A Hybrid Energy Storage‐Based 800 V DC Distribution Architecture and Hierarchical Rolling Scheduling Strategy for Intelligent Data Centres

ABSTRACT Intelligent computing loads, which are characterised by high‐power density and impulsive, pose a significant challenge to high‐power impact resilience and fast dynamic response capability of the internal power supply system in data centres. To tackle this challenge, this paper proposes an 800 V HVDC distribution architecture integrated with hybrid energy storage system (HESS), together with a hierarchical rolling scheduling strategy for HESS. First, the architecture and the scheduling framework are proposed, where the architecture features distributed deployment of HESS and a design of integrated backup and energy storage batteries. Then, the modelling for the HESS hierarchical rolling scheduling strategy is presented across month‐ahead, minute‐ahead and second‐ahead layers. Herein, the strategy mainly differs from existing approaches in its tight coupling with intelligent computing load and the power supply architecture. Simulation results demonstrate the practical potential of the proposed architecture and strategy, which collectively suppress high‐frequency fluctuations and shape low‐frequency components of intelligent computing load profiles; achieve economic efficiency by leveraging existing buses, lines and supply equipment capacities; and embed load‐side flexibility, supporting grid‐friendly operation and compliance with national policies on artificial intelligence and power grid coordinated operation.