Edge Cloud Collaboration Framework and Key Technology for Flexible Interconnected Distribution Network

As advancements in new power system construction continue, the demand for AC/DC hybrid connections and bidirectional interaction between source and load in traditional distribution network (DN) is increasingly significant. Building the flexible interconnected distribution network (FIDN) by energy router and other devices can effectively improve the control flexibility of DN, which supports the safe, controllable and economic operation. However, traditional centralized optimization encounters challenges associated with dimensional disaster and privacy disclosure, while distributed optimization faces difficulties in rapid convergence within complex systems. This paper focuses on the optimization decision-making of FIDN based on edge-cloud collaboration. It analyzes the characteristics of edge-cloud deployment mode, optimization objective, interactive instruction and edge-cloud collaboration model solution. On this basis, it proposes a framework of edge-cloud collaboration optimization for FIDN and a key technology system based on cloud-centric meta reinforcement learning (RL) and edge-end distributed collaborative control. Additionally, the interaction process of edge-cloud collaboration is also analyzed. It provides systematic guidance for operation optimization and engineering application for FIDN.