为促进能源互联网环境下海量信息与能源安全的一体化,提出了一种基于电网安全运行仿真大数据的知识管理方法。首先利用电网运行大数据,建立考虑了气象因素这一非电气量因素的电网运行状态描述;接着应用支持向量机模型,依照关键断面是否出现对电网安全运行知识进行分层;然后应用聚类分析方法,对海量电网运行状态进行聚类,形成不同形式的精细规则,建立电网安全运行知识库;在此基础上,将实时电网运行数据与历史数据进行实时匹配,从而实现快速、准确的电网超前安全预警。IEEE 9节点系统和广东省网实际系统的算例结果验证了所提方法的有效性。
To promote the combination of massive amounts of information and energy security, this paper proposes a power grid security operation knowledge management method based on big simulation data for energy internet. Firstly establish a power grid operational state model considering weather based on big operation data; secondly apply SVM to divide power grid security operation knowledge into different hierarchies according to appearance of critical power flow interfaces; thirdly, apply clustering analysis method to cluster massive power grid operational states, form different fine operational rules, and establish a knowledge base. Finally, match real-time power grid operational states with historical operational states promptly, and achieve power grid security early warning fast and accurately. Applied in IEEE 9 power system and Guangdong power system, the method in the paper shows its effectiveness.