计及云边协同的综合能源系统双层优化

多区域互联的综合能源系统具备两大优势：区域内部多能互补与区域之间能量互济，本文针对综合能源系统设计了一种云边协同调控方法以兑现这两大优势。首先，本文设计了能量管理云平台-边缘控制器双层架构分别统筹子区域内和子区域间的能量流，子区域以预测控制方法(model predictive control, MPC)实现边缘控制器内部能量的实时调控。其次，基于所提双层架构设计了包括集中调控层-分布协同层的双层能量优化模型，集中调控层负责区域间能量互济以降低整个系统对外部配电网的能量需求，提高系统运行经济性；分布协同层负责控制区域内可控机组以降低实际运行工况与日前调度计划值的偏差，降低系统运行波动性。然后，研究了交替方向乘子法(alternating direction method of multipliers, ADMM)求解所提双层能量优化模型的具体步骤。最后，仿真验证了所提云边协同能量调控方法对多区域互联综合能源系统经济性与稳定性的改善作用。

Bi-level optimization for integrated energy system considering cloud-edge collaboration

The multi-area interconnected integrated energy system has two major advantages: multi-energy complementarity within the region and energy complementarity between regions. This paper proposes a operation method considering cloud-edge collaboration for integrated energy system to realize the advantages. Firstly, this paper designs the bi-level architecture of cloud and edge controller to coordinate the energy flow between sub-regions and within sub-region respectively. The model predictive control method (model predictive control, MPC) is used to realize the real-time energy regulation of the edge controller. Secondly, based on the proposed bi-level architecture, a bi-level energy optimization model including a centralized regulation layer and a distributed coordination layer is designed. The optimation target of centralized regulation layer is minimizing the energy demand of the entire system from the external distribution network to improve the economy. The optimation target of distribution coordination layer is minimizing the energy deviation between the actual power and the planned schedulings power of the controllable units within the area to reduce the fluctuation. Then, the specific steps of the alternating direction method of multipliers (ADMM) to solve the proposed bi-level energy optimization model are studied. Finally, the simulation verifies the economy and stability improvements of the multi-area interconnected integrated energy system by adapting the proposed operation method considering cloud-edge collaboration.