDL: Nov. 5, 2025