面向“双碳”目标的光储一体化建筑集群电力协同调度方法研究综述

Review on Power Cooperative Scheduling Method for Building Cluster Integrated PV and Storage Dual to the Carbon Peaking and Carbon Neutrality Goals

为实现建筑用能领域的“双碳”目标,需将大规模的城乡建筑集群配置为光储一体化建筑集群,同时考虑到光伏发电的间歇性和随机性,还需实现此类建筑集群的大规模电力协同调度,从而有效的消纳可再生能源。通过文献综述对比了现有发用储融合的用电模式,总结了光储一体化建筑集群电力协同调度的概念和优势,进一步对比分析了针对光储一体化建筑集群电力系统的协同调度方法,认为基于多智能体强化学习的去中心化电力调度方法具有灵活性强、适用度高等特点,并进一步梳理多智能体强化学习应用在电力协同调度中的关键问题,提出基于数字孪生技术能在数据涌现进行算法训练层面有效助力实现基于多智能体强化学习的光储一体化建筑集群电力协同调度。

In order to achieve the Carbon Peaking and Carbon Neutrality Goals in the field of buildings energy use,large-scale urban and rural building clusters should be configured as building cluster integrated PV and storage.Considering the intermittence and randomness of PV power generation,it is necessary to implement the large-scale collaborative power dispatching in such building cluster to absorb renewable energy effectively.The existing power consumption modes with the integration of power generation,use and storage are compared,and the concept and advantages of collaborative power dispatching in building cluster integrated PV and storage are summarized.The collaborative scheduling methods for the power system of such building cluster are analyzed,and it is considered that the decentralized power scheduling method based on Multi-Agent Reinforcement Learning(MARL)has strong flexibility and applicability.Furthermore,the key problems in the application of MARL in power collaborative scheduling are summarized.It is proposed that digital twin technology can effectively help realize power cooperative scheduling of building cluster integrated PV and storage based on MARL at the level of algorithm training on data emergence.