基于数字孪生的电力系统自动化管理系统研究

Research on Power System Automation Management System Based on Digital Twins

电力系统复杂多变的运行模式增加了其自动化管理难度。因此,搭建了一套基于数字孪生的电力系统自动化管理系统,其系统架构包含感知层、数据层、运算层、功能层、应用层,其将电力系统的数字孪生模型与实际物理世界相结合,实现了实时自动化的监测、分析和控制。为了使电力系统自动化管理系统更加可靠有效,提出了基于知识库和基于数据驱动两种电力系统异常故障处理流程,并分析了其优劣。结果表明,二者各有优劣,基于知识库的故障处理流程,适合处理已知的、历史上频繁出现的故障类型,执行效率较高,对专家经验和知识有很大的依赖。基于数据驱动的故障处理流程,其未知故障诊断能力十分优越,但是需要大量样本数据的支持,硬件投入及算法优化的工作较为复杂。

The complex and changeable operation mode of power system increases the difficulty of its automation management.Therefore,a set of power system automation management system based on digital twin is constructed,and its system architecture contains five layers:perception layer,data layer,computing layer,function layer,and application layer,which combines the digital twin model of the power system with the actual physical world to achieve real-time automated monitoring,analysis,and control.In order to make the power system automation management system more reliable and effective,two kinds of power system abnormal fault handling processes based on knowledge base and based on data-driven are proposed,and their advantages and disadvantages are analysed.The results show that both have their advantages and disadvantages,and the knowledge base-based fault handling process,which is suitable for dealing with known and historically frequent fault types,is more efficient in execution and relies heavily on expert experience and knowledge.The data-driven fault handling process based on the unknown fault diagnosis capability is very superior,but requires the support of a large amount of sample data,hardware investment and algorithm optimisation is more complicated.