面向智能电网的空间隐蔽型恶性数据注入攻击在线防御研究  

Online Defense Research of Spatial-hidden Malicious Data Injection Attacks in Smart Grid

作　　者：刘鑫蕊[1] 吴泽群 LIU Xinrui;WU Zequn(College of Information Science and Engineering(Northeastern University),Shenyang 110819,Liaoning Province,China)

机构地区：[1]东北大学信息科学与工程学院,辽宁省沈阳市110819

出　　处：《中国电机工程学报》2020年第8期2546-2558,共13页Proceedings of the CSEE

基　　金：国家自然科学基金项目(61573094,61703289)。

摘　　要：智能电网的安全运行高度依赖信息环节功能所提供的强大技术保障,致使电网在运行过程中易受到恶性数据注入等网络攻击的威胁,其中空间隐蔽型恶性数据注入攻击是最普遍的一种。为保证该类恶性数据注入攻击在电网运行中能被高效实时检测处理,提出一套面向监视控制与数据采集(supervisory control and data acquisition,SCADA)和相量测量单元(phasor measurement unit,PMU)混合量测的智能电网恶性数据在线防御流程。首先通过历史状态量获取与状态预测实现状态量挖掘,再进行SCADA仪表与PMU量测量的恶性数据检测、剔除与修正。此外,该文提出一种适用于混合量测系统的多重匹配状态预测方法,其预测结果作为状态参考用以打破恶性数据隐蔽性。IEEE-14和IEEE-118节点测试系统仿真结果验证了所提方法预测准确性及在线检测空间隐蔽型恶性数据的有效性。The safe operation of smart grid relies heavily on the powerful technical support provided by cyber part, which makes it vulnerable to cyber-attacks such as malicious data injection during operation, of which the spatial-hidden malicious data injection attack is the most common one. In order to effectively detect and process it in a real-time system, a novel online defense process was proposed against malicious data based on hybrid SCADA and PMU measurements. Firstly, the mining of state variables was realized by historical state acquisition and state forecast, then the malicious data of SCADA and PMU measurements was detected, eliminated and corrected. Besides, a multiple matching state forecast method for hybrid measurement system was proposed, which is regarded as state reference to realize stealthiness corruption. Numerical simulation results on the IEEE 14-bus and 118-bus systems have demonstrated the forecast accuracy and effectiveness of the proposed method against spatial-hidden malicious data.

关 键 词：电网信息物理系统 空间隐蔽型恶性数据 在线防御流程 监视控制与数据采集(SCADA)和相量测量单元(PMU)混合量测 多重匹配状态预测 

分 类 号：TM734[电气工程—电力系统及自动化]