基于子空间跟踪算法的盲虚假数据攻击  

Blind False Data Injection Attacks Based on Subspace Tracking Algorithm

作　　者：邓彬 饶爽 胡亚荣 赵晶玲 孟琦 Deng Bin;Rao Shuang;Hu Yarong;Zhao Jingling;Meng Qi(Shenzhen Power Supply Bureau Co.,Ltd.,Shenzhen Guangdong 518000,China;Faculty of School of Electrical and Information Engineering,Xi’an Jiaotong University,Xi’an Shaanxi 710049,China;Research Institution of Data Communications Science and Technology,Beijing 100191,China)

机构地区：[1]深圳供电局有限公司,广东深圳518000 [2]西安交通大学电信学部,陕西西安710049 [3]数据通信科学技术研究所,北京100191

出　　处：《电气自动化》2023年第2期82-84,共3页Electrical Automation

基　　金：国家自然科学基金资助项目(62071370)。

摘　　要：虚假数据注入攻击是电力系统网络安全领域的热门课题。当电网保持稳定时,盲虚假数据注入攻击在未知电网雅可比矩阵的情况下,依旧能绕过坏数据检测系统,诱导系统状态估计出错进行错误的调度,从而破坏电网的稳定安全,但该算法在估计雅可比矩阵时,存在运算复杂度高、耗时长的问题。为此,提出了基于紧缩投影逼近子空间跟踪算法的新盲虚假数据攻击算法,以递推更新的方式构造出近似的雅可比矩阵。仿真试验表明:在IEEE 14系统上新算法运算时间仅为传统主成分分析法的1/2,随着电网规模增大这个数字还会下降。所提出的算法较传统算法简化了运算量,提升了攻击成功率。False data injection attack(FDIA)is a hot topic in the field of power system network security.When the power grid remains stable,the blind false data injection attack can still bypass the bad data detection system in the case of unknown power grid Jacobian matrix,induce the system state estimation error and perform wrong scheduling,so as to destroy the stability and security of the power grid.However,the algorithm has the problems of high computational complexity and long time when estimating the Jacobian matrix.Therefore,a new blind false data attack algorithm based on the compact projection approximation subspace tracking algorithm was proposed,and the approximate Jacobian matrix was constructed by recursive update.The simulation experiments show that the operation time of the new algorithm on the IEEE 14 system is only 1/2 of that of the traditional principal component analysis method,and this number will decrease as the power grid scale increases.Compared with the traditional algorithm,the proposed algorithm simplifies the computation and improves the attack success rate.

关 键 词：虚假数据注入攻击 子空间跟踪算法 状态估计 攻击向量构造 坏数据检测 

分 类 号：TM711[电气工程—电力系统及自动化]